Probe E# — E-Sharp AB
Probe is made by E# (E-Sharp AB) in Sweden. Want your own boards reviewed like this? Book a meeting

AFCK

OHWR0032 · revision R1

AFCK (OHWR0032 R1) is an open-hardware circuit board from the CERN Open Hardware Repository, designed in Altium Designer: a 14-layer board of 180.7 × 148.6 mm with 1406 components and 1416 nets, published under CERN-OHL-1.1.

review 80% walked · 35 findings: 13 major · 22 minor (223 items)

Open in the interactive viewerRead the full review

AFCK (OHWR0032 R1) circuit board, top view: copper, silkscreen and components
Top view, rendered by Probe from the design's own fabrication data.

Design review

Probe recorded 35 findings: 13 major · 22 minor (223 items) on this design. Each is shown here with its reasoning; the full review links every one to the pins, nets and datasheet pages it rests on.

  1. observationOBS-426majorOpen

    P2V0 bulk capacitors C50/C55 are 2 V-rated polymer parts on a rail regulated to about 2.01 V

    The P2V0 bulk capacitors C50 and C55 are 2 V-rated polymer parts on a rail regulated to about 2.01 V, so they run at or above their rated voltage all the time.

    The numbers

    QuantityValueSource
    NCP3170 feedback reference, min/typ/max at 25 C0.792 / 0.8 / 0.808 VNCP3170 datasheet, p.4
    Feedback divider on IC57 FB (FB_1V8)R62 = 24k9 to P2V0, R105 = 16k5 to GNDnetlist: FB_1V8
    P2V0 nominal0.8 x (1 + 24.9/16.5) = 2.007 Vderived
    P2V0 at the reference maximumabout 2.03 Vderived
    C50, C55 rating'POLY, 150uF, 2V, ESR: 9mR', no MPNBOM: C50, C55

    Why it matters

    A polymer capacitor held at or over its rated voltage loses its life and leakage margin; manufacturers ask for derating, never over-rating. P2V0 feeds the IC81 LDO (P1V8 / VCCAUX, MGTVCCAUX) and VCCAUX_IO through L23, so a capacitor failure takes the FPGA auxiliary supplies with it.

    Suggested fix: 2.5 V or higher rated parts (the FMC 2.5 V/3.3 V rails already use 4 V polymer), or confirm the fitted MPN's rating, since these BOM rows carry no MPN.

  2. observationOBS-451majorOpen

    FPGA configuration bank 14 is powered from FMC2's adjustable VADJ (PVADJ2), but the SPI configuration flash and CFGBVS are fixed at 3.3 V

    FPGA configuration bank 14 (IC79) is powered from FMC2's adjustable VADJ rail PVADJ2, but the SPI configuration flash IC13 and CFGBVS_0 are fixed at 3.3 V: configuration only works, and bank 14 is only within ratings, when VADJ2 happens to be 3.3 V.

    Where it is

    ``mermaid flowchart LR D["IC10 ch2 set through AD8402 IC11<br/>about 1.6 V to 3.3 V"] -- "PVADJ2" --> B14["IC79 bank 14 VCCO_14<br/>P30, R27, T24, U21, V28, W25"] D -- "PVADJ2" --> O["VCCO_12/13<br/>and FMC2 connector"] P["P3V3"] --> F["IC13 M25P128<br/>pin 8 VCC"] P -- "R162 10k" --> C["CFGBVS_0, ball L10<br/>declares 2.5/3.3 V"] F -- "pin 2 Q, R125 10R<br/>FSPI.FLASH_Q/D1" --> B14 ``

    • Bank 14 VCCO: the schematic's IC79 symbol (sheet 16) puts all six VCCO_14 balls (P30, R27, T24, U21, V28, W25) on PVADJ2. PVADJ2 is the FMC2 VADJ rail, set by IC10 ch2 through the AD8402 (FB2: R248 1k5 to PVADJ2, R262 453R to FB2_R; FB2_R = R260 910R to GND + IC11 pin 12), and shared with VCCO_12/13 and the FMC2 connector.
    • Bank 14 carries the master-SPI configuration pins: D00/MOSI (P24), D01/DIN (R25), D02/D03 (R20/R21), FCS_B (U19) and PUDC_B (R23).
    • Fixed at 3.3 V: IC13 VCC (pin 8, P3V3), and CFGBVS_0 (L10), pulled to P3V3 by R162 10k, i.e. declared 2.5/3.3 V.

    What happens at each VADJ2 setting

    VADJ2Effect
    3.3 VConfiguration works and bank 14 is within ratings
    2.5 VThe flash's 3.3 V Q output drives D01/DIN above the HR-bank limit of VCCO + 0.55 V = 3.05 V. The FPGA's 2.5 V MOSI/FCS_B barely meet the flash's 0.7 x VCC = 2.31 V VIH
    1.8 V, a common FMC VADJCFGBVS tied high is the configuration UG475 warns about: 'To avoid device damage, this pin must be connected correctly'

    The limits

    QuantityValueSource
    CFGBVS_0 connectionto VCCO_0 when banks 0/14/15 are 2.5 or 3.3 V, to GND when <= 1.8 VUG475 Table 1-12, p.22
    HR-bank I/O input absolute maximum-0.40 V to VCCO + 0.55 VDS182 (XC7K325T), p.1
    M25P128 VIH0.7 x VCCM25P128 datasheet, p.36

    Why it matters

    FMC2 mezzanines needing a VADJ other than 3.3 V are incompatible with configuring from IC13.

    Suggested fix: power bank 14 from P3V3 (or a fixed rail matching the flash), or document VADJ2 = 3.3 V as mandatory.

  3. observationOBS-454majorOpen

    Three 1.8 V bank-32 balls are wired to translators whose FPGA side runs at PVADJ2 (up to 3.3 V): IC41 cannot see their highs, and IC70 can drive 3.3 V into an HP ball

    Three 1.8 V bank-32 FPGA balls are wired to translators whose FPGA side runs at PVADJ2 (up to 3.3 V): IC41 cannot see their highs, and IC70 can drive 3.3 V into an HP ball rated 2.35 V absolute maximum.

    Bank 32 is an HP bank powered from PVADJ2_1V8 (1.8 V, IC2 output, 1.795 V), per the vendor package data on the interface cards and the schematic's VCCO_32 balls. PVADJ2 is the FMC2 VADJ (1.6-3.3 V) and must be 3.3 V for configuration (OBS-451).

    ``mermaid flowchart LR subgraph B32["IC79 bank 32, HP, VCCO = PVADJ2_1V8 = 1.8 V"] S1["AB14 DAC1_SYNC_N<br/>Y14 DAC2_SYNC_N<br/>AE16 CLK_UPDATEN"] S2["AE15 PVADJ2_IO10"] end S1 -- "1.8 V highs" --> A["IC41 SN74AVC8T245<br/>VCCA = PVADJ2"] C["IC70 TXB0104<br/>VCCA = PVADJ2"] -- "A4 driven to VCCA, up to 3.3 V" --> S2 R["RTM connector, RTM_IO10"] -- "B4, pin 10" --> C ``

    (a) IC41 cannot read the bank-32 highs

    IC41 (SN74AVC8T245, VCCA = PVADJ2 on pins 1-2) takes three A inputs from bank 32:

    • pin 3: PLL_CTRL.PLL_VADJ2_DAC2_SYNC_N (Y14)
    • pin 5: PLL_CTRL.PLL_VADJ2_DAC1_SYNC_N (AB14)
    • pin 9: PLL_CTRL.VADJ2_CLK_UPDATEN (AE16)

    Its other four A inputs come from HR banks 12/13 at PVADJ2.

    (b) IC70 can overdrive an HP ball

    IC70 (TXB0104, VCCA = PVADJ2 on pin 1) has A4 (pin 5) = PVADJ2_IO10 on ball AE15 in bank 32, and B4 (pin 10) = RTM_IO10 to the RTM connector.

    The numbers

    QuantityValueSource
    Bank 32 VCCOPVADJ2_1V8, 1.795 V (IC2 output)schematic: IC79 symbol, VCCO_32 balls
    AVC8T245 data-input VIH, VCCI 2.7-3.6 V2 VSN74AVC8T245 datasheet, p.5
    AVC8T245 data-input VIH, VCCI 1.95-2.7 V1.6 VSN74AVC8T245 datasheet, p.5
    HP-bank I/O input absolute maximum-0.55 V to VCCO + 0.55 VDS182 (XC7K325T), p.1
    Same limit for bank 321.8 + 0.55 = 2.35 Vderived

    Why it matters

    • (a) At VCCA = 2.7-3.6 V the AVC8T245 needs VIH >= 2 V. A 1.8 V LVCMOS output never reaches it, so the two DAC SYNCs and CLK_UPDATEN read as permanently low on the B side: both DACs selected together, update strobe stuck.
    • (b) The TXB0104 drives its A side to VCCA, so a high from the RTM on RTM_IO10 puts up to 3.3 V on an HP ball whose absolute maximum is VCCO + 0.55 V = 2.35 V: an overstress of the FPGA.

    Suggested fix: move these nets to PVADJ2 (HR) balls, or power the translators' A side from PVADJ2_1V8.

  4. EsdProtectorPlacementmajorOpen

    D7B.2 on 'USB_D_RB_P': 1 via(s) between CON2B.3 and D7B.2

    D7B (PRTR5V0U2X) protects 'USB_D_RB_P' at CON2B.3. Along the copper: 4.83 mm of track from the contact, 1 via(s) before it, on a 0.62 mm branch off the line; the line runs 6.15 mm on to R285B.1; D7B sits at CON2B, within 0.9 mm of the nearest place outside it to the contact: the track to it is mostly the escape out of CON2B's own pins, which no placement shortens.

    1 via(s) between CON2B.3 and D7B.2: the strike reaches an inner layer before it is clamped.

    Measured on the copper the ODB states: tracks, the pads they land on, and the netlist's vias. Not a straight line between centroids.

    At D7B

  5. DecapViaFanoutmajorOpen

    DecapViaFanout: 11 components×11

    11 checked by the review by Claude walker W4 (Cern session) — 11 different reasons

    At C237, C239, C241, C244, C2A, C371 and 5 more

    Each finding, with its own headline
    • C237C237 on 'VCCBRAM': its rail pad reaches its via over 3.18 mm of track, adding about 1.08 nH to its loop, past Probe's 0.5 nH starting value; put the vias at the pads

      C237 on 'VCCBRAM': its rail pad reaches its via over 3.18 mm of track, adding about 1.08 nH to its loop, past Probe's 0.5 nH starting value; put the vias at the pads. C237.1 (rail) runs 3.18 mm along its copper to its nearest via, 2.18 mm past the 1 mm beside the pad: 1.08 nH; C237.2 (ground) not measured: the pad sits in a pour of its net on its own layer, which has no track length to measure. Each run past the pad as one conductor of a pair 1.50 mm apart (the pad pitch), 0.2 l ln(D/R) nH with R 0.127 mm (the track's width is not read): 1.08 nH in all. Put the vias at the pads (or in them), one per pad at least, the rail's and the ground's close together; the capacitor's own inductance as mounted is 1.18 nH (pad 0.10 nH + no via of the first net within 1 mm of its pad: the study's 0.6 nH per mm of 0.95 mm depth, ASSUMED). 0.5 nH is Probe's starting value, about the whole mounted inductance of a well-mounted 0402.

    • C239C239 on 'VCCBRAM': its rail pad reaches its via over 4.71 mm of track, adding about 1.83 nH to its loop, past Probe's 0.5 nH starting value; put the vias at the pads

      C239 on 'VCCBRAM': its rail pad reaches its via over 4.71 mm of track, adding about 1.83 nH to its loop, past Probe's 0.5 nH starting value; put the vias at the pads. C239.1 (rail) runs 4.71 mm along its copper to its nearest via, 3.71 mm past the 1 mm beside the pad: 1.83 nH; C239.2 (ground) not measured: the pad sits in a pour of its net on its own layer, which has no track length to measure. Each run past the pad as one conductor of a pair 1.50 mm apart (the pad pitch), 0.2 l ln(D/R) nH with R 0.127 mm (the track's width is not read): 1.83 nH in all. Put the vias at the pads (or in them), one per pad at least, the rail's and the ground's close together; the capacitor's own inductance as mounted is 1.18 nH (pad 0.10 nH + no via of the first net within 1 mm of its pad: the study's 0.6 nH per mm of 0.95 mm depth, ASSUMED). 0.5 nH is Probe's starting value, about the whole mounted inductance of a well-mounted 0402.

    • C241C241 on 'VCCBRAM': its rail pad reaches its via over 3.74 mm of track, adding about 1.35 nH to its loop, past Probe's 0.5 nH starting value; put the vias at the pads

      C241 on 'VCCBRAM': its rail pad reaches its via over 3.74 mm of track, adding about 1.35 nH to its loop, past Probe's 0.5 nH starting value; put the vias at the pads. C241.1 (rail) runs 3.74 mm along its copper to its nearest via, 2.74 mm past the 1 mm beside the pad: 1.35 nH; C241.2 (ground) not measured: the pad sits in a pour of its net on its own layer, which has no track length to measure. Each run past the pad as one conductor of a pair 1.50 mm apart (the pad pitch), 0.2 l ln(D/R) nH with R 0.127 mm (the track's width is not read): 1.35 nH in all. Put the vias at the pads (or in them), one per pad at least, the rail's and the ground's close together; the capacitor's own inductance as mounted is 1.18 nH (pad 0.10 nH + no via of the first net within 1 mm of its pad: the study's 0.6 nH per mm of 0.95 mm depth, ASSUMED). 0.5 nH is Probe's starting value, about the whole mounted inductance of a well-mounted 0402.

    • C244C244 on 'P3V3': its rail pad reaches its via over 3.29 mm of track, adding about 1.16 nH to its loop, past Probe's 0.5 nH starting value; put the vias at the pads

      C244 on 'P3V3': its rail pad reaches its via over 3.29 mm of track, adding about 1.16 nH to its loop, past Probe's 0.5 nH starting value; put the vias at the pads. C244.2 (rail) runs 3.29 mm along its copper to its nearest via, 2.29 mm past the 1 mm beside the pad: 1.16 nH; C244.1 (ground) not measured: the pad sits in a pour of its net on its own layer, which has no track length to measure. Each run past the pad as one conductor of a pair 1.60 mm apart (the pad pitch), 0.2 l ln(D/R) nH with R 0.127 mm (the track's width is not read): 1.16 nH in all. Put the vias at the pads (or in them), one per pad at least, the rail's and the ground's close together; the capacitor's own inductance as mounted is 0.74 nH (pad 0.10 nH + no via of the first net within 1 mm of its pad: the study's 0.6 nH per mm of 0.21 mm depth, ASSUMED). 0.5 nH is Probe's starting value, about the whole mounted inductance of a well-mounted 0402.

    • C2AC2A on 'FMC1_P3V3': its rail pad reaches its via over 2.52 mm of track, adding about 0.77 nH to its loop, past Probe's 0.5 nH starting value; put the vias at the pads

      C2A on 'FMC1_P3V3': its rail pad reaches its via over 2.52 mm of track, adding about 0.77 nH to its loop, past Probe's 0.5 nH starting value; put the vias at the pads. C2A.1 (rail) runs 2.52 mm along its copper to its nearest via, 1.52 mm past the 1 mm beside the pad: 0.77 nH; C2A.2 (ground) not measured: the pad sits in a pour of its net on its own layer, which has no track length to measure. Each run past the pad as one conductor of a pair 1.60 mm apart (the pad pitch), 0.2 l ln(D/R) nH with R 0.127 mm (the track's width is not read): 0.77 nH in all. Put the vias at the pads (or in them), one per pad at least, the rail's and the ground's close together; the capacitor's own inductance as mounted is 1.05 nH (pad 0.10 nH + no via of the first net within 1 mm of its pad: the study's 0.6 nH per mm of 0.74 mm depth, ASSUMED). 0.5 nH is Probe's starting value, about the whole mounted inductance of a well-mounted 0402.

    • C371C371 on 'VTT': its rail pad reaches its via over 5.50 mm of track, adding about 1.54 nH to its loop, past Probe's 0.5 nH starting value; put the vias at the pads

      C371 on 'VTT': its rail pad reaches its via over 5.50 mm of track, adding about 1.54 nH to its loop, past Probe's 0.5 nH starting value; put the vias at the pads. C371.1 (rail) runs 5.50 mm along its copper to its nearest via, 4.50 mm past the 1 mm beside the pad: 1.54 nH; C371.2 (ground) not measured: the pad sits in a pour of its net on its own layer, which has no track length to measure. Each run past the pad as one conductor of a pair 0.70 mm apart (the pad pitch), 0.2 l ln(D/R) nH with R 0.127 mm (the track's width is not read): 1.54 nH in all. Put the vias at the pads (or in them), one per pad at least, the rail's and the ground's close together; the capacitor's own inductance as mounted is 0.39 nH (pad 0.05 nH + no via of the first net within 1 mm of its pad: the study's 0.6 nH per mm of 0.21 mm depth, ASSUMED). 0.5 nH is Probe's starting value, about the whole mounted inductance of a well-mounted 0402.

    • C448C448 on 'VCCAUX_IO': its rail pad reaches its via over 2.14 mm of track, adding about 0.72 nH to its loop, past Probe's 0.5 nH starting value; put the vias at the pads

      C448 on 'VCCAUX_IO': its rail pad reaches its via over 2.14 mm of track, adding about 0.72 nH to its loop, past Probe's 0.5 nH starting value; put the vias at the pads. C448.2 (rail) runs 2.14 mm along its copper to its nearest via, 1.14 mm past the 1 mm beside the pad: 0.72 nH; C448.1 (ground) not measured: the pad sits in a pour of its net on its own layer, which has no track length to measure. Each run past the pad as one conductor of a pair 3.00 mm apart (the pad pitch), 0.2 l ln(D/R) nH with R 0.127 mm (the track's width is not read): 0.72 nH in all. Put the vias at the pads (or in them), one per pad at least, the rail's and the ground's close together; the capacitor's own inductance as mounted is 1.32 nH (pad 0.20 nH + no via of the first net within 1 mm of its pad: the study's 0.6 nH per mm of 0.53 mm depth, ASSUMED). 0.5 nH is Probe's starting value, about the whole mounted inductance of a well-mounted 0402.

    • C476C476 on 'P3V3MP': its rail pad reaches its via over 6.00 mm of track, adding about 1.71 nH to its loop, past Probe's 0.5 nH starting value; put the vias at the pads

      C476 on 'P3V3MP': its rail pad reaches its via over 6.00 mm of track, adding about 1.71 nH to its loop, past Probe's 0.5 nH starting value; put the vias at the pads. C476.1 (rail) runs at least 6.00 mm along its copper to its nearest via, 5.00 mm past the 1 mm beside the pad: 1.71 nH; C476.2 (ground) not measured: the pad sits in a pour of its net on its own layer, which has no track length to measure. Each run past the pad as one conductor of a pair 0.70 mm apart (the pad pitch), 0.2 l ln(D/R) nH with R 0.127 mm (the track's width is not read): 1.71 nH in all. Put the vias at the pads (or in them), one per pad at least, the rail's and the ground's close together; the capacitor's own inductance as mounted is 0.39 nH (pad 0.05 nH + no via of the first net within 1 mm of its pad: the study's 0.6 nH per mm of 0.21 mm depth, ASSUMED). 0.5 nH is Probe's starting value, about the whole mounted inductance of a well-mounted 0402.

    • C607C607 on 'P3V3_CLEAN': its rail pad reaches its via over 3.65 mm of track, adding about 0.90 nH to its loop, past Probe's 0.5 nH starting value; put the vias at the pads

      C607 on 'P3V3_CLEAN': its rail pad reaches its via over 3.65 mm of track, adding about 0.90 nH to its loop, past Probe's 0.5 nH starting value; put the vias at the pads. C607.1 (rail) runs 3.65 mm along its copper to its nearest via, 2.65 mm past the 1 mm beside the pad: 0.90 nH; C607.2 (ground) not measured: the pad sits in a pour of its net on its own layer, which has no track length to measure. Each run past the pad as one conductor of a pair 0.70 mm apart (the pad pitch), 0.2 l ln(D/R) nH with R 0.127 mm (the track's width is not read): 0.90 nH in all. Put the vias at the pads (or in them), one per pad at least, the rail's and the ground's close together; the capacitor's own inductance as mounted is 0.64 nH (pad 0.05 nH + no via of the first net within 1 mm of its pad: the study's 0.6 nH per mm of 0.63 mm depth, ASSUMED). 0.5 nH is Probe's starting value, about the whole mounted inductance of a well-mounted 0402.

    • C774C774 on 'P1V5': its rail pad reaches its via over 6.00 mm of track, adding about 1.71 nH to its loop, past Probe's 0.5 nH starting value; put the vias at the pads

      C774 on 'P1V5': its rail pad reaches its via over 6.00 mm of track, adding about 1.71 nH to its loop, past Probe's 0.5 nH starting value; put the vias at the pads. C774.1 (rail) runs at least 6.00 mm along its copper to its nearest via, 5.00 mm past the 1 mm beside the pad: 1.71 nH; C774.2 (ground) not measured: the pad sits in a pour of its net on its own layer, which has no track length to measure. Each run past the pad as one conductor of a pair 0.70 mm apart (the pad pitch), 0.2 l ln(D/R) nH with R 0.127 mm (the track's width is not read): 1.71 nH in all. Put the vias at the pads (or in them), one per pad at least, the rail's and the ground's close together; the capacitor's own inductance as mounted is 0.70 nH (pad 0.05 nH + no via of the first net within 1 mm of its pad: the study's 0.6 nH per mm of 0.74 mm depth, ASSUMED). 0.5 nH is Probe's starting value, about the whole mounted inductance of a well-mounted 0402.

    • C9AC9A on 'PVADJ1': its rail pad reaches its via over 2.34 mm of track, adding about 0.68 nH to its loop, past Probe's 0.5 nH starting value; put the vias at the pads

      C9A on 'PVADJ1': its rail pad reaches its via over 2.34 mm of track, adding about 0.68 nH to its loop, past Probe's 0.5 nH starting value; put the vias at the pads. C9A.1 (rail) runs 2.34 mm along its copper to its nearest via, 1.34 mm past the 1 mm beside the pad: 0.68 nH; C9A.2 (ground) not measured: the pad sits in a pour of its net on its own layer, which has no track length to measure. Each run past the pad as one conductor of a pair 1.60 mm apart (the pad pitch), 0.2 l ln(D/R) nH with R 0.127 mm (the track's width is not read): 0.68 nH in all. Put the vias at the pads (or in them), one per pad at least, the rail's and the ground's close together; the capacitor's own inductance as mounted is 1.05 nH (pad 0.10 nH + no via of the first net within 1 mm of its pad: the study's 0.6 nH per mm of 0.74 mm depth, ASSUMED). 0.5 nH is Probe's starting value, about the whole mounted inductance of a well-mounted 0402.

  6. EsdProtectorPlacementmajorOpen

    EsdProtectorPlacement: 15 components×15

    15 checked by the review by Claude walker W4 (Cern session) — 4 different reasons

    At D7A, IC4, IC5, IC7

    Each finding, with its own headline
    • D7AD7A.2 on 'USB_D_RA_P': 1 via(s) between CON2A.3 and D7A.2

      D7A (PRTR5V0U2X) protects 'USB_D_RA_P' at CON2A.3. Along the copper: 4.83 mm of track from the contact, 1 via(s) before it, on a 0.62 mm branch off the line; the line runs 10.19 mm on to R285A.1; D7A sits at CON2A, within 0.9 mm of the nearest place outside it to the contact: the track to it is mostly the escape out of CON2A's own pins, which no placement shortens.

      1 via(s) between CON2A.3 and D7A.2: the strike reaches an inner layer before it is clamped.

      Measured on the copper the ODB states: tracks, the pads they land on, and the netlist's vias. Not a straight line between centroids.

    • D7AD7A.3 on 'USB_D_RA_N': 1 via(s) between CON2A.2 and D7A.3

      D7A (PRTR5V0U2X) protects 'USB_D_RA_N' at CON2A.2. Along the copper: 4.83 mm of track from the contact, 1 via(s) before it, on a 0.62 mm branch off the line; the line runs 10.55 mm on to R284A.1; D7A sits at CON2A, within 0.9 mm of the nearest place outside it to the contact: the track to it is mostly the escape out of CON2A's own pins, which no placement shortens.

      1 via(s) between CON2A.2 and D7A.3: the strike reaches an inner layer before it is clamped.

      Measured on the copper the ODB states: tracks, the pads they land on, and the netlist's vias. Not a straight line between centroids.

    • D7AD7A.4 on 'USB_VBUSA': D7A.4 hangs 1.07 mm off the line (limit 1 mm, default

      D7A (PRTR5V0U2X) protects 'USB_VBUSA' at CON2A.1. Along the copper: 2.24 mm of track from the contact, 0 via(s) before it, on a 1.07 mm branch off the line; the line runs 9.26 mm on to C622A.2; D7A sits at CON2A, within 0.9 mm of the nearest place outside it to the contact: the track to it is mostly the escape out of CON2A's own pins, which no placement shortens.

      D7A.4 hangs 1.07 mm off the line (limit 1 mm, default: no datasheet figure on file): the line carries the strike past it while the clamp is reached through the branch.

      Measured on the copper the ODB states: tracks, the pads they land on, and the netlist's vias. Not a straight line between centroids.

    • IC4IC4.4 on 'IPMB.SCL': 2 via(s) between P1.56 and IC4.4

      IC4 (BZA462A) protects 'IPMB.SCL' at P1.56. Along the copper: 48.61 mm of track from the contact, 2 via(s) before it, on a 3.48 mm branch off the line; the line runs 306.57 mm on to IC67.16; IC4 sits 35.07 mm farther from the contact than the nearest place outside P1.

      2 via(s) between P1.56 and IC4.4: the strike reaches an inner layer before it is clamped. IC4.4 hangs 3.48 mm off the line (limit 1 mm, default: no datasheet figure on file): the line carries the strike past it while the clamp is reached through the branch.

      Measured on the copper the ODB states: tracks, the pads they land on, and the netlist's vias. Not a straight line between centroids.

    • IC4IC4.6 on 'IPMB.SDA': 2 via(s) between P1.71 and IC4.6

      IC4 (BZA462A) protects 'IPMB.SDA' at P1.71. Along the copper: 65.57 mm of track from the contact, 2 via(s) before it, on a 5.94 mm branch off the line; the line runs 321.46 mm on to IC67.15; IC4 sits 46.14 mm farther from the contact than the nearest place outside P1.

      2 via(s) between P1.71 and IC4.6: the strike reaches an inner layer before it is clamped. IC4.6 hangs 5.94 mm off the line (limit 1 mm, default: no datasheet figure on file): the line carries the strike past it while the clamp is reached through the branch.

      Measured on the copper the ODB states: tracks, the pads they land on, and the netlist's vias. Not a straight line between centroids.

    • IC5IC5.1 on 'ADD.GA2': IC5.1 hangs 3.23 mm off the line (limit 1 mm, default

      IC5 (BZA462A) protects 'ADD.GA2' at P1.26. Along the copper: 10.28 mm of track from the contact, 0 via(s) before it, on a 3.23 mm branch off the line; the line runs 256.87 mm on to IC67.68; IC5 sits 6.84 mm farther from the contact than the nearest place outside P1.

      IC5.1 hangs 3.23 mm off the line (limit 1 mm, default: no datasheet figure on file): the line carries the strike past it while the clamp is reached through the branch.

      Measured on the copper the ODB states: tracks, the pads they land on, and the netlist's vias. Not a straight line between centroids.

    • IC5IC5.3 on 'ADD.GA1': IC5.3 hangs 2.43 mm off the line (limit 1 mm, default

      IC5 (BZA462A) protects 'ADD.GA1' at P1.17. Along the copper: 7.79 mm of track from the contact, 0 via(s) before it, on a 2.43 mm branch off the line; the line runs 253.07 mm on to IC67.69; IC5 sits 5.85 mm farther from the contact than the nearest place outside P1.

      IC5.3 hangs 2.43 mm off the line (limit 1 mm, default: no datasheet figure on file): the line carries the strike past it while the clamp is reached through the branch.

      Measured on the copper the ODB states: tracks, the pads they land on, and the netlist's vias. Not a straight line between centroids.

    • IC5IC5.4 on 'ADD.GA0': IC5.4 hangs 6.76 mm off the line (limit 1 mm, default

      IC5 (BZA462A) protects 'ADD.GA0' at P1.5. Along the copper: 16.42 mm of track from the contact, 0 via(s) before it, on a 6.76 mm branch off the line; the line runs 251.12 mm on to IC67.70; IC5 sits 10.38 mm farther from the contact than the nearest place outside P1.

      IC5.4 hangs 6.76 mm off the line (limit 1 mm, default: no datasheet figure on file): the line carries the strike past it while the clamp is reached through the branch.

      Measured on the copper the ODB states: tracks, the pads they land on, and the netlist's vias. Not a straight line between centroids.

    • IC7IC7.1 on 'AJTAG.TMS': 2 via(s) between P1.166 and IC7.1

      IC7 (BZA462A) protects 'AJTAG.TMS' at P1.166. Along the copper: 258.29 mm of track from the contact, 2 via(s) before it, the line runs through its pad (flow-through); the line runs 262.38 mm on to R229.1; IC7 sits 167.49 mm farther from the contact than the nearest place outside P1.

      2 via(s) between P1.166 and IC7.1: the strike reaches an inner layer before it is clamped. IC7 joins the line 258.29 mm from P1.166 on a 262.38 mm run to R229.1: nearer the far end than the contact, so the surge travels most of the line before it is clamped.

      Measured on the copper the ODB states: tracks, the pads they land on, and the netlist's vias. Not a straight line between centroids.

    • IC7IC7.1 on 'AJTAG.TMS': 2 via(s) between J6.4 and IC7.1

      IC7 (BZA462A) protects 'AJTAG.TMS' at J6.4. Along the copper: 162.45 mm of track from the contact, 2 via(s) before it, the line runs through its pad (flow-through); the line runs 166.55 mm on to R229.1; IC7 sits 106.16 mm farther from the contact than the nearest place outside J6.

      2 via(s) between J6.4 and IC7.1: the strike reaches an inner layer before it is clamped. IC7 joins the line 162.45 mm from J6.4 on a 166.55 mm run to R229.1: nearer the far end than the contact, so the surge travels most of the line before it is clamped.

      Measured on the copper the ODB states: tracks, the pads they land on, and the netlist's vias. Not a straight line between centroids.

    • IC7IC7.3 on 'AJTAG.TDI': 2 via(s) between P1.169 and IC7.3

      IC7 (BZA462A) protects 'AJTAG.TDI' at P1.169. Along the copper: 265.66 mm of track from the contact, 2 via(s) before it, the line runs through its pad (flow-through); the line runs 267.17 mm on to R230.1; IC7 sits 168.82 mm farther from the contact than the nearest place outside P1.

      2 via(s) between P1.169 and IC7.3: the strike reaches an inner layer before it is clamped. IC7 joins the line 265.66 mm from P1.169 on a 267.17 mm run to R230.1: nearer the far end than the contact, so the surge travels most of the line before it is clamped.

      Measured on the copper the ODB states: tracks, the pads they land on, and the netlist's vias. Not a straight line between centroids.

    • IC7IC7.3 on 'AJTAG.TDI': 2 via(s) between J6.10 and IC7.3

      IC7 (BZA462A) protects 'AJTAG.TDI' at J6.10. Along the copper: 163.49 mm of track from the contact, 2 via(s) before it, the line runs through its pad (flow-through); the line runs 165 mm on to R230.1; IC7 sits 100.75 mm farther from the contact than the nearest place outside J6.

      2 via(s) between J6.10 and IC7.3: the strike reaches an inner layer before it is clamped. IC7 joins the line 163.49 mm from J6.10 on a 165 mm run to R230.1: nearer the far end than the contact, so the surge travels most of the line before it is clamped.

      Measured on the copper the ODB states: tracks, the pads they land on, and the netlist's vias. Not a straight line between centroids.

    • IC7IC7.4 on 'AJTAG.TRST#': 2 via(s) between P1.167 and IC7.4

      IC7 (BZA462A) protects 'AJTAG.TRST#' at P1.167. Along the copper: 266.12 mm of track from the contact, 2 via(s) before it, on a 0.83 mm branch off the line; the line runs 266.81 mm on to R60.2; IC7 sits 167.93 mm farther from the contact than the nearest place outside P1.

      2 via(s) between P1.167 and IC7.4: the strike reaches an inner layer before it is clamped. IC7 joins the line 265.29 mm from P1.167 on a 266.81 mm run to R60.2: nearer the far end than the contact, so the surge travels most of the line before it is clamped.

      Measured on the copper the ODB states: tracks, the pads they land on, and the netlist's vias. Not a straight line between centroids.

    • IC7IC7.6 on 'AJTAG.TCK': 2 via(s) between P1.165 and IC7.6

      IC7 (BZA462A) protects 'AJTAG.TCK' at P1.165. Along the copper: 264.44 mm of track from the contact, 2 via(s) before it, the line runs through its pad (flow-through); the line runs 268.6 mm on to R227.1; IC7 sits 167.05 mm farther from the contact than the nearest place outside P1.

      2 via(s) between P1.165 and IC7.6: the strike reaches an inner layer before it is clamped. IC7 joins the line 264.44 mm from P1.165 on a 268.6 mm run to R227.1: nearer the far end than the contact, so the surge travels most of the line before it is clamped.

      Measured on the copper the ODB states: tracks, the pads they land on, and the netlist's vias. Not a straight line between centroids.

    • IC7IC7.6 on 'AJTAG.TCK': 2 via(s) between J6.6 and IC7.6

      IC7 (BZA462A) protects 'AJTAG.TCK' at J6.6. Along the copper: 167.14 mm of track from the contact, 2 via(s) before it, the line runs through its pad (flow-through); the line runs 171.31 mm on to R227.1; IC7 sits 104.35 mm farther from the contact than the nearest place outside J6.

      2 via(s) between J6.6 and IC7.6: the strike reaches an inner layer before it is clamped. IC7 joins the line 167.14 mm from J6.6 on a 171.31 mm run to R227.1: nearer the far end than the contact, so the surge travels most of the line before it is clamped.

      Measured on the copper the ODB states: tracks, the pads they land on, and the netlist's vias. Not a straight line between centroids.

  7. DecapOrdermajorOpen

    IC14 supply pins 13, 14, 15, 16, 17 on 'P12V': the smallest capacitor C209 sits 2.0 mm away, behind C167 at 1.0 mm

    IC14 supply pins 13, 14, 15, 16, 17 on 'P12V': the smallest capacitor C209 sits 2.0 mm away, behind C167 at 1.0 mm. IC14.13: the smallest value, C209 (1 uF), is 2.0 mm from the pin and C167 (10 uF) 1.0 mm: the high-frequency part sits behind more inductance than the bulk part, which needs it least. IC14.14: the smallest value, C209 (1 uF), is 2.5 mm from the pin and C167 (10 uF) 1.1 mm: the high-frequency part sits behind more inductance than the bulk part, which needs it least. IC14.15: the smallest value, C209 (1 uF), is 3.0 mm from the pin and C167 (10 uF) 1.4 mm: the high-frequency part sits behind more inductance than the bulk part, which needs it least. IC14.16: the smallest value, C209 (1 uF), is 3.5 mm from the pin and C167 (10 uF) 1.8 mm: the high-frequency part sits behind more inductance than the bulk part, which needs it least. IC14.17: the smallest value, C209 (1 uF), is 4.0 mm from the pin and C167 (10 uF) 2.2 mm: the high-frequency part sits behind more inductance than the bulk part, which needs it least. Swap them, or bring the smallest value to the pin: it serves the fastest current and is the one the loop inductance defeats. Distances are pad centre to pin centre in the board plane. The 5 mm radius, the 10x value step and the 1 mm / 1.5x tie tolerance are Probe's starting values.

    At IC14

  8. PartLifecyclemajorOpen

    M25P128-VME6G on 2 placements (IC13, IC64): the distributor catalogue reports "Obsolete" for Micron Technology Inc.'s M25P128-VME6G

    M25P128-VME6G on 2 placements (IC13, IC64): the distributor catalogue reports "Obsolete" for Micron Technology Inc.'s M25P128-VME6G — a sourcing/longevity risk to confirm against the MANUFACTURER before acting. This board's BOM names NUMONYX, and that is one line with Micron Technology Inc.: Micron Technology completed its acquisition of Numonyx on 2010-05-07; Numonyx's NOR and PCM lines are sold under the Micron name. A distributor delisting is not the same claim as the part being unavailable; check the vendor's own lifecycle/product-change notice for the status and any named successor. Re-read the source: GET /api/odb/parts/M25P128-VME6G/providers (which provider said it, the maker and part number it matched, its own words, and when we asked). No approved alternate is recorded for this position.

    At IC13

  9. PartLifecyclemajorOpen

    PartLifecycle: 2 part numbers×2

    2 Obsolete · on 8 placements

    At IC19, IC55

    Each finding, with its own headline
    • IC19INA220AIDGST on 6 placements (IC19, IC20, IC21, IC59, IC60, IC61): the distributor catalogue reports "Obsolete" for Texas Instruments's INA220AIDGST

      INA220AIDGST on 6 placements (IC19, IC20, IC21, IC59, IC60, IC61): the distributor catalogue reports "Obsolete" for Texas Instruments's INA220AIDGST — a sourcing/longevity risk to confirm against the MANUFACTURER before acting. This board's BOM names no manufacturer for it, so whether that is the fitted maker's product was not checked. A distributor delisting is not the same claim as the part being unavailable; check the vendor's own lifecycle/product-change notice for the status and any named successor. Re-read the source: GET /api/odb/parts/INA220AIDGST/providers (which provider said it, the maker and part number it matched, its own words, and when we asked). THE LAST BUY DATE HAS PASSED (2025-12-30, 320 day(s) ago), so this is a substitution rather than a buy. The distributor reports NONE left. No approved alternate is recorded for this position.

    • IC55NCP3170ADR2G on 2 placements (IC55, IC57): the distributor catalogue reports "Obsolete" for onsemi's NCP3170ADR2G

      NCP3170ADR2G on 2 placements (IC55, IC57): the distributor catalogue reports "Obsolete" for onsemi's NCP3170ADR2G — a sourcing/longevity risk to confirm against the MANUFACTURER before acting. This board's BOM names no manufacturer for it, so whether that is the fitted maker's product was not checked. A distributor delisting is not the same claim as the part being unavailable; check the vendor's own lifecycle/product-change notice for the status and any named successor. Re-read the source: GET /api/odb/parts/NCP3170ADR2G/providers (which provider said it, the maker and part number it matched, its own words, and when we asked). THE LAST BUY DATE HAS PASSED (2022-04-01, 1,689 day(s) ago), so this is a substitution rather than a buy. The distributor reports NONE left. No approved alternate is recorded for this position.

  10. PartLifecyclemajorOpen

    PartLifecycle: 4 part numbers×4

    1 Discontinued at DigiKey, 1 Last Time Buy, 1 Obsolete, 1 Not For New Designs · on 10 placements

    At IC11, IC51, IC62A, T1

    Each finding, with its own headline
    • IC11AD8402ARZ1 on IC11: the distributor catalogue reports "Last Time Buy" for Analog Devices Inc.'s AD8402ARZ1

      AD8402ARZ1 on IC11: the distributor catalogue reports "Last Time Buy" for Analog Devices Inc.'s AD8402ARZ1 — a sourcing/longevity risk to confirm against the MANUFACTURER before acting. This board's BOM names no manufacturer for it, so whether that is the fitted maker's product was not checked. A distributor delisting is not the same claim as the part being unavailable; check the vendor's own lifecycle/product-change notice for the status and any named successor. Re-read the source: GET /api/odb/parts/AD8402ARZ1/providers (which provider said it, the maker and part number it matched, its own words, and when we asked). LAST BUY 2027-07-29 — 256 day(s) remaining to place a final order. The distributor reports 63 left. No approved alternate is recorded for this position.

    • IC51LP5951MF-3.0/NOPB on IC51: the distributor catalogue reports "Obsolete" for National Semiconductor's LP5951MF-3.0/NOPB

      LP5951MF-3.0/NOPB on IC51: the distributor catalogue reports "Obsolete" for National Semiconductor's LP5951MF-3.0/NOPB — a sourcing/longevity risk to confirm against the MANUFACTURER before acting. That is the maker this board's BOM names (NATIONAL SEMICONDUCTOR). A distributor delisting is not the same claim as the part being unavailable; check the vendor's own lifecycle/product-change notice for the status and any named successor. Re-read the source: GET /api/odb/parts/LP5951MF-3.0/NOPB/providers (which provider said it, the maker and part number it matched, its own words, and when we asked). No approved alternate is recorded for this position.

    • IC62ACP2102-GM on 2 placements (IC62A, IC62B): the distributor catalogue reports "Not For New Designs" for Silicon Labs's CP2102-GM

      CP2102-GM on 2 placements (IC62A, IC62B): the distributor catalogue reports "Not For New Designs" for Silicon Labs's CP2102-GM — a sourcing/longevity risk to confirm against the MANUFACTURER before acting. That is the maker this board's BOM names (SILICON LABS). A distributor delisting is not the same claim as the part being unavailable; check the vendor's own lifecycle/product-change notice for the status and any named successor. Re-read the source: GET /api/odb/parts/CP2102-GM/providers (which provider said it, the maker and part number it matched, its own words, and when we asked). No approved alternate is recorded for this position.

    • T1IRF9910Pbf on 6 placements (T1, T2, T3, T4, T5, T6): the distributor catalogue reports "Discontinued at DigiKey" for Infineon Technologies's IRF9910PBF

      IRF9910Pbf on 6 placements (T1, T2, T3, T4, T5, T6): the distributor catalogue reports "Discontinued at DigiKey" for Infineon Technologies's IRF9910PBF — a sourcing/longevity risk to confirm against the MANUFACTURER before acting. This board's BOM names INTERNATIONAL RECTIFIER, and that is one line with Infineon Technologies: Infineon completed its acquisitions of International Rectifier on 2015-01-13 and Cypress Semiconductor on 2020-04-16. A distributor delisting is not the same claim as the part being unavailable; check the vendor's own lifecycle/product-change notice for the status and any named successor. Re-read the source: GET /api/odb/parts/IRF9910Pbf/providers (which provider said it, the maker and part number it matched, its own words, and when we asked). No approved alternate is recorded for this position.

  11. ReferencePlaneDiscontinuitymajorOpen

    ReferencePlaneDiscontinuity: 10 boards×10

    10 checked by the review by Claude walker W4 (Cern session) — 10 different reasons

    At along@P4:FMC1_P3V3|PVADJ1, along@P4:P1V5|PVADJ2_1V8, along@P6:MGTAVTT|MGTVCCAUX, along@P6:MGTVCCAUX|P1V5, along@P6:P1V5|VCCAUX_IO, along@P6:VCCAUX|VCCAUX_IO and 4 more

    Each finding, with its own headline
    • along@P4:FMC1_P3V3|PVADJ12 nets run ALONG the split between the FMC1_P3V3 and PVADJ1 planes on P4 at 2 places, the longest 11.17 mm; the impedance rises from 44 ohm to 50 ohm there at worst, 5.7 % of each edge reflected (fastest: HB1.HB20_P at 100 ps 10-90 %, ASSUMED)

      On P4 the FMC1_P3V3 and PVADJ1 planes part, and 2 nets run down the slot between them at 2 place(s), rather than across it. This is not a crossing: the return still follows the track, in the plane on its other side and in the slot's edges, so it does not detour. What changes is the track's own impedance. With the plane gone from under it on this side for that length, its field reaches further, to the slot's edges and to the next plane beyond, so its capacitance falls and its impedance rises; every edge travelling down it is partly reflected where it enters the stretch and again where it leaves, by (Z2 - Z1)/(Z2 + Z1). The longer the stretch against the edge's rise, the more of the edge sees the full step. A crossing is generally the worse fault; a mismatch is not free. Over solid plane each track is 44 ohm to 52 ohm; along the slot 50 ohm to 58 ohm, reflecting up to 5.7 % of each edge (a method-of-moments field solve of the cross-section, one dielectric; were the slot's edges to carry none of the return, up to 58 ohm). No place stays within 10 %, the tolerance controlled impedance is commonly built to (the stackup states none), so the finding is a warning. One decision covers the split line's slot: whether these tracks may run down it, or should be moved over solid plane.

      • 'I2C_CPU.SCL' on L3 at (-58.41, 6.84), 11.76 mm, running ALONG the split for 11.17 mm, not across it, 44 ohm over solid plane, 50 ohm along it: 5.7 % of each edge reflected; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB1.HB20_P' on L3 at (-100.65, 91.02), 1.65 mm, running ALONG the split for 1.09 mm, not across it, 52 ohm over solid plane, 58 ohm along it: 5 % of each edge reflected; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=along%40P4%3AFMC1_P3V3%7CPVADJ1 serves each with its own layer_image region; all of them lie in x -102.17..-57.21, y -0.05..92.76 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • along@P4:P1V5|PVADJ2_1V8'IO13_33' runs ALONG the split between the P1V5 and PVADJ2_1V8 planes on P4 at 1 place, the longest 4.32 mm; the impedance rises from 45 ohm to 50 ohm there at worst, 5.5 % of each edge reflected (fastest: IO13_33 at 100 ps 10-90 %, ASSUMED)

      On P4 the P1V5 and PVADJ2_1V8 planes part, and 'IO13_33' runs down the slot between them at 1 place(s), rather than across it. This is not a crossing: the return still follows the track, in the plane on its other side and in the slot's edges, so it does not detour. What changes is the track's own impedance. With the plane gone from under it on this side for that length, its field reaches further, to the slot's edges and to the next plane beyond, so its capacitance falls and its impedance rises; every edge travelling down it is partly reflected where it enters the stretch and again where it leaves, by (Z2 - Z1)/(Z2 + Z1). The longer the stretch against the edge's rise, the more of the edge sees the full step. A crossing is generally the worse fault; a mismatch is not free. Over solid plane each track is 45 ohm; along the slot 50 ohm, reflecting up to 5.5 % of each edge (a method-of-moments field solve of the cross-section, one dielectric; were the slot's edges to carry none of the return, up to 50 ohm). No place stays within 10 %, the tolerance controlled impedance is commonly built to (the stackup states none), so the finding is a warning. One decision covers the split line's slot: whether these tracks may run down it, or should be moved over solid plane.

      • 'IO13_33' on L3 at (-55.68, 68.61), 6.08 mm, running ALONG the split for 4.32 mm, not across it, 45 ohm over solid plane, 50 ohm along it: 5.5 % of each edge reflected; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=along%40P4%3AP1V5%7CPVADJ2_1V8 serves each with its own layer_image region; all of them lie in x -59.6..-51.74, y 67.36..69.76 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • along@P6:MGTAVTT|MGTVCCAUX3 nets run ALONG the split between the MGTAVTT and MGTVCCAUX planes on P6 at 3 places, the longest 3.71 mm; the impedance rises from 41 ohm to 46 ohm there at worst, 5.9 % of each edge reflected (fastest: FDDR.D2 at 100 ps 10-90 %, ASSUMED)

      On P6 the MGTAVTT and MGTVCCAUX planes part, and 3 nets run down the slot between them at 3 place(s), rather than across it. This is not a crossing: the return still follows the track, in the plane on its other side and in the slot's edges, so it does not detour. What changes is the track's own impedance. With the plane gone from under it on this side for that length, its field reaches further, to the slot's edges and to the next plane beyond, so its capacitance falls and its impedance rises; every edge travelling down it is partly reflected where it enters the stretch and again where it leaves, by (Z2 - Z1)/(Z2 + Z1). The longer the stretch against the edge's rise, the more of the edge sees the full step. A crossing is generally the worse fault; a mismatch is not free. Over solid plane each track is 41 ohm to 44 ohm; along the slot 46 ohm to 49 ohm, reflecting up to 5.9 % of each edge (a method-of-moments field solve of the cross-section, one dielectric; were the slot's edges to carry none of the return, up to 50 ohm). At 2 of the 3 place(s) the rise is within 10 %, the tolerance controlled impedance is commonly built to (the stackup states none); the rest exceed it or cannot be said, so the finding is a warning. One decision covers the split line's slot: whether these tracks may run down it, or should be moved over solid plane.

      • 'FDDR.D2' on L5 at (-61.96, 75.56), 4.28 mm, running ALONG the split for 3.71 mm, not across it, 41 ohm over solid plane, 46 ohm along it: 5.9 % of each edge reflected; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'FPGA_I2C_SCL' on L5 at (-78.78, 88.81), 3.31 mm, running ALONG the split for 3.04 mm, not across it, 44 ohm over solid plane, 49 ohm along it: 4.6 % of each edge reflected; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'FPGA_I2C_SDA' on L5 at (-78.96, 88.66), 3.79 mm, running ALONG the split for 2.94 mm, not across it, 44 ohm over solid plane, 48 ohm along it: 3.8 % of each edge reflected; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=along%40P6%3AMGTAVTT%7CMGTVCCAUX serves each with its own layer_image region; all of them lie in x -81.41..-59.43, y 74.56..91.1 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • along@P6:MGTVCCAUX|P1V5'FDDR.D5' runs ALONG the split between the MGTVCCAUX and P1V5 planes on P6 at 1 place, the longest 2.77 mm; the impedance rises from 41 ohm to 46 ohm there at worst, 5.6 % of each edge reflected (fastest: FDDR.D5 at 100 ps 10-90 %, ASSUMED)

      On P6 the MGTVCCAUX and P1V5 planes part, and 'FDDR.D5' runs down the slot between them at 1 place(s), rather than across it. This is not a crossing: the return still follows the track, in the plane on its other side and in the slot's edges, so it does not detour. What changes is the track's own impedance. With the plane gone from under it on this side for that length, its field reaches further, to the slot's edges and to the next plane beyond, so its capacitance falls and its impedance rises; every edge travelling down it is partly reflected where it enters the stretch and again where it leaves, by (Z2 - Z1)/(Z2 + Z1). The longer the stretch against the edge's rise, the more of the edge sees the full step. A crossing is generally the worse fault; a mismatch is not free. Over solid plane each track is 41 ohm; along the slot 46 ohm, reflecting up to 5.6 % of each edge (a method-of-moments field solve of the cross-section, one dielectric; were the slot's edges to carry none of the return, up to 47 ohm). No place stays within 10 %, the tolerance controlled impedance is commonly built to (the stackup states none), so the finding is a warning. One decision covers the split line's slot: whether these tracks may run down it, or should be moved over solid plane.

      • 'FDDR.D5' on L5 at (-57.63, 78.5), 4.6 mm, running ALONG the split for 2.77 mm, not across it, 41 ohm over solid plane, 46 ohm along it: 5.6 % of each edge reflected; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=along%40P6%3AMGTVCCAUX%7CP1V5 serves each with its own layer_image region; all of them lie in x -58.89..-56.63, y 75.33..81.82 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • along@P6:P1V5|VCCAUX_IO3 nets run ALONG the split between the P1V5 and VCCAUX_IO planes on P6 at 3 places, the longest 1.42 mm; the impedance rises from 41 ohm to 47 ohm there at worst, 6.6 % of each edge reflected (fastest: FDDR.A8 at 100 ps 10-90 %, ASSUMED)

      On P6 the P1V5 and VCCAUX_IO planes part, and 3 nets run down the slot between them at 3 place(s), rather than across it. This is not a crossing: the return still follows the track, in the plane on its other side and in the slot's edges, so it does not detour. What changes is the track's own impedance. With the plane gone from under it on this side for that length, its field reaches further, to the slot's edges and to the next plane beyond, so its capacitance falls and its impedance rises; every edge travelling down it is partly reflected where it enters the stretch and again where it leaves, by (Z2 - Z1)/(Z2 + Z1). The longer the stretch against the edge's rise, the more of the edge sees the full step. A crossing is generally the worse fault; a mismatch is not free. Over solid plane each track is 41 ohm; along the slot 46 ohm, reflecting up to 6.6 % of each edge (a method-of-moments field solve of the cross-section, one dielectric; were the slot's edges to carry none of the return, up to 47 ohm). No place stays within 10 %, the tolerance controlled impedance is commonly built to (the stackup states none), so the finding is a warning. One decision covers the split line's slot: whether these tracks may run down it, or should be moved over solid plane.

      • 'FDDR.BA2' on L5 at (-57.56, 73.54), 2.07 mm, running ALONG the split for 1.31 mm, not across it, 41 ohm over solid plane, 47 ohm along it: 6.6 % of each edge reflected; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'FDDR.CKE0' on L5 at (-57.8, 73.37), 2.48 mm, running ALONG the split for 1.42 mm, not across it, 41 ohm over solid plane, 47 ohm along it: 6.6 % of each edge reflected; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'FDDR.A8' on L5 at (-42.15, 68.44), 1.63 mm, running ALONG the split for 1.05 mm, not across it, 41 ohm over solid plane, 46 ohm along it: 5.7 % of each edge reflected; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=along%40P6%3AP1V5%7CVCCAUX_IO serves each with its own layer_image region; all of them lie in x -58.8..-40.29, y 67.44..75.67 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • along@P6:VCCAUX|VCCAUX_IO3 nets run ALONG the split between the VCCAUX and VCCAUX_IO planes on P6 at 3 places, the longest 6.46 mm; the impedance rises from 52 ohm to 58 ohm there at worst, 4.9 % of each edge reflected (fastest: FDDR.A4 at 100 ps 10-90 %, ASSUMED)

      On P6 the VCCAUX and VCCAUX_IO planes part, and 3 nets run down the slot between them at 3 place(s), rather than across it. This is not a crossing: the return still follows the track, in the plane on its other side and in the slot's edges, so it does not detour. What changes is the track's own impedance. With the plane gone from under it on this side for that length, its field reaches further, to the slot's edges and to the next plane beyond, so its capacitance falls and its impedance rises; every edge travelling down it is partly reflected where it enters the stretch and again where it leaves, by (Z2 - Z1)/(Z2 + Z1). The longer the stretch against the edge's rise, the more of the edge sees the full step. A crossing is generally the worse fault; a mismatch is not free. Over solid plane each track is 41 ohm to 52 ohm; along the slot 44 ohm to 58 ohm, reflecting up to 4.9 % of each edge (a method-of-moments field solve of the cross-section, one dielectric; were the slot's edges to carry none of the return, up to 58 ohm). At 1 of the 3 place(s) the rise is within 10 %, the tolerance controlled impedance is commonly built to (the stackup states none); the rest exceed it or cannot be said, so the finding is a warning. One decision covers the split line's slot: whether these tracks may run down it, or should be moved over solid plane.

      • 'LA2.LA15_N' on L5 at (-67.89, 37.92), 1.84 mm, running ALONG the split for 1.34 mm, not across it, 52 ohm over solid plane, 58 ohm along it: 4.9 % of each edge reflected; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'VADJ1_IO5' on L5 at (-62.18, 69.46), 7.33 mm, running ALONG the split for 6.46 mm, not across it, 44 ohm over solid plane, 49 ohm along it: 4.9 % of each edge reflected; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'FDDR.A4' on L5 at (-53.52, 69.69), 4.75 mm, running ALONG the split for 3.21 mm, not across it, 41 ohm over solid plane, 44 ohm along it: 3.8 % of each edge reflected; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=along%40P6%3AVCCAUX%7CVCCAUX_IO serves each with its own layer_image region; all of them lie in x -69.54..-50.17, y 36.27..70.75 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -52.57,64.65|-51.77,65.45'HB2.HB01_P' crosses one gap in P6's negative-plane field at 1 place, the longest 1.48 mm (fastest: HB2.HB01_P at 100 ps 10-90 %, ASSUMED)

      Along 'HB2.HB01_P' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 1.48 mm under the track at (-52.17, 64.67). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 1 of them run ALONG the bare copper, parallel to its edges, rather than across it (the longest 0.83 mm). What cleared it: the hole at x -52.57..-51.77, y 64.65..65.45. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. No crossing has a plane on the track's other side, so the whole return meets the break.

      • 'HB2.HB01_P' on L5 at (-52.17, 64.67), 1.48 mm, running ALONG the bare copper for 0.83 mm, not across it; other side: no plane on G4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-52.57%2C64.65%7C-51.77%2C65.45 serves each with its own layer_image region; all of them lie in x -53.56..-50.19, y 63.36..65.76 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -81.57,55.65|-80.77,56.45'HA1.HA11_N' crosses one gap in P6's negative-plane field at 1 place, the longest 0.54 mm (fastest: HA1.HA11_N at 100 ps 10-90 %, ASSUMED)

      Along 'HA1.HA11_N' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.54 mm under the track at (-81.17, 55.66). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. What cleared it: the hole at x -81.57..-80.77, y 55.65..56.45. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. No crossing has a plane on the track's other side, so the whole return meets the break.

      • 'HA1.HA11_N' on L5 at (-81.17, 55.66), 0.54 mm; other side: no plane on G4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-81.57%2C55.65%7C-80.77%2C56.45 serves each with its own layer_image region; all of them lie in x -82.44..-79.9, y 54.66..56.66 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • merged:-102.11,43.02|-102.11,45.123 nets cross the slot of 4 drills from (-102.11, 43.02) to (-102.11, 45.12), whose anti-pads run together in G3's negative-plane field and P4's negative-plane field, between 3 pairs of drills (fastest: FLASH_D2 at 100 ps 10-90 %, ASSUMED)

      G3's negative-plane field and P4's negative-plane field are cleared round a field of 4 drills from (-102.11, 43.02) to (-102.11, 45.12), and the clearances run together: the straight line from one drill to the next is bare all the way. 3 nets cross between drills of it at 3 place(s), 6 crossings counting each plane beside the track, each listed below with its two drills. Past one drill's anti-pad the return flows round on either side; here it cannot pass between the drills and detours round the whole merged clearance - the slot a row of pins or vias cuts in a plane. One decision covers the slot: whether these tracks may cross it, or the clearances should leave copper between the drills. No crossing has a plane on the track's other side, so the whole return meets the break.

      • 'FLASH_D2' on L3 at (-102.11, 44.07), 0.14 mm between the drills at (-102.11, 43.72) and (-102.11, 44.42) in G3; other side: no plane on P4
      • 'FLASH_D2' on L3 at (-102.11, 44.07), 0.14 mm between the drills at (-102.11, 43.72) and (-102.11, 44.42) in P4; other side: no plane on G3
      • 'FSPI.DONE' on L3 at (-102.11, 44.77), 0.14 mm between the drills at (-102.11, 44.42) and (-102.11, 45.12) in G3; other side: no plane on P4
      • 'FSPI.DONE' on L3 at (-102.11, 44.77), 0.14 mm between the drills at (-102.11, 44.42) and (-102.11, 45.12) in P4; other side: no plane on G3
      • 'FSPI.FLASH_SI/D0' on L3 at (-102.11, 43.37), 0.14 mm between the drills at (-102.11, 43.02) and (-102.11, 43.72) in G3; other side: no plane on P4
      • 'FSPI.FLASH_SI/D0' on L3 at (-102.11, 43.37), 0.14 mm between the drills at (-102.11, 43.02) and (-102.11, 43.72) in P4; other side: no plane on G3

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=merged%3A-102.11%2C43.02%7C-102.11%2C45.12 serves each with its own layer_image region; all of them lie in x -103.18..-101.04, y 42.37..45.77 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • split@P4:FMC1_P3V3|PVADJ183 nets cross the split between the FMC1_P3V3 and PVADJ1 planes on P4 at 97 places; 91 with a plane on the track's other side (fastest: FLASH_D2 at 100 ps 10-90 %, ASSUMED)

      On P4 the FMC1_P3V3 and PVADJ1 planes part, and 83 nets cross from one to the other at 97 place(s). The return current cannot follow a track across; it detours to wherever the two planes are joined (a stitching capacitor, or the source), and the loop it opens radiates and couples. One decision covers the boundary: whether it is stitched near these crossings, or the nets crossing it are slow enough not to care. Tracks that run ALONG this split, down the slot rather than across it, are a finding of their own: theirs is an impedance step, not a detour. At 91 of the 97 crossing(s) the track's other neighbour carries a plane, which the return can use there. 91 of the 97 crossing(s) have a plane on the track's other side at most twice as far away as the broken one (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The rest leave most or all of the return to the break, so the finding is a warning.

      • 'CLKS1.CLK0_M2C_N' on L3 at (-86.68, 51.83), 1.18 mm; other side: no plane on G3
      • 'CLKS1.CLK0_M2C_P' on L3 at (-86.68, 51.34), 1.16 mm; other side: no plane on G3
      • 'CLKS1.CLK1_M2C_N' on L3 at (-86.69, 49.79), 1.18 mm; other side: no plane on G3
      • 'CLKS1.CLK1_M2C_P' on L3 at (-86.69, 49.28), 1.19 mm; other side: no plane on G3
      • 'FJTAG.TCK' on L3 at (-102.24, 95.63), 0.62 mm; other side: no plane on G3
      • 'FJTAG.TDI' on L3 at (-102.24, 96.2), 0.4 mm; other side: no plane on G3
      • 'CLKS1.CLK2_BIDIR_N' on L3 at (-86.29, 53.42), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'CLKS1.CLK2_BIDIR_P' on L3 at (-86.29, 53.2), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'CLKS1.CLK3_BIDIR_N' on L3 at (-85.74, 48.85), 0.41 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'CLKS1.CLK3_BIDIR_P' on L3 at (-85.5, 48.75), 0.41 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FJTAG.TCK' on L3 at (-97.14, 90.74), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FJTAG.TCK' on L3 at (-86.29, 76.03), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • and 85 more

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=split%40P4%3AFMC1_P3V3%7CPVADJ1 serves each with its own layer_image region; all of them lie in x -109.12..-57.21, y 13.66..100.23 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

  12. ReferencePlaneDiscontinuitymajorOpen

    ReferencePlaneDiscontinuity: 8 boards×8

    e.g. “'CLK_EXT_P' crosses one gap in G5's negative-plane field at 1 place, the longest 0.64 mm (fastest: CLK_EXT_P at 1 ns 10-90 %, suggested by its name)” · 7 different statements

    At gap:G1:hole -14.37,85.63|-13.73,85.9, gap:G1:hole -9.87,85.63|-9.23,85.9, gap:G3:hole -17.77,64.53|-17.5,65.17, gap:G3:hole -17.77,69.53|-17.5,70.17, gap:G3:hole -8.77,62.03|-8.5,62.67, gap:G3:hole -8.77,94.53|-8.5,95.17, gap:G5:hole -14.37,56.3|-13.73,56.57, gap:G5:hole -18.87,56.3|-18.23,56.57

    Each finding, with its own headline
    • gap:G1:hole -14.37,85.63|-13.73,85.9'FPGA_CLK1_N' crosses one gap in G1's negative-plane field at 1 place, the longest 0.28 mm (fastest: FPGA_CLK1_N at 1 ns 10-90 %, suggested by its name)

      Along 'FPGA_CLK1_N' on TOP, G1 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.28 mm under the track at (-14.05, 85.89). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 1 of them run ALONG the bare copper, parallel to its edges, rather than across it (the longest 0.28 mm). What cleared it: the hole at x -14.37..-13.73, y 85.63..85.9. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. No crossing has a plane on the track's other side, so the whole return meets the break.

      • 'FPGA_CLK1_N' on TOP at (-14.05, 85.89), 0.28 mm, running ALONG the bare copper for 0.28 mm, not across it; other side: none: the signal layer is outermost

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AG1%3Ahole%20-14.37%2C85.63%7C-13.73%2C85.9 serves each with its own layer_image region; all of them lie in x -15.19..-12.91, y 84.89..86.89 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:G1:hole -9.87,85.63|-9.23,85.9'FPGA_CLK1_N' crosses one gap in G1's negative-plane field at 1 place, the longest 0.37 mm (fastest: FPGA_CLK1_N at 1 ns 10-90 %, suggested by its name)

      Along 'FPGA_CLK1_N' on TOP, G1 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.37 mm under the track at (-9.59, 85.89). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 1 of them run ALONG the bare copper, parallel to its edges, rather than across it (the longest 0.37 mm). What cleared it: the hole at x -9.87..-9.23, y 85.63..85.9. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. No crossing has a plane on the track's other side, so the whole return meets the break.

      • 'FPGA_CLK1_N' on TOP at (-9.59, 85.89), 0.37 mm, running ALONG the bare copper for 0.37 mm, not across it; other side: none: the signal layer is outermost

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AG1%3Ahole%20-9.87%2C85.63%7C-9.23%2C85.9 serves each with its own layer_image region; all of them lie in x -10.69..-8.32, y 84.88..86.89 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:G3:hole -17.77,64.53|-17.5,65.17'RTM_LVDS.RTM_LVD1_N' crosses one gap in G3's negative-plane field at 1 place, the longest 0.32 mm (fastest: RTM_LVDS.RTM_LVD1_N at 100 ps 10-90 %, ASSUMED)

      Along 'RTM_LVDS.RTM_LVD1_N' on L3, G3 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.32 mm under the track at (-17.56, 64.9). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. What cleared it: the hole at x -17.77..-17.5, y 64.53..65.17. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. No crossing has a plane on the track's other side, so the whole return meets the break.

      • 'RTM_LVDS.RTM_LVD1_N' on L3 at (-17.56, 64.9), 0.32 mm; other side: no plane on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AG3%3Ahole%20-17.77%2C64.53%7C-17.5%2C65.17 serves each with its own layer_image region; all of them lie in x -18.6..-16.56, y 63.85..66.16 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:G3:hole -17.77,69.53|-17.5,70.17'RTM_LVDS.RTM_LVD5_N' crosses one gap in G3's negative-plane field at 1 place, the longest 0.31 mm (fastest: RTM_LVDS.RTM_LVD5_N at 100 ps 10-90 %, ASSUMED)

      Along 'RTM_LVDS.RTM_LVD5_N' on L3, G3 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.31 mm under the track at (-17.57, 70). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. What cleared it: the hole at x -17.77..-17.5, y 69.53..70.17. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. No crossing has a plane on the track's other side, so the whole return meets the break.

      • 'RTM_LVDS.RTM_LVD5_N' on L3 at (-17.57, 70), 0.31 mm; other side: no plane on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AG3%3Ahole%20-17.77%2C69.53%7C-17.5%2C70.17 serves each with its own layer_image region; all of them lie in x -18.6..-16.56, y 68.86..71.16 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:G3:hole -8.77,62.03|-8.5,62.67'RTM_LVDS.RTM_LVD0_N' crosses one gap in G3's negative-plane field at 1 place, the longest 0.29 mm (fastest: RTM_LVDS.RTM_LVD0_N at 100 ps 10-90 %, ASSUMED)

      Along 'RTM_LVDS.RTM_LVD0_N' on L3, G3 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.29 mm under the track at (-8.56, 62.5). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. What cleared it: the hole at x -8.77..-8.5, y 62.03..62.67. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. No crossing has a plane on the track's other side, so the whole return meets the break.

      • 'RTM_LVDS.RTM_LVD0_N' on L3 at (-8.56, 62.5), 0.29 mm; other side: no plane on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AG3%3Ahole%20-8.77%2C62.03%7C-8.5%2C62.67 serves each with its own layer_image region; all of them lie in x -9.59..-7.56, y 61.37..63.66 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:G3:hole -8.77,94.53|-8.5,95.17'FPGA_CLK2_N' crosses one gap in G3's negative-plane field at 1 place, the longest 0.32 mm (fastest: FPGA_CLK2_N at 1 ns 10-90 %, suggested by its name)

      Along 'FPGA_CLK2_N' on L3, G3 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.32 mm under the track at (-8.56, 94.8). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. What cleared it: the hole at x -8.77..-8.5, y 94.53..95.17. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. No crossing has a plane on the track's other side, so the whole return meets the break.

      • 'FPGA_CLK2_N' on L3 at (-8.56, 94.8), 0.32 mm; other side: no plane on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AG3%3Ahole%20-8.77%2C94.53%7C-8.5%2C95.17 serves each with its own layer_image region; all of them lie in x -9.6..-7.56, y 93.54..95.85 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:G5:hole -14.37,56.3|-13.73,56.57'CLK_EXT_P' crosses one gap in G5's negative-plane field at 1 place, the longest 0.64 mm (fastest: CLK_EXT_P at 1 ns 10-90 %, suggested by its name)

      Along 'CLK_EXT_P' on BTM, G5 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.64 mm under the track at (-14.05, 56.41). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. What cleared it: the hole at x -14.37..-13.73, y 56.3..56.57. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. No crossing has a plane on the track's other side, so the whole return meets the break.

      • 'CLK_EXT_P' on BTM at (-14.05, 56.41), 0.64 mm; other side: none: the signal layer is outermost

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AG5%3Ahole%20-14.37%2C56.3%7C-13.73%2C56.57 serves each with its own layer_image region; all of them lie in x -15.37..-12.73, y 55.41..57.41 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:G5:hole -18.87,56.3|-18.23,56.57'CLK_EXT_P' crosses one gap in G5's negative-plane field at 1 place, the longest 0.64 mm (fastest: CLK_EXT_P at 1 ns 10-90 %, suggested by its name)

      Along 'CLK_EXT_P' on BTM, G5 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.64 mm under the track at (-18.55, 56.41). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. What cleared it: the hole at x -18.87..-18.23, y 56.3..56.57. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. No crossing has a plane on the track's other side, so the whole return meets the break.

      • 'CLK_EXT_P' on BTM at (-18.55, 56.41), 0.64 mm; other side: none: the signal layer is outermost

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AG5%3Ahole%20-18.87%2C56.3%7C-18.23%2C56.57 serves each with its own layer_image region; all of them lie in x -19.87..-17.23, y 55.41..57.41 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

  13. SinglePointNetmajorOpen

    SinglePointNet: 2 pins×2

    Checked by the review, stands · Claude walker (cern session) — Unused inputs left open (netlist): XADC VP_0/VN_0 on the XC7K325T IC79 (R15/T14), LPC1764 RTCX1 (IC1.16, RTC crystal unused), CP2102 DCD (IC62A/B pin 1). None is a driver; the datasheets on file do not state the termination for the unused case (CP2102 shows 25-50 uA input current, p.8), so left checked.

    At IC79.R15, IC79.T14

    Each finding, with its own headline
    • IC79.R15IC79.R15 (VP_0) is an IC input left unconnected

      Pin IC79.R15 (VP_0) is an IC INPUT left UNCONNECTED (on the no-connect bucket). Marked no-connect by the designer, but an input normally needs a defined level - confirm it is truly meant to float. If the datasheet defines this input left open (an internal pull, "read as low when left floating", or a tri-level input whose open state is one of its settings), file it on the pin as if_unused (pull_up, pull_down or may_float) with submit_pin_behavior, citing the page, and this finding clears.

    • IC79.T14IC79.T14 (VN_0) is an IC input left unconnected

      Pin IC79.T14 (VN_0) is an IC INPUT left UNCONNECTED (on the no-connect bucket). Marked no-connect by the designer, but an input normally needs a defined level - confirm it is truly meant to float. If the datasheet defines this input left open (an internal pull, "read as low when left floating", or a tri-level input whose open state is one of its settings), file it on the pin as if_unused (pull_up, pull_down or may_float) with submit_pin_behavior, citing the page, and this finding clears.

  14. observationOBS-427minorOpen

    Standalone 3.3 V management LDO IC74 (LD1117) has 1 uF at its output; the datasheet asks for 10 uF minimum for stability

    The stand-alone 3.3 V management LDO IC74 (LD1117) has only C817 = 1 µF at its output, while ST asks for 10 µF minimum for stability. This is the MMC's own supply in stand-alone (J4) operation.

    The numbers

    QuantityValueSource
    IC74 inputP12V_AUX, from the J4 bench-supply connectornetlist
    Capacitance on P3V3MP_AUXC817 = 1 µF 0603 X7R, the only onenetlist: P3V3MP_AUX
    Decoupling on P3V3MP, behind the PTC F3100 nF and 1 µF parts, about 4.2 µF in totalnetlist: P3V3MP
    Total counting everything behind the PTCabout 5 µFderived
    LD1117 minimum output capacitor10 µFLD1117 datasheet, p.1

    Why it matters

    An LD1117 below its minimum output capacitance can oscillate, and here that is the MMC's own supply in stand-alone operation.

    Suggested fix: a 10 µF (or larger, within the datasheet ESR window) capacitor at IC74's output.

  15. observationOBS-428minorOpen

    P1V8 LDO IC81 (TPS74401) has 1 uF at its output; current TPS74401 silicon requires at least 2.2 uF

    The P1V8 LDO IC81 (TPS74401) has only C296 = 1 µF at its output, while current TPS74401 silicon (the new die) requires at least 2.2 µF.

    The numbers

    QuantityValueSource
    P1V8 set pointR346 3k57 / R347 2k87 from 0.8 V: 1.795 Vderived
    Capacitance on the P1V8 netC296 = 1 µF 0603 X7R, the only onenetlist: P1V8
    Output capacitor, legacy dieany or none (0 µF)TPS74401 datasheet (SBVS066T), p.4
    Output capacitor, new die2.2 µF minimumTPS74401 datasheet (SBVS066T), p.1 and p.4

    The rest of the load sits behind a bead and an inductor: MGTVCCAUX via ferrite FB1, VCCAUX via L12.

    Why it matters

    Boards built with the new TPS74401 die (anything bought now) would have the VCCAUX/MGTVCCAUX LDO below its stated minimum output capacitance, risking oscillation on an FPGA auxiliary supply. On legacy-die builds it is within spec.

    Suggested fix: add a 2.2 µF+ capacitor directly at IC81's OUT pins. (IC2 on PVADJ2_1V8 already has 100 µF.)

  16. observationOBS-440minorOpen

    AMC M-LVDS transmitters (IC22/IC23) switch on and drive the backplane lines while the FPGA is unconfigured: PUDC_B is tied low and the DE/RE/DIR controls have no pull-down

    The AMC M-LVDS transmitters IC22/IC23 switch on and drive the backplane lines while the FPGA is unconfigured: PUDC_B is tied low, which pulls every FPGA pin high, and the DE/RE/DIR controls have no pull-down.

    Where it is

    ``mermaid flowchart LR P["PUDC_B, ball R23<br/>R354 0R to GND"] -- "internal pull-ups on" --> F["IC79 FPGA ball<br/>e.g. Y11"] F -- "A input" --> X["IC40 SN74AVC8T245<br/>DIR = VCCA, OE = GND"] F -- "DIR" --> Y["SN74AVC1T45<br/>IC28 etc."] X -- "DE1..DE8, no pull-down" --> M["IC22/IC23 SN65MLVD040<br/>DE and /RE tied together"] M -- "AMC ports 17-20" --> B["Backplane M-LVDS"] ``

    • IC22/IC23 (SN65MLVD040): each channel has DE and /RE tied together (e.g. DE1 = pins 12 and 21) and D and R tied together (DR1 = pins 25 and 26).
    • DE1..DE8 come from IC40 (SN74AVC8T245, DIR on P1V5 = VCCA so A->B, OE on GND). Its A inputs are FPGA balls, e.g. Y11 (MLVDS_F.R\E\_DE_RX_P17).
    • The same FPGA ball drives the DIR pin of the matching SN74AVC1T45 (IC28 etc.).
    • No pull-down is fitted on any of these nets (DE1 = IC22.12 + IC22.21 + IC40.14 only).
    • PUDC (IC79.R23) has R354 0R to GND and R180 1k to PVADJ2, both on the BOM as fitted and shown on schematic sheet 20 without a DNP mark.

    What happens before the bitstream loads

    Per UG475 (p.22), PUDC_B low enables the internal pull-ups on every SelectIO pin after power-up and during configuration. So until the bitstream loads, every DE/RE and DIR is pulled high:

    • all eight drivers enabled (DE is active-high, per the SN65MLVD040 pin table, p.3)
    • receivers disabled (/RE is active-low)
    • each 1T45 drives D from an FPGA pin that is itself pulled high

    Why it matters

    On an MTCA.4 carrier, AMC ports 17-20 are bussed M-LVDS lines shared with other modules. This board would assert them high every time its FPGA is powered, reset or reconfigured, which corrupts clocks/triggers owned by another module during that time. M-LVDS tolerates the contention electrically.

    Suggested fix: a pull-down on the DE nets (or PUDC_B high, with pull-downs where needed) would keep the drivers off until the FPGA takes control. Designer's judgement needed on how these ports are used.

  17. observationOBS-441minorOpen

    TXB0104 translators IC68/IC69/IC70 can see VCCA above VCCB: their A side is the adjustable FMC VADJ rail (up to about 3.3 V) and their B side is P3V3

    The TXB0104 translators IC68, IC69 and IC70 can see VCCA above VCCB: their A side is the adjustable FMC VADJ rail (up to about 3.3 V) and their B side is P3V3, while TI requires VCCA <= VCCB.

    Where it is

    TranslatorVCCA (pin 1) and OE (pin 8)VCCB (pin 14)
    IC68PVADJ1P3V3
    IC69PVADJ1P3V3
    IC70PVADJ2P3V3

    The numbers

    QuantityValueSource
    TXB0104 A-port supply1.2 V <= VCCA <= 3.6 V and VCCA <= VCCBTXB0104 datasheet, pin table, p.4
    TPS53126 VFB, min/typ/max at 25 C0.755 / 0.765 / 0.775 VTPS53126 datasheet, p.5
    PVADJ1/PVADJ2 rangeIC10 with R247/R248 1k5 over 453R + (910R ‖ AD8402 RDAC) from 0.765 V: 1.61 V up to about 3.30 V, 3.34 V at the reference maximumnetlist: FB1 (R247, R263, R261, IC11 pin 4); derived
    P3V3IC43 with 4k3/1k3: 3.30 V nominal, about 3.24-3.34 Vderived

    At the top VADJ setting the two rails are equal nominally, and tolerance can put VCCA above VCCB. VADJ can also ramp while P3V3 is still off, since the MMC enables them independently.

    Why it matters

    The TXB0104's one-shot edge accelerators assume the A side is the lower rail. Operating with VCCA > VCCB is outside the stated conditions: behaviour and leakage are not specified. It only matters when an FMC mezzanine asks for VADJ = 3.3 V, or during sequencing.

    Options: cap VADJ below 3.3 V in firmware while these translators are fitted, or use a translator without the VCCA <= VCCB rule. Needs a designer's view on the intended VADJ range.

  18. observationOBS-444minorOpen

    FMC VADJ current monitors are cross-labelled: the 'FMC1' VADJ monitor (IC19) measures FMC2's shunt and the 'FMC2' monitor (IC60) measures FMC1's

    The FMC VADJ current monitors are cross-labelled: the 'FMC1' VADJ monitor IC19 measures FMC2's shunt R257, and the 'FMC2' monitor IC60 measures FMC1's shunt R75.

    The mapping

    ShuntSits betweenPort label on sheet 26NetGoes to
    R752.5V_FMC1 and PVADJ1 (FMC1 VADJ)I_FMC2_PVADJ+/-I_FMC2_PADJP/NIC60 (pins 8, 10)
    R2572.5V_FMC2 and PVADJ2 (FMC2 VADJ)I_FMC1_PVADJ+/-I_FMC1_PADJP/NIC19 (pins 8, 10)

    Not swapped

    The 3.3 V and 12 V monitors are correctly assigned:

    • I_FMC1_P3V3 on R94 in the FMC1_P3V3 path (3.3V_FMC1 -> FMC1_P3V3), to IC21
    • I_FMC1_P12V0 on R266 in the FMC1 12 V path, to IC61
    • likewise for FMC2

    Why it matters

    MMC firmware that follows the net names will report FMC1's VADJ current as FMC2's and vice versa, while the 3.3 V and 12 V readings are correctly assigned. That is an easy source of wrong power-budget or overcurrent decisions per mezzanine.

    Suggested fix: swap the port labels on sheet 26, or document the mapping for firmware (IC19 = FMC2 VADJ, IC60 = FMC1 VADJ).

  19. observationOBS-446minorOpen

    RGB status LED LD1: green and blue emitters (Vf 3.0-3.6 V) run from 3.3 V through the same 330R as red, so they are barely lit or dark

    The RGB status LED LD1 runs all three colours from P3V3 through the same 330R, but green and blue have a forward voltage of 3.0-3.6 V, so they are barely lit or dark while red gets about 4 mA.

    Where it is

    LD1 (OVSRRGBCC3) has its common anode (pin 4) on P3V3. Each cathode goes through 330R to a BSS138 low-side switch whose gate comes from the FPGA:

    ColourLD1 pinResistorSwitchFPGA ball
    Red1 (LED_LR)R277 330RQ7G23
    Green2 (LED_GR)R278 330RQ10G25
    Blue3 (LED_BR)R279 330RQ11F23, R282 10k to GND on the gate

    The numbers

    QuantityValueSource
    Red Vf at 20 mA, min/typ/max1.8 / 2.0 / 2.4 VOVSRRGBCC3 datasheet, p.3
    Green and blue Vf at 20 mA, min/typ/max3.0 / 3.3 / 3.6 VOVSRRGBCC3 datasheet, p.3
    Red current(3.3 - 2.0) / 330 = about 4 mAderived
    Green/blue headroom across 330R plus the FET0-0.3 V: well under 1 mA, and none at the maximum Vfderived

    Why it matters

    Identical resistors across colours with materially different Vf leave two of the three colours near-dark and the mix unpredictable part to part. Nothing on file records a weighting intent.

    Suggested fix: a higher anode supply for green/blue (5 V is not available here), much smaller resistors, or a lower-Vf part.

  20. observationOBS-452minorOpen

    Config-flash clock FLASH_CLK_R carries a 100R/100R Thevenin termination: a 50-ohm load to 1.65 V that CMOS SPI drivers (FPGA CCLK, MMC SCK1) cannot swing rail to rail

    The configuration-flash clock FLASH_CLK_R has a fitted 100R/100R Thevenin termination (R138 to P3V3, R134 to GND): a 50 Ω load to 1.65 V that the CMOS SPI drivers (FPGA CCLK, MMC SCK1) cannot swing rail to rail.

    Where it is

    ``mermaid flowchart LR F1["FLASH_CLK<br/>FPGA ball A27"] -- "R350 0R" --> N["FLASH_CLK_R"] F2["FSPI.FPGA_CCLK<br/>FPGA CCLK, ball B10"] -- "R351 10R" --> N M["FSPI.SCK1<br/>MMC-side SPI clock"] -- "R352 10R" --> N N --> IC["IC13 M25P128 pin 6 C<br/>and unfitted IC17 footprint"] N --> T["R138 100R to P3V3<br/>R134 100R to GND<br/>both fitted"] ``

    The numbers

    QuantityValueSource
    Termination seen by the drivers50 Ω to 1.65 Vnetlist: FLASH_CLK_R, R134/R138 100R; derived
    Standing current from P3V316.5 mA, continuousderived
    LPC1764 SPI pin drive4 mA: VOH = VDD - 0.4 V at IOH = -4 mALPC1764 datasheet, p.48
    Clock swing from the MMC sideabout ±0.2 V around 1.65 Vderived
    M25P128 input thresholdsVIH 0.7 x VCC = 2.31 V; VIL 0.3 x VCCM25P128 datasheet, p.36
    FPGA CCLK peak current into the network33 mAderived
    Standing dissipation54 mWderived

    From the MMC side the clock never reaches the flash's VIH or VIL.

    Control

    The second flash IC64's clock, FLASH_CLK_RA, has only the R293 0R / R294 10R / R295 10R feeds and no termination. That suggests R134/R138 are an unintended fit or a leftover option.

    Why it matters

    If the MMC is meant to (re)program the configuration flash through SCK1, it cannot clock IC13 with this termination fitted. The FPGA's CCLK driver also has to fight 33 mA peak into the network, costing swing margin at the flash, plus 54 mW of standing dissipation.

    Suggested fix: remove R134/R138, or replace them with a series-source / AC termination.

  21. observationOBS-457minorOpen

    I2C pull-ups stack to about 660 ohm (4.4 mA) when the PCA9547 selects the I2C_CPU channel with the payload clock switch closed

    When the PCA9547 (IC80) selects channel 4 (I2C_CPU) while the PCA9306 switch IC56 is closed, three sets of pull-ups stack to about 660 Ω, so a slave must sink 4.4 mA to reach 0.4 V; MAX6642 (IC18) is specified for 1 mA there.

    The pull-ups that end up in parallel

    SegmentWherePull-upsRail
    FPGA_SCL/FPGA_SDAIC80 upstream SCL/SDA, pins 19/20R361/R360 2k2P3V3MP
    I2C_CPUIC80 channel 4, SD4/SC4 pins 10/11R174/R176 2k2 plus R265/R147 10kP3V3MP
    CLK_SDA/CLK_SCLjoined to I2C_CPU through IC56 whenever P3V3 is upR314/R313 2k2P3V3

    The numbers

    QuantityValueSource
    Combined pull-up, channel 4 selected, IC56 closed2k2 ‖ 2k2 ‖ 10k ‖ 2k2 = about 660 Ωderived
    Sink current to reach VOL = 0.4 V(3.3 - 0.4) / 660 = 4.4 mAderived
    I2C_CPU alone1.8 k, 1.6 mAderived
    I2C_CPU with IC56 closed0.99 k, 2.9 mAderived
    MAX6642 (IC18) sink1 mA at 0.4 V, 4 mA at 0.6 VMAX6642 datasheet, p.2
    PCA9547BS pinning19 SCL, 20 SDA upstream; 10 SD4, 11 SC4PCA9547BS datasheet, Fig. 4, p.4

    Each segment alone is fine. Only the combined state is too strong.

    Why it matters

    Above the 3 mA standard/fast-mode sink level, the weaker slaves (MAX6642, LM75A, INA220, EEPROMs) cannot pull the line to a clean low. That erodes the low-level noise margin, only in this one routing state.

    Suggested fix: drop a pair of the duplicated 2k2s (e.g. on FPGA_SCL/SDA or CLK_SCL/SDA), or avoid selecting channel 4 while IC56 is closed.

  22. observationOBS-458minorOpen

    USB-UART B: CP2102 TXD (3.3 V) enters FPGA bank 18 (FMC1 VADJ) directly through 2k2, while the opposite direction is translated by IC69

    The CP2102 (IC62B) TXD drives 3.3 V straight into FPGA ball F16 in bank 18 (VCCO = PVADJ1) through R275B 2k2 only, while the opposite direction of the same UART goes through the TXB0104 translator IC69.

    Where it is

    ``mermaid flowchart LR U["IC62B CP2102<br/>I/O from own 3.3 V regulator, pin 6 VDD"] -- "pin 26 TXD, USB_TXDB" --> R1["R275B 2k2"] R1 -- "TXD_USBB, also TP4" --> F["IC79 ball F16<br/>IO_25_18, bank 18<br/>VCCO = PVADJ1"] F2["FPGA, VADJ1_RXD"] --> T["IC69 TXB0104<br/>A4 to B4"] T -- "RXD_USBB" --> R2["R276B 2k2"] R2 -- "USB_RXDB" --> U2["IC62B pin 25 RXD"] ``

    • CP2102 to FPGA (TXD, pin 26): no translator.
    • FPGA to CP2102 (RXD, pin 25): translated by IC69 (A4 VADJ1_RXD -> B4 RXD_USBB).

    The design translates one direction of the link and not the other.

    The numbers

    QuantityValueSource
    CP2102 pin 26TXD, 'Asynchronous data output (UART Transmit)'CP2102 datasheet, p.11
    HR-bank I/O input limitVCCO + 0.55 VDS182 (XC7K325T), p.1
    Bank 18 VCCOPVADJ1, the FMC1 VADJschematic

    Why it matters

    Whenever FMC1 VADJ is below 3.3 V (1.8 V or 2.5 V mezzanines), a 3.3 V high on F16 exceeds the HR-bank limit of VCCO + 0.55 V and is clamped by the FPGA through 2k2: sub-mA, but outside ratings. At VADJ1 = 1.8 V the FPGA input sees a signal well above its supply.

    Suggested fix: route USB_TXDB through a spare translator channel (IC69 has no free channel; IC71 B4/pin 10 is unused) or add a divider.

  23. observationOBS-459minorOpen

    Silkscreen touches solder-mask openings on 147 pads, mostly the 0603 LED outlines and the 2512 current-sense shunts

    Silkscreen crosses the solder-mask openings of 147 pads (95 top, 52 bottom), worst on the 0603 LED outlines and the 2512 current-sense shunts. Cosmetic and ink-on-pad risk, not a functional fault.

    How it was measured

    The TOP/BOTTOM_OVERLAY strokes, buffered by their stroke width, were intersected with the TOP/BOTTOM_SOLDER mask openings in this board's ODB++ layers. This is measured from the layer geometry, not from a render. Overlaps reach about 70 mil² (about 5 % of a 0603 pad).

    Worst cases

    WherePartsNetsExample overlap
    Top, led_0603R outline hugging both padsLED3/LED4/LED5LSP_LED_A0-A2LSP_LED_A0 (LED3 anode side) at (-121.60, 68.40) mm: ~65 of 1225 mil²
    Top, LED outlineLED6/LED7FMC1/2_12P0V_LED
    Top, LED outlineLED16DONE_LED/DONE_LED_R
    Bottom, LED outlinesLED8-LED15e.g. NetLED12_A/13_A, LED11_R, Q5_LEDNetLED12_A (LED12) at (-38.12, -2.07) mm: ~62 of 1225 mil²
    2512 shunt outlinesR75/R257/R266/R2672.5V_FMC1/2, PVADJ1/2, NetQ2_5, POWER2.12P0VPVADJ1 (R75 pad) at (-41.98, -1.75) mm: ~44 of 11057 mil²
    Single padFPGA_SDAnear (-154.5, 88.7) mm

    The LED polarity dots sit clear of the pads, so the polarity information survives.

    Why it matters

    Board houses clip silk from mask openings, so the result is ragged outlines and ink risk on solderable surfaces rather than a functional fault. Pulling the footprint outlines back by about 0.1 mm removes it.

  24. observationOBS-460minorOpen

    No bring-up test points on the supply rails, resets or configuration signals: the board's test points are GND studs plus two UART lines

    The AFCK has no bring-up test point on any regulated rail, reset or configuration signal: every test-point part is on GND, on one of two UART lines, or on the unfitted IC67 footprint.

    What the test points are on

    Every test-point part on the board was checked against its net:

    • TPD2 ... TPD38 (21 pads) and B5-B8: all GND
    • TP4: TXD_USBB
    • TP13: VADJ1_RXD
    • TP1, TP2, TP10: only on pins of the unfitted IC67 footprint (TP1 = IC67.86, DNP MCU)

    What has no probe point

    • Regulated rails: P1V0, P1V2, P1V5, P1V8, P2V0, P3V3, P3V3MP, PVADJ1/2, VTT
    • Resets: RESETN, FPGA_RESETN, FDDR.RST_N
    • Configuration signals: DONE, INIT_B, PROGRAM_B

    Why it matters

    Bring-up then relies on probing 0201/0402 capacitor pads or the indicator LEDs. A pad per rail and per reset makes first power-on and field diagnosis much quicker and safer, and is low cost to add on a respin.

  25. observationOBS-461minorOpen

    Configuration-flash sizing note on schematic sheet 5 is for an XC7A200T, not the fitted XC7K325T

    The flash-sizing note beside the configuration flashes IC13/IC64 on sheet 5 is written for an XC7A200T, but the FPGA fitted on this board (IC79) is an XC7K325T-2FFG900C.

    ItemValueSource
    Note beside IC13/IC64'XC7A200T 65,330,912 bits -> 16MB memory'schematic sheet 5
    FPGA fitted, IC79XC7K325T-2FFG900CBOM: IC79

    The note appears to be carried over from the Artix-7 based AFC design.

    Why it matters

    The note is the only statement on the drawing of why a 128 Mbit flash was chosen. The Kintex-7 325T bitstream is larger than the 200T's, so the figure a reader takes from the drawing is wrong. The 128 Mbit part still holds one 325T image, but with less spare room than the note implies.

    Suggested fix: update the note to the XC7K325T's bitstream size.

  26. observationOBS-479minorOpen

    P1V2 and P2V0 bucks (IC55, IC57) default ON while the MMC is in reset: their enables have no pull-down and the LPC1764 pins power up with pull-ups

    On the AFCK, the P1V2 buck IC55 and the P2V0 buck IC57 (both NCP3170) turn on by themselves whenever P12V is present and the MMC is in reset: their enables have no pull-down, and the MMC's pins power up with their pull-ups on. Every other payload rail enable on the board has a pull-down.

    Where it is

    EnableMMC pinSeries resistorBuckOutputPull-down
    PM.EN_P1V2 to EN_P1V2_RIC1 pin 9R103, 10 kIC55 EN pin 6P1V2 (via L2)none
    PM.EN_P1V8 to EN_P1V8_RIC1 pin 8R61, 10 kIC57 EN pin 6P2V0 (via L7), which feeds the P1V8 LDO IC81none

    The only other connection on each net is the footprint IC67 (part string DNP-ATxmega128A1U-AU, the unfitted alternative MCU). By contrast the other payload enables have 10 k or 4k7 pull-downs: R49, R69, R71, R78, R85, R92 and R114 (PM.EN_P1V0).

    The numbers

    QuantityValueSource
    LPC1764 standard I/O at resetinput with pull-up enabledLPC1764 datasheet p.75
    LPC1764 pull-up current, min/typ/max15 / 50 / 85 µALPC1764 datasheet p.48
    NCP3170 EN threshold1.41 VNCP3170 datasheet p.2
    NCP3170 EN pin guidance"Pull EN to logic low to disable the device. Do not leave it open."NCP3170 datasheet p.2
    EN level with only the MMC pull-up behind 10 kpulled toward 3.3 V, well above 1.41 Vderived

    ``mermaid flowchart LR A["IC1 LPC1764 pin 9, pull-up on in reset"] -- "R103 10 k" --> B["IC55 EN, no pull-down"] B --> C["P1V2 on (MGTAVTT via FB7)"] D["IC1 LPC1764 pin 8, pull-up on in reset"] -- "R61 10 k" --> E["IC57 EN, no pull-down"] E --> F["P2V0 on, then P1V8 via IC81"] ``

    Why it matters

    • While the MMC is held in reset, unprogrammed or rebooting with P12V present, P1V2, P2V0 and through IC81 P1V8 (VCCAUX) come up without the MMC's sequencing, while VCCINT (P1V0, IC14, enable held low by R114) stays off. That is the opposite of the Kintex-7 recommended VCCINT-first order.
    • P1V2 is MGTAVTT (via FB7) and P1V0 is MGTAVCC (via FB6). DS182 p.7 warns that with MGTAVTT powered before MGTAVCC the MGTAVTT current "can increase by 460 mA per transceiver" during the MGTAVCC ramp.
    • In a crate P12V normally follows MMC activation. But the J4 stand-alone supply feeds P12V (via D5/D6) and the MMC together, and an MMC reset with payload power on hits the same path.

    Suggested fix: a pull-down on PM.EN_P1V2 and PM.EN_P1V8, like the other enables. Needs the designer's view on the intended behaviour.

  27. observationOBS-480minorOpen

    The six TPS53126 enables (IC10, IC12, IC43) sit between the 0.3 V low and 2.0 V high thresholds while the MMC is in reset

    While the MMC (IC1) is in reset, the six TPS53126 enables of IC10, IC12 and IC43 are not guaranteed off: the LPC1764 pull-ups fight the enable pull-downs and can hold them above the 0.3 V guaranteed-off level.

    Where it is

    Each TPS53126 is a dual converter with two enables, so the three parts have six. Each enable is an MMC GPIO with a pull-down to GND:

    Enable netConverter pinMMC pinPull-down
    PM.EN_FMC1_PVADJIC10 pin 24IC1 pin 20R71, 10 k
    PM.EN_FMC2_PVADJIC10 pin 7IC1 pin 44R78, 10 k
    PM.EN_FMC1_P3V3IC12 pin 7IC1 pin 6R92, 10 k
    PM.EN_FMC2_P3V3IC12 pin 24IC1 pin 7R85, 10 k
    PM.EN_P3V3IC43 pin 7IC1 pin 43R69, 4k7
    PM.1V5_VTT_ENIC43 pin 24IC1 pin 45R49, 4k7

    Each net also reaches the unfitted alternative MCU footprint IC67.

    The numbers

    QuantityValueSource
    LPC1764 pin state in resetinput with pull-up enabledLPC1764 datasheet p.75
    LPC1764 pull-up current, min/typ/max15 / 50 / 85 µALPC1764 datasheet p.48
    EN level with 10 k (four enables)about 0.15 / 0.50 / 0.85 Vderived, I x R
    EN level with 4k7 (two enables)about 0.07 / 0.24 / 0.40 Vderived, I x R
    TPS53126 ENx L-level max / H-level min0.3 V / 2.0 VTPS53126 datasheet p.6

    The pull-up current is specified at 0 V, so at the enable voltage it is a little lower. The figures are upper bounds.

    Why it matters

    • The four 10 k enables (FMC1/FMC2 VADJ and FMC1/FMC2 3.3 V) sit near 0.5 V with typical pull-up current, outside the guaranteed-off band.
    • The two 4k7 enables (P3V3 and 1V5 VTT) are inside the band at typical current and only leave it at the 85 µA maximum.
    • So these rails are not guaranteed off before firmware runs. In practice the real enable threshold is probably well above 0.5 V, which is why this is minor.

    Suggested fix: a smaller pull-down (about 2k2) would restore the guaranteed state. Designer's judgement needed.

  28. observationOBS-482minorOpen

    JTAG ESD array IC7 sits beside the SCANSTA111 (IC6), about 105 mm from J6 and 171 mm from the AMC-edge JTAG contacts it is meant to protect

    On the AFCK, the JTAG ESD array IC7 (BZA462A) is placed next to the device it protects, the SCANSTA111 IC6, and not at either port where a transient comes in: it is about 105 mm from the JTAG header J6 and about 171 mm from the AMC-edge JTAG contacts on P1 (straight line).

    Where it is

    IC7 clamps AJTAG.TMS, TDI, TCK and TRST#. These nets enter at J6 (bottom side) and at the AMC edge P1 (pads 165-169), then pass 100R series resistors (R227 on TCK, R229 on TMS, ...) into IC6.

    ItemPosition (x, y), mmSource
    IC7 BZA462A (top)(-138.5, 72.2)placement
    IC6 SCANSTA111 (top)(-133.5, 73.1)placement
    R227 / R229 100R (top)(-135.6, 72.7) / (-135.6, 71.7)placement
    J6 JTAG header (bottom)(-93.9, -23.3)placement
    P1 pad 165 AJTAG.TCK / 166 AJTAG.TMS(-0.76, -27.75) / (-0.76, -28.5)pad list
    QuantityValueSource
    IC7 to J6, straight lineabout 105 mmderived from placement
    IC6 to J6, straight lineabout 104 mmderived from placement
    IC7 to P1 pad 166, straight lineabout 171 mmderived from placement and pads
    IC6 to P1 pad 166, straight lineabout 167 mmderived from placement and pads
    IC7 to IC6about 5 mmderived from placement

    Straight-line distances are a lower bound on the routed length.

    ``mermaid flowchart LR A["J6 header (bottom)"] -- "about 105 mm, unclamped" --> C["IC7 BZA462A clamp"] B["AMC edge P1 pads 165-169"] -- "about 171 mm, unclamped" --> C C -- "about 3 mm" --> D["R227 / R229 100R"] D --> E["IC6 SCANSTA111"] ``

    What the datasheet says

    The BZA462A layout guidance (datasheet p.8) says "Circuit board layout is of extreme importance in the suppression of transients", that the parasitic inductance of series conductors adds to the clamping voltage, and "Place the suppression element close to the input terminals or connectors."

    Why it matters

    • A transient injected at J6 or at the AMC edge travels 100 mm or more of unprotected trace before it is clamped, and can couple into neighbouring nets on the way.
    • IC7's ground pins return the clamp current next to IC6 and the FPGA, not at either port.
    • The 100R series resistors and the TVS still protect IC6 electrically, which is why this is minor.

    Suggested fix: move the array, or add a second one, next to J6 and next to P1, so the clamp sits at the port.

  29. observationOBS-664minorOpen

    OSC4 is an Si570 (pin 1 NC, 'make no external connection') but pin 1 is wired to the PLL DAC tuning node PLLDAC_OUT3

    OSC4 is fitted as an Si570 XO (570BBC000121DG), whose pin 1 is a no-connect the datasheet says must have no external connection, yet the board wires pin 1 to the PLL DAC tuning node PLLDAC_OUT3; the footprint is laid out for an Si571 VCXO, where pin 1 is the control voltage VC.

    ItemValueSource
    OSC4 part570BBC000121DG (Si570, I2C-programmable XO)BOM row OSC4
    Si570 pin 1NC: "No Connect. Make no external connection to this pin."Si570/571 datasheet d8337913 p29, Table 14
    OSC4.1 netPLLDAC_OUT3, with C425 and C606 (100 nF), R355 240 ohm, and R224 100 ohm on to SI57X_TUNE (IC41 B6, R356 to PLL_DAC3_SYNC_N)netlist
    Designer's note"Can be used to fine tune Si571, which has a control voltage input."schematic p24

    Why it matters

    The tuning path is a deliberate Si571 option (the designer's note on p24), but the BOM fits an Si570, for which the vendor asks for pin 1 to be left open. Driving it from the DAC node is probably harmless if the pin is unbonded, but that is not stated, and the tuning network does nothing with this oscillator. Either the BOM should name an Si571 (VCXO) when voltage tuning is wanted, or the pin-1 connection should be left unpopulated for the Si570 build. A person should confirm which part the build intends.

  30. observationOBS-665minorOpen

    AMC geographic-address ESD array IC5 (BZA462A) returns to ground through a single via

    IC5 (BZA462A quad TVS on the AMC geographic-address lines GA0-GA2 from P1) reaches the GND plane through a single via, so that via's inductance sits in series with every clamp event; the datasheet asks to minimise the transient return path to ground.

    ItemValueSource
    IC5 ground pins2 and 5 on GND, 2.7 mm apart, bottom sidenetlist
    GND vias within 3 mm of pin 2one, 0.89 mmstored drill-layer vias
    GND vias within 3 mm of pin 5nonestored drill-layer vias
    Lines protectedADD.GA0 / GA1 / GA2 (AMC P1.5 / P1.17 / P1.26)netlist

    Contrast on the same board: IC4 (same part, IPMB) has five vias within 3 mm of pin 2 and one at pin 5; D7A/D7B (USB) have four each.

    Why it matters

    The geographic-address lines are touched on every insertion into the crate. Nexperia's BZA462A guidelines (datasheet p8) ask to minimise the length of the transient return path to ground. A second via at the ground pins, as at IC4, removes the single-via inductance. Whether this matters at the ESD level the crate needs is the owner's call.

  31. observationOBS-675minorOpen

    BC847B went end-of-life (Not For New Designs)

    Vendor (DigiKey) lifecycle for BC847B (Nexperia USA Inc.'s listing) changed from "Active" to "Not For New Designs" since this board was reviewed. Placed as Q12, Q5, Q6. A sourcing/longevity risk; consider a replacement before production.

  32. concernCON-001minorOpen

    [eol:BC847B:Not For New Designs] BC847B lifecycle changed from "Active" to "Not For New Designs" since this board was reviewed: OHWR0032 R1 AFCK. A sourcing and longevity risk. Should this be replaced before the next build?

  33. ReferencePlaneDiscontinuityminorOpen

    3 nets cross one gap in P6's negative-plane field at 3 places, the longest 0.38 mm (fastest: FPGA_I2C_SCL at 10 ns 10-90 %, suggested by its name)

    Along 'FPGA_I2C_SDA' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.38 mm under the track at (-102.36, 100.51). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 3 places on 3 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -114.16..-82.56, y 100.32..100.7. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

    • 'FPGA_I2C_SCL' on L5 at (-102.12, 100.51), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
    • 'FPGA_I2C_SDA' on L5 at (-102.36, 100.51), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
    • 'RST' on L5 at (-100.73, 100.51), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

    Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-114.16%2C100.32%7C-82.56%2C100.7 serves each with its own layer_image region; all of them lie in x -103.36..-99.73, y 99.32..101.7 mm (board frame, Y up).

    Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    At gap:P6:hole -114.16,100.32|-82.56,100.7

  34. DecapViaFanoutminorOpen

    DecapViaFanout: 3 components×3

    3 checked by the review by Claude walker W4 (Cern session) — 3 different reasons

    At C235, C4B, C770

    Each finding, with its own headline
    • C235C235 on 'VCCBRAM': its rail pad reaches its via over 1.95 mm of track, adding about 0.60 nH to its loop, past Probe's 0.5 nH starting value; put the vias at the pads (a 100 uF bulk capacitor: its inductance matters only far above the band it serves)

      C235 on 'VCCBRAM': its rail pad reaches its via over 1.95 mm of track, adding about 0.60 nH to its loop, past Probe's 0.5 nH starting value; put the vias at the pads (a 100 uF bulk capacitor: its inductance matters only far above the band it serves). C235.1 (rail) runs 1.95 mm along its copper to its nearest via, 0.95 mm past the 1 mm beside the pad: 0.60 nH; C235.2 (ground) not measured: the pad sits in a pour of its net on its own layer, which has no track length to measure. Each run past the pad as one conductor of a pair 3.00 mm apart (the pad pitch), 0.2 l ln(D/R) nH with R 0.127 mm (the track's width is not read): 0.60 nH in all. Put the vias at the pads (or in them), one per pad at least, the rail's and the ground's close together; the capacitor's own inductance as mounted is 1.57 nH (pad 0.20 nH + no via of the first net within 1 mm of its pad: the study's 0.6 nH per mm of 0.95 mm depth, ASSUMED). 0.5 nH is Probe's starting value, about the whole mounted inductance of a well-mounted 0402.

    • C4BC4B on 'FMC2_P3V3': its rail pad reaches its via over 2.62 mm of track, adding about 0.55 nH to its loop, past Probe's 0.5 nH starting value; put the vias at the pads (within a quarter of the starting value)

      C4B on 'FMC2_P3V3': its rail pad reaches its via over 2.62 mm of track, adding about 0.55 nH to its loop, past Probe's 0.5 nH starting value; put the vias at the pads (within a quarter of the starting value). C4B.2 (rail) runs 2.62 mm along its copper to its nearest via, 1.62 mm past the 1 mm beside the pad: 0.55 nH; C4B.1 (ground) not measured: the pad sits in a pour of its net on its own layer, which has no track length to measure. Each run past the pad as one conductor of a pair 0.70 mm apart (the pad pitch), 0.2 l ln(D/R) nH with R 0.127 mm (the track's width is not read): 0.55 nH in all. Put the vias at the pads (or in them), one per pad at least, the rail's and the ground's close together; the capacitor's own inductance as mounted is 0.70 nH (pad 0.05 nH + no via of the first net within 1 mm of its pad: the study's 0.6 nH per mm of 0.74 mm depth, ASSUMED). 0.5 nH is Probe's starting value, about the whole mounted inductance of a well-mounted 0402.

    • C770C770 on 'PVADJ1': its rail pad reaches its via over 2.79 mm of track, adding about 0.61 nH to its loop, past Probe's 0.5 nH starting value; put the vias at the pads (within a quarter of the starting value)

      C770 on 'PVADJ1': its rail pad reaches its via over 2.79 mm of track, adding about 0.61 nH to its loop, past Probe's 0.5 nH starting value; put the vias at the pads (within a quarter of the starting value). C770.1 (rail) runs 2.79 mm along its copper to its nearest via, 1.79 mm past the 1 mm beside the pad: 0.61 nH; C770.2 (ground) not measured: the pad sits in a pour of its net on its own layer, which has no track length to measure. Each run past the pad as one conductor of a pair 0.70 mm apart (the pad pitch), 0.2 l ln(D/R) nH with R 0.127 mm (the track's width is not read): 0.61 nH in all. Put the vias at the pads (or in them), one per pad at least, the rail's and the ground's close together; the capacitor's own inductance as mounted is 0.70 nH (pad 0.05 nH + no via of the first net within 1 mm of its pad: the study's 0.6 nH per mm of 0.74 mm depth, ASSUMED). 0.5 nH is Probe's starting value, about the whole mounted inductance of a well-mounted 0402.

  35. ReferencePlaneDiscontinuityminorOpen

    ReferencePlaneDiscontinuity: 142 boards×142

    142 checked by the review by Claude walker W4 (Cern session) — 124 different reasons

    At along@P4:MGTAVTT|P1V5, along@P4:PVADJ1|PVADJ2_1V8, along@P6:GND|MGTVCCAUX, along@P6:GND|P12V, along@P6:VCCAUX|VCCBRAM, gap:G1:hole -15.6,8.59|-13.16,9.5 and 136 more

    Each finding, with its own headline — 100 of 142 shown
    • along@P4:MGTAVTT|P1V5'M1' runs ALONG the split between the MGTAVTT and P1V5 planes on P4 at 1 place, the longest 9.76 mm; the impedance rises from 44 ohm to 49 ohm there at worst, 4.7 % of each edge reflected (fastest: M1 at 100 ps 10-90 %, ASSUMED)

      On P4 the MGTAVTT and P1V5 planes part, and 'M1' runs down the slot between them at 1 place(s), rather than across it. This is not a crossing: the return still follows the track, in the plane on its other side and in the slot's edges, so it does not detour. What changes is the track's own impedance. With the plane gone from under it on this side for that length, its field reaches further, to the slot's edges and to the next plane beyond, so its capacitance falls and its impedance rises; every edge travelling down it is partly reflected where it enters the stretch and again where it leaves, by (Z2 - Z1)/(Z2 + Z1). The longer the stretch against the edge's rise, the more of the edge sees the full step. A crossing is generally the worse fault; a mismatch is not free. Over solid plane each track is 44 ohm; along the slot 49 ohm, reflecting up to 4.7 % of each edge (a method-of-moments field solve of the cross-section, one dielectric; were the slot's edges to carry none of the return, up to 50 ohm). At every place the rise is within 10 %, the tolerance controlled impedance is commonly built to (the stackup states none), so the finding is the lower tier (info). One decision covers the split line's slot: whether these tracks may run down it, or should be moved over solid plane.

      • 'M1' on L3 at (-78.7, 92.97), 10.53 mm, running ALONG the split for 9.76 mm, not across it, 44 ohm over solid plane, 49 ohm along it: 4.7 % of each edge reflected; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=along%40P4%3AMGTAVTT%7CP1V5 serves each with its own layer_image region; all of them lie in x -86.76..-74.36, y 91.97..94.27 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • along@P4:PVADJ1|PVADJ2_1V8'FSPI.FCS_B\' runs ALONG the split between the PVADJ1 and PVADJ2_1V8 planes on P4 at 1 place, the longest 1.36 mm; the impedance rises from 49 ohm to 55 ohm there at worst, 5.3 % of each edge reflected (fastest: FSPI.FCS_B\ at 2 ns 10-90 %, suggested by its name)

      On P4 the PVADJ1 and PVADJ2_1V8 planes part, and 'FSPI.FCS_B\' runs down the slot between them at 1 place(s), rather than across it. This is not a crossing: the return still follows the track, in the plane on its other side and in the slot's edges, so it does not detour. What changes is the track's own impedance. With the plane gone from under it on this side for that length, its field reaches further, to the slot's edges and to the next plane beyond, so its capacitance falls and its impedance rises; every edge travelling down it is partly reflected where it enters the stretch and again where it leaves, by (Z2 - Z1)/(Z2 + Z1). The longer the stretch against the edge's rise, the more of the edge sees the full step. A crossing is generally the worse fault; a mismatch is not free. Over solid plane each track is 49 ohm; along the slot 55 ohm, reflecting up to 5.3 % of each edge (a method-of-moments field solve of the cross-section, one dielectric; were the slot's edges to carry none of the return, up to 55 ohm). No place stays within 10 %, the tolerance controlled impedance is commonly built to (the stackup states none), so the finding is a warning. One decision covers the split line's slot: whether these tracks may run down it, or should be moved over solid plane.

      • 'FSPI.FCS_B\' on L3 at (-66.32, 61.56), 1.99 mm, running ALONG the split for 1.36 mm, not across it, 49 ohm over solid plane, 55 ohm along it: 5.3 % of each edge reflected; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=along%40P4%3APVADJ1%7CPVADJ2_1V8 serves each with its own layer_image region; all of them lie in x -68.38..-64.47, y 60.56..62.75 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • along@P6:GND|MGTVCCAUX2 nets run ALONG the split between the GND and MGTVCCAUX planes on P6 at 2 places, the longest 8.89 mm; the impedance rises from 44 ohm to 49 ohm there at worst, 5.4 % of each edge reflected (fastest: FPGA_I2C_SCL at 10 ns 10-90 %, suggested by its name)

      On P6 the GND and MGTVCCAUX planes part, and 2 nets run down the slot between them at 2 place(s), rather than across it. This is not a crossing: the return still follows the track, in the plane on its other side and in the slot's edges, so it does not detour. What changes is the track's own impedance. With the plane gone from under it on this side for that length, its field reaches further, to the slot's edges and to the next plane beyond, so its capacitance falls and its impedance rises; every edge travelling down it is partly reflected where it enters the stretch and again where it leaves, by (Z2 - Z1)/(Z2 + Z1). The longer the stretch against the edge's rise, the more of the edge sees the full step. A crossing is generally the worse fault; a mismatch is not free. Over solid plane each track is 44 ohm; along the slot 46 ohm to 49 ohm, reflecting up to 5.4 % of each edge (a method-of-moments field solve of the cross-section, one dielectric; were the slot's edges to carry none of the return, up to 50 ohm). At 1 of the 2 place(s) the rise is within 10 %, the tolerance controlled impedance is commonly built to (the stackup states none); the rest exceed it or cannot be said, so the finding is a warning. One decision covers the split line's slot: whether these tracks may run down it, or should be moved over solid plane.

      • 'FPGA_I2C_SCL' on L5 at (-86.52, 92.1), 9.26 mm, running ALONG the split for 8.89 mm, not across it, 44 ohm over solid plane, 49 ohm along it: 5.4 % of each edge reflected; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'FPGA_I2C_SDA' on L5 at (-86.51, 91.86), 8.83 mm, running ALONG the split for 8.67 mm, not across it, 44 ohm over solid plane, 46 ohm along it: 2.4 % of each edge reflected; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=along%40P6%3AGND%7CMGTVCCAUX serves each with its own layer_image region; all of them lie in x -91.97..-80.81, y 90.84..93.1 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • along@P6:GND|P12V'FCLKA_P' runs ALONG the split between the GND and P12V planes on P6 at 1 place, the longest 9.33 mm; the impedance rises from 52 ohm to 57 ohm there at worst, 4.3 % of each edge reflected (fastest: FCLKA_P at 1 ns 10-90 %, suggested by its name)

      On P6 the GND and P12V planes part, and 'FCLKA_P' runs down the slot between them at 1 place(s), rather than across it. This is not a crossing: the return still follows the track, in the plane on its other side and in the slot's edges, so it does not detour. What changes is the track's own impedance. With the plane gone from under it on this side for that length, its field reaches further, to the slot's edges and to the next plane beyond, so its capacitance falls and its impedance rises; every edge travelling down it is partly reflected where it enters the stretch and again where it leaves, by (Z2 - Z1)/(Z2 + Z1). The longer the stretch against the edge's rise, the more of the edge sees the full step. A crossing is generally the worse fault; a mismatch is not free. Over solid plane each track is 52 ohm; along the slot 57 ohm, reflecting up to 4.3 % of each edge (a method-of-moments field solve of the cross-section, one dielectric; were the slot's edges to carry none of the return, up to 58 ohm). At every place the rise is within 10 %, the tolerance controlled impedance is commonly built to (the stackup states none), so the finding is the lower tier (info). One decision covers the split line's slot: whether these tracks may run down it, or should be moved over solid plane.

      • 'FCLKA_P' on L5 at (-91.28, -1.95), 9.8 mm, running ALONG the split for 9.33 mm, not across it, 52 ohm over solid plane, 57 ohm along it: 4.3 % of each edge reflected; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=along%40P6%3AGND%7CP12V serves each with its own layer_image region; all of them lie in x -96.06..-87.07, y -6.59..2.26 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • along@P6:VCCAUX|VCCBRAM3 nets run ALONG the split between the VCCAUX and VCCBRAM planes on P6 at 3 places, the longest 3.76 mm; the impedance rises from 52 ohm to 56 ohm there at worst, 3 % of each edge reflected (fastest: HB2.HB02_N at 100 ps 10-90 %, ASSUMED)

      On P6 the VCCAUX and VCCBRAM planes part, and 3 nets run down the slot between them at 3 place(s), rather than across it. This is not a crossing: the return still follows the track, in the plane on its other side and in the slot's edges, so it does not detour. What changes is the track's own impedance. With the plane gone from under it on this side for that length, its field reaches further, to the slot's edges and to the next plane beyond, so its capacitance falls and its impedance rises; every edge travelling down it is partly reflected where it enters the stretch and again where it leaves, by (Z2 - Z1)/(Z2 + Z1). The longer the stretch against the edge's rise, the more of the edge sees the full step. A crossing is generally the worse fault; a mismatch is not free. Over solid plane each track is 52 ohm; along the slot 54 ohm to 56 ohm, reflecting up to 3 % of each edge (a method-of-moments field solve of the cross-section, one dielectric; were the slot's edges to carry none of the return, up to 58 ohm). At every place the rise is within 10 %, the tolerance controlled impedance is commonly built to (the stackup states none), so the finding is the lower tier (info). One decision covers the split line's slot: whether these tracks may run down it, or should be moved over solid plane.

      • 'HB2.HB02_N' on L5 at (-52.5, 66.67), 4.33 mm, running ALONG the split for 3.76 mm, not across it, 52 ohm over solid plane, 56 ohm along it: 3 % of each edge reflected; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB16_P' on L5 at (-60.2, 66.66), 2.84 mm, running ALONG the split for 2.64 mm, not across it, 52 ohm over solid plane, 55 ohm along it: 2.1 % of each edge reflected; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB16_N' on L5 at (-60.67, 66.44), 3.2 mm, running ALONG the split for 2.72 mm, not across it, 52 ohm over solid plane, 54 ohm along it: 1.8 % of each edge reflected; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=along%40P6%3AVCCAUX%7CVCCBRAM serves each with its own layer_image region; all of them lie in x -62.72..-48.58, y 65.03..67.72 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:G1:hole -15.6,8.59|-13.16,9.52 nets cross one gap in G1's negative-plane field at 2 places, the longest 0.96 mm (fastest: TX13_N at 3 ns 10-90 %, suggested by its name)

      Along 'TX13_N' on TOP, G1 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.96 mm under the track at (-15.46, 9.08). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 2 places on 2 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -15.6..-13.16, y 8.59..9.5. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. No crossing has a plane on the track's other side, so the whole return meets the break.

      • 'TX13_N' on TOP at (-15.46, 9.08), 0.96 mm; other side: none: the signal layer is outermost
      • 'TX13_P' on TOP at (-15.46, 8.66), 0.29 mm; other side: none: the signal layer is outermost

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AG1%3Ahole%20-15.6%2C8.59%7C-13.16%2C9.5 serves each with its own layer_image region; all of them lie in x -16.6..-13.87, y 7.59..10.21 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:G1:hole -15.62,9.86|-13.18,10.772 nets cross one gap in G1's negative-plane field at 2 places, the longest 1.75 mm (fastest: RX13_N at 3 ns 10-90 %, suggested by its name)

      Along 'RX13_P' on TOP, G1 is a negative layer (copper everywhere inside the profile), and its field is cleared for 1.75 mm under the track at (-14.96, 10.68). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 2 places on 2 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -15.62..-13.18, y 9.86..10.77. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. No crossing has a plane on the track's other side, so the whole return meets the break.

      • 'RX13_P' on TOP at (-14.96, 10.68), 1.75 mm; other side: none: the signal layer is outermost
      • 'RX13_N' on TOP at (-13.36, 10.11), 1.62 mm; other side: none: the signal layer is outermost

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AG1%3Ahole%20-15.62%2C9.86%7C-13.18%2C10.77 serves each with its own layer_image region; all of them lie in x -16.05..-12.18, y 8.86..11.77 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:G1:hole -15.65,6.05|-13.21,6.962 nets cross one gap in G1's negative-plane field at 2 places, the longest 2.32 mm (fastest: RX14_N at 3 ns 10-90 %, suggested by its name)

      Along 'RX14_N' on TOP, G1 is a negative layer (copper everywhere inside the profile), and its field is cleared for 2.32 mm under the track at (-13.42, 6.28). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 2 places on 2 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -15.65..-13.21, y 6.05..6.96. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. No crossing has a plane on the track's other side, so the whole return meets the break.

      • 'RX14_N' on TOP at (-13.42, 6.28), 2.32 mm; other side: none: the signal layer is outermost
      • 'RX14_P' on TOP at (-14.49, 6.12), 1.64 mm; other side: none: the signal layer is outermost

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AG1%3Ahole%20-15.65%2C6.05%7C-13.21%2C6.96 serves each with its own layer_image region; all of them lie in x -16.65..-12.21, y 5.05..7.96 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:G1:hole -26.98,8.59|-24.54,9.52 nets cross one gap in G1's negative-plane field at 2 places, the longest 1.09 mm (fastest: TX15_N at 3 ns 10-90 %, suggested by its name)

      Along 'TX15_P' on TOP, G1 is a negative layer (copper everywhere inside the profile), and its field is cleared for 1.09 mm under the track at (-24.92, 9.11). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 2 places on 2 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -26.98..-24.54, y 8.59..9.5. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. No crossing has a plane on the track's other side, so the whole return meets the break.

      • 'TX15_P' on TOP at (-24.92, 9.11), 1.09 mm; other side: none: the signal layer is outermost
      • 'TX15_N' on TOP at (-24.71, 9.33), 0.49 mm; other side: none: the signal layer is outermost

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AG1%3Ahole%20-26.98%2C8.59%7C-24.54%2C9.5 serves each with its own layer_image region; all of them lie in x -26.31..-23.54, y 7.73..10.5 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:G1:hole -27,9.86|-24.56,10.772 nets cross one gap in G1's negative-plane field at 2 places, the longest 1.29 mm (fastest: TX15_N at 3 ns 10-90 %, suggested by its name)

      Along 'TX15_N' on TOP, G1 is a negative layer (copper everywhere inside the profile), and its field is cleared for 1.29 mm under the track at (-25.7, 10.31). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 2 places on 2 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -27..-24.56, y 9.86..10.77. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. No crossing has a plane on the track's other side, so the whole return meets the break.

      • 'TX15_N' on TOP at (-25.7, 10.31), 1.29 mm; other side: none: the signal layer is outermost
      • 'TX15_P' on TOP at (-26.12, 10.31), 1.29 mm; other side: none: the signal layer is outermost

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AG1%3Ahole%20-27%2C9.86%7C-24.56%2C10.77 serves each with its own layer_image region; all of them lie in x -27.58..-24.24, y 8.86..11.77 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:G5:hole -26.78,46.53|-25.78,47.65'DP2.DP1_M2C_P' crosses one gap in G5's negative-plane field at 1 place, the longest 1.12 mm (fastest: DP2.DP1_M2C_P at 100 ps 10-90 %, ASSUMED)

      Along 'DP2.DP1_M2C_P' on L6, G5 is a negative layer (copper everywhere inside the profile), and its field is cleared for 1.12 mm under the track at (-26.6, 47.09). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 1 of them run ALONG the bare copper, parallel to its edges, rather than across it (the longest 1.12 mm). What cleared it: the hole at x -26.78..-25.78, y 46.53..47.65. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 52% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'DP2.DP1_M2C_P' on L6 at (-26.6, 47.09), 1.12 mm, running ALONG the bare copper for 1.12 mm, not across it; other side: (negative-layer field) on G4, 0.11 mm from the track against 0.1 mm to G5: 52% of the return on G5

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AG5%3Ahole%20-26.78%2C46.53%7C-25.78%2C47.65 serves each with its own layer_image region; all of them lie in x -27.6..-25.6, y 45.53..48.65 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:G5:hole -94.49,-10.8|-93.37,-9.82 nets cross one gap in G5's negative-plane field at 2 places, the longest 1 mm (fastest: DP2.DP2_M2C_N at 100 ps 10-90 %, ASSUMED)

      Along 'DP2.DP2_M2C_N' on L6, G5 is a negative layer (copper everywhere inside the profile), and its field is cleared for 1 mm under the track at (-93.72, -10.45). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 2 places on 2 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -94.49..-93.37, y -10.8..-9.8. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 52% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'DP2.DP2_M2C_N' on L6 at (-93.72, -10.45), 1 mm; other side: (negative-layer field) on G4, 0.11 mm from the track against 0.1 mm to G5: 52% of the return on G5
      • 'DP2.DP2_M2C_P' on L6 at (-93.56, -10.61), 0.55 mm; other side: (negative-layer field) on G4, 0.11 mm from the track against 0.1 mm to G5: 52% of the return on G5

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AG5%3Ahole%20-94.49%2C-10.8%7C-93.37%2C-9.8 serves each with its own layer_image region; all of them lie in x -95.07..-92.37, y -11.8..-9.1 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:G5:hole -94.49,-5.93|-93.37,-4.932 nets cross one gap in G5's negative-plane field at 2 places, the longest 1.36 mm (fastest: DP2.DP3_M2C_N at 100 ps 10-90 %, ASSUMED)

      Along 'DP2.DP3_M2C_N' on L6, G5 is a negative layer (copper everywhere inside the profile), and its field is cleared for 1.36 mm under the track at (-94.01, -5.41). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 2 places on 2 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -94.49..-93.37, y -5.93..-4.93. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 52% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'DP2.DP3_M2C_N' on L6 at (-94.01, -5.41), 1.36 mm; other side: (negative-layer field) on G4, 0.11 mm from the track against 0.1 mm to G5: 52% of the return on G5
      • 'DP2.DP3_M2C_P' on L6 at (-93.79, -5.51), 1.19 mm; other side: (negative-layer field) on G4, 0.11 mm from the track against 0.1 mm to G5: 52% of the return on G5

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AG5%3Ahole%20-94.49%2C-5.93%7C-93.37%2C-4.93 serves each with its own layer_image region; all of them lie in x -95.49..-92.37, y -6.93..-3.93 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:G5:hole -94.49,56.48|-93.37,57.484 nets cross one gap in G5's negative-plane field at 4 places, the longest 0.65 mm (fastest: HA1.HA08_N at 100 ps 10-90 %, ASSUMED)

      Along 'HA1.HA08_P' on L6, G5 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.65 mm under the track at (-94.26, 56.71). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 4 places on 4 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -94.49..-93.37, y 56.48..57.48. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 52% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HA1.HA08_P' on L6 at (-94.26, 56.71), 0.65 mm; other side: (negative-layer field) on G4, 0.11 mm from the track against 0.1 mm to G5: 52% of the return on G5
      • 'HA1.HA13_P' on L6 at (-93.55, 57.29), 0.53 mm; other side: (negative-layer field) on G4, 0.11 mm from the track against 0.1 mm to G5: 52% of the return on G5
      • 'HA1.HA08_N' on L6 at (-94.41, 56.55), 0.2 mm; other side: (negative-layer field) on G4, 0.11 mm from the track against 0.1 mm to G5: 52% of the return on G5
      • 'HA1.HA13_N' on L6 at (-93.4, 57.45), 0.08 mm; other side: (negative-layer field) on G4, 0.11 mm from the track against 0.1 mm to G5: 52% of the return on G5

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AG5%3Ahole%20-94.49%2C56.48%7C-93.37%2C57.48 serves each with its own layer_image region; all of them lie in x -95.49..-92.37, y 55.48..58.48 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:G5:hole -94.49,61.21|-93.37,62.213 nets cross one gap in G5's negative-plane field at 3 places, the longest 1.13 mm (fastest: HA1.HA15_N at 100 ps 10-90 %, ASSUMED)

      Along 'HA1.HA15_N' on L6, G5 is a negative layer (copper everywhere inside the profile), and its field is cleared for 1.13 mm under the track at (-93.77, 61.81). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 3 places on 3 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -94.49..-93.37, y 61.21..62.21. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 52% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HA1.HA15_N' on L6 at (-93.77, 61.81), 1.13 mm; other side: (negative-layer field) on G4, 0.11 mm from the track against 0.1 mm to G5: 52% of the return on G5
      • 'HA1.HA15_P' on L6 at (-93.61, 61.97), 0.68 mm; other side: (negative-layer field) on G4, 0.11 mm from the track against 0.1 mm to G5: 52% of the return on G5
      • 'HA1.HA16_P' on L6 at (-94.39, 61.3), 0.26 mm; other side: (negative-layer field) on G4, 0.11 mm from the track against 0.1 mm to G5: 52% of the return on G5

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AG5%3Ahole%20-94.49%2C61.21%7C-93.37%2C62.21 serves each with its own layer_image region; all of them lie in x -95.49..-92.37, y 60.21..63.21 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:G5:hole -95.35,30.55|-94.23,31.55'VREF2.VREF_B_M2C' crosses one gap in G5's negative-plane field at 1 place, the longest 1.12 mm (fastest: VREF2.VREF_B_M2C at 100 ps 10-90 %, ASSUMED)

      Along 'VREF2.VREF_B_M2C' on L6, G5 is a negative layer (copper everywhere inside the profile), and its field is cleared for 1.12 mm under the track at (-94.79, 30.94). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 1 of them run ALONG the bare copper, parallel to its edges, rather than across it (the longest 1.12 mm). What cleared it: the hole at x -95.35..-94.23, y 30.55..31.55. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 52% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'VREF2.VREF_B_M2C' on L6 at (-94.79, 30.94), 1.12 mm, running ALONG the bare copper for 1.12 mm, not across it; other side: (negative-layer field) on G4, 0.11 mm from the track against 0.1 mm to G5: 52% of the return on G5

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AG5%3Ahole%20-95.35%2C30.55%7C-94.23%2C31.55 serves each with its own layer_image region; all of them lie in x -96.35..-93.23, y 29.94..31.94 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:G5:hole -96.03,39.39|-95.03,40.512 nets cross one gap in G5's negative-plane field at 2 places, the longest 1 mm (fastest: MGT.LINK01_CLK_N at 1 ns 10-90 %, suggested by its name)

      Along 'MGT.LINK01_CLK_N' on L6, G5 is a negative layer (copper everywhere inside the profile), and its field is cleared for 1 mm under the track at (-95.53, 40.11). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 2 places on 2 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -96.03..-95.03, y 39.39..40.51. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 52% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'MGT.LINK01_CLK_N' on L6 at (-95.53, 40.11), 1 mm; other side: (negative-layer field) on G4, 0.11 mm from the track against 0.1 mm to G5: 52% of the return on G5
      • 'MGT.LINK01_CLK_P' on L6 at (-95.53, 39.89), 1 mm; other side: (negative-layer field) on G4, 0.11 mm from the track against 0.1 mm to G5: 52% of the return on G5

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AG5%3Ahole%20-96.03%2C39.39%7C-95.03%2C40.51 serves each with its own layer_image region; all of them lie in x -97.03..-94.03, y 38.89..41.11 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:G5:hole -97.49,30.55|-96.37,31.55'VREF2.VREF_B_M2C' crosses one gap in G5's negative-plane field at 1 place, the longest 1.12 mm (fastest: VREF2.VREF_B_M2C at 100 ps 10-90 %, ASSUMED)

      Along 'VREF2.VREF_B_M2C' on L6, G5 is a negative layer (copper everywhere inside the profile), and its field is cleared for 1.12 mm under the track at (-96.93, 30.94). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 1 of them run ALONG the bare copper, parallel to its edges, rather than across it (the longest 1.12 mm). What cleared it: the hole at x -97.49..-96.37, y 30.55..31.55. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 52% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'VREF2.VREF_B_M2C' on L6 at (-96.93, 30.94), 1.12 mm, running ALONG the bare copper for 1.12 mm, not across it; other side: (negative-layer field) on G4, 0.11 mm from the track against 0.1 mm to G5: 52% of the return on G5

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AG5%3Ahole%20-97.49%2C30.55%7C-96.37%2C31.55 serves each with its own layer_image region; all of them lie in x -98.49..-95.37, y 29.94..31.94 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -105.72,88.21|-91.06,88.6112 nets cross one gap in P4's negative-plane field at 13 places, the longest 4.17 mm (fastest: HB1.HB10_N at 100 ps 10-90 %, ASSUMED)

      Along 'HB1.HB10_P' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 4.17 mm under the track at (-103.94, 88.45). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 3 of them run ALONG the bare copper, parallel to its edges, rather than across it (the longest 3.27 mm). 13 places on 12 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -105.72..-91.06, y 88.21..88.61. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HB1.HB10_P' on L3 at (-103.94, 88.45), 4.17 mm, running ALONG the bare copper for 3.27 mm, not across it; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB1.HB10_N' on L3 at (-103.12, 88.22), 1.36 mm, running ALONG the bare copper for 1.31 mm, not across it; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB1.HB14_N' on L3 at (-100.93, 88.41), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB1.HB14_P' on L3 at (-100.61, 88.41), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB1.HB18_N' on L3 at (-99.69, 88.41), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB1.HB18_P' on L3 at (-99.35, 88.41), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB1.HB20_N' on L3 at (-98.39, 88.45), 0.56 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB1.HB20_P' on L3 at (-98.12, 88.49), 0.55 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB1.HB10_N' on L3 at (-104.78, 88.22), 0.54 mm, running ALONG the bare copper for 0.43 mm, not across it; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FJTAG.TCK' on L3 at (-95.76, 88.41), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FJTAG.TDI' on L3 at (-95.19, 88.41), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FJTAG.TDO' on L3 at (-95.38, 88.41), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • and 1 more

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-105.72%2C88.21%7C-91.06%2C88.61 serves each with its own layer_image region; all of them lie in x -106.83..-94.19, y 87.21..89.61 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -106.86,21.93|-106.46,25.494 nets cross one gap in P4's negative-plane field at 4 places, the longest 0.4 mm (fastest: HB2.HB20_N at 100 ps 10-90 %, ASSUMED)

      Along 'HB2.HB20_P' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.4 mm under the track at (-106.66, 24.54). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 4 places on 4 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -106.86..-106.46, y 21.93..25.49. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HB2.HB20_N' on L3 at (-106.66, 24.77), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB20_P' on L3 at (-106.66, 24.54), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB21_N' on L3 at (-106.66, 23.5), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB21_P' on L3 at (-106.66, 23.27), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-106.86%2C21.93%7C-106.46%2C25.49 serves each with its own layer_image region; all of them lie in x -107.86..-105.46, y 22.27..25.77 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -106.86,21.93|-83.44,22.332 nets cross one gap in P4's negative-plane field at 2 places, the longest 23.58 mm (fastest: HB2.HB17_CC_N at 100 ps 10-90 %, ASSUMED)

      Along 'HB2.HB17_CC_N' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 23.58 mm under the track at (-94.98, 22.03). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 2 of them run ALONG the bare copper, parallel to its edges, rather than across it (the longest 23.03 mm). 2 places on 2 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -106.86..-83.44, y 21.93..22.33. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HB2.HB17_CC_N' on L3 at (-94.98, 22.03), 23.58 mm, running ALONG the bare copper for 23.03 mm, not across it; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB17_CC_P' on L3 at (-104.26, 22.25), 4.97 mm, running ALONG the bare copper for 4.63 mm, not across it; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-106.86%2C21.93%7C-83.44%2C22.33 serves each with its own layer_image region; all of them lie in x -107.86..-82.25, y 21.03..23.33 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -106.86,25.09|-106.46,28.994 nets cross one gap in P4's negative-plane field at 4 places, the longest 0.4 mm (fastest: HB2.HB14_N at 100 ps 10-90 %, ASSUMED)

      Along 'HB2.HB14_P' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.4 mm under the track at (-106.66, 27.31). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 4 places on 4 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -106.86..-106.46, y 25.09..28.99. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HB2.HB14_N' on L3 at (-106.66, 27.09), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB14_P' on L3 at (-106.66, 27.31), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB18_N' on L3 at (-106.66, 26.04), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB18_P' on L3 at (-106.66, 25.82), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-106.86%2C25.09%7C-106.46%2C28.99 serves each with its own layer_image region; all of them lie in x -107.86..-105.46, y 24.82..28.31 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -106.99,88.21|-105.32,89.88'HB1.HB14_N' crosses one gap in P4's negative-plane field at 1 place, the longest 0.57 mm (fastest: HB1.HB14_N at 100 ps 10-90 %, ASSUMED)

      Along 'HB1.HB14_N' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.57 mm under the track at (-106.54, 89.44). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. What cleared it: the hole at x -106.99..-105.32, y 88.21..89.88. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HB1.HB14_N' on L3 at (-106.54, 89.44), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-106.99%2C88.21%7C-105.32%2C89.88 serves each with its own layer_image region; all of them lie in x -107.82..-105.26, y 88.44..90.44 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -106.99,89.48|-106.59,98.056 nets cross one gap in P4's negative-plane field at 6 places, the longest 0.5 mm (fastest: HB1.HB14_P at 100 ps 10-90 %, ASSUMED)

      Along 'HB1.HB14_P' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.5 mm under the track at (-106.79, 89.66). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 6 places on 6 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -106.99..-106.59, y 89.48..98.05. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HB1.HB14_P' on L3 at (-106.79, 89.66), 0.5 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FJTAG.TCK' on L3 at (-106.79, 95.63), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB1.HB18_N' on L3 at (-106.79, 91.98), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB1.HB18_P' on L3 at (-106.79, 92.2), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB1.HB20_N' on L3 at (-106.79, 93.25), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB1.HB20_P' on L3 at (-106.79, 93.47), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-106.99%2C89.48%7C-106.59%2C98.05 serves each with its own layer_image region; all of them lie in x -107.99..-105.48, y 88.66..96.63 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -112.6,25.09|-106.46,25.492 nets cross one gap in P4's negative-plane field at 2 places, the longest 0.4 mm (fastest: HB2.HB18_N at 100 ps 10-90 %, ASSUMED)

      Along 'HB2.HB18_N' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.4 mm under the track at (-111.87, 25.29). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 2 places on 2 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -112.6..-106.46, y 25.09..25.49. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HB2.HB18_N' on L3 at (-111.87, 25.29), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB18_P' on L3 at (-111.65, 25.29), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-112.6%2C25.09%7C-106.46%2C25.49 serves each with its own layer_image region; all of them lie in x -112.87..-110.64, y 24.09..26.49 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -112.6,25.09|-112.2,34.164 nets cross one gap in P4's negative-plane field at 4 places, the longest 0.4 mm (fastest: HB2.HB14_N at 100 ps 10-90 %, ASSUMED)

      Along 'HB2.HB14_N' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.4 mm under the track at (-112.4, 27.09). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 4 places on 4 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -112.6..-112.2, y 25.09..34.16. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'GBT2.GBTCLK1_M2C_N' on L3 at (-112.4, 30.8), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'GBT2.GBTCLK1_M2C_P' on L3 at (-112.4, 30.57), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB14_N' on L3 at (-112.4, 27.09), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB14_P' on L3 at (-112.4, 27.31), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-112.6%2C25.09%7C-112.2%2C34.16 serves each with its own layer_image region; all of them lie in x -113.6..-111.2, y 26.09..31.8 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -112.6,33.76|-60.93,34.1622 nets cross one gap in P4's negative-plane field at 22 places, the longest 0.78 mm (fastest: HB2.HB00_CC_N at 100 ps 10-90 %, ASSUMED)

      Along 'FMCC2.CLK1_M2C_N' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.78 mm under the track at (-61.28, 33.96). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 22 places on 22 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -112.6..-60.93, y 33.76..34.16. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FMCC2.CLK1_M2C_N' on L3 at (-61.28, 33.96), 0.78 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FMCC2.CLK1_M2C_P' on L3 at (-61.59, 33.96), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB00_CC_N' on L3 at (-64.43, 33.96), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB00_CC_P' on L3 at (-64.75, 33.96), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB06_CC_N' on L3 at (-65.58, 33.96), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB06_CC_P' on L3 at (-65.9, 33.96), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB11_N' on L3 at (-69.12, 33.96), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB11_P' on L3 at (-69.44, 33.96), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB13_N' on L3 at (-66.73, 33.96), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB13_P' on L3 at (-67.05, 33.96), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB14_N' on L3 at (-76.9, 33.96), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB14_P' on L3 at (-77.22, 33.96), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • and 10 more

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-112.6%2C33.76%7C-60.93%2C34.16 serves each with its own layer_image region; all of them lie in x -91.81..-59.92, y 32.76..35.31 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -113.13,95.75|-106.59,96.152 nets cross one gap in P4's negative-plane field at 2 places, the longest 6.54 mm (fastest: FJTAG.TDO at 2 ns 10-90 %, suggested by its name)

      Along 'FJTAG.TDO' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 6.54 mm under the track at (-109.66, 96.01). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 2 of them run ALONG the bare copper, parallel to its edges, rather than across it (the longest 6.2 mm). 2 places on 2 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -113.13..-106.59, y 95.75..96.15. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FJTAG.TDO' on L3 at (-109.66, 96.01), 6.54 mm, running ALONG the bare copper for 6.19 mm, not across it; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FJTAG.TMS' on L3 at (-109.71, 95.82), 6.49 mm, running ALONG the bare copper for 6.2 mm, not across it; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-113.13%2C95.75%7C-106.59%2C96.15 serves each with its own layer_image region; all of them lie in x -114.13..-105.59, y 94.82..97.01 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -113.13,95.75|-112.73,104.176 nets cross one gap in P4's negative-plane field at 6 places, the longest 0.4 mm (fastest: RTM_PS# at 100 ps 10-90 %, ASSUMED)

      Along 'RTM_TDI' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.4 mm under the track at (-112.93, 97.24). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 6 places on 6 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -113.13..-112.73, y 95.75..104.17. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FJTAG.TDI' on L3 at (-112.93, 96.2), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'RTM_PS#' on L3 at (-112.93, 99.33), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'RTM_TCK' on L3 at (-112.93, 98.17), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'RTM_TDI' on L3 at (-112.93, 97.24), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'RTM_TDO' on L3 at (-112.93, 99.06), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'RTM_TMS' on L3 at (-112.93, 97.43), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-113.13%2C95.75%7C-112.73%2C104.17 serves each with its own layer_image region; all of them lie in x -114.13..-111.73, y 95.2..100.33 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -117.01,-16.83|-116.61,102.8519 nets cross one gap in P4's negative-plane field at 22 places, the longest 49.07 mm (fastest: LED_B at 100 ps 10-90 %, ASSUMED)

      Along 'FMCC2.CLK0_M2C_P' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 49.07 mm under the track at (-116.88, 8.38). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 3 of them run ALONG the bare copper, parallel to its edges, rather than across it (the longest 47.89 mm). 22 places on 19 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -117.01..-116.61, y -16.83..102.85. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FMCC2.CLK0_M2C_P' on L3 at (-116.88, 8.38), 49.07 mm, running ALONG the bare copper for 47.89 mm, not across it; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FMCC2.CLK0_M2C_N' on L3 at (-116.65, 10.32), 43.8 mm, running ALONG the bare copper for 43.58 mm, not across it; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FMCC2.CLK0_M2C_N' on L3 at (-116.65, -12.15), 1.09 mm, running ALONG the bare copper for 0.85 mm, not across it; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FJTAG.TCK' on L3 at (-116.81, 92.84), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FJTAG.TDI' on L3 at (-116.81, 93.99), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FJTAG.TDO' on L3 at (-116.81, 93.62), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FJTAG.TMS' on L3 at (-116.81, 93.23), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FMCC2.CLK2_BIDIR_N' on L3 at (-116.81, 36.37), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'LED_B' on L3 at (-116.81, 43.1), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'LED_G' on L3 at (-116.81, 42.4), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'LED_R' on L3 at (-116.81, 42.75), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'RTM_PS#' on L3 at (-116.81, 98.23), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • and 10 more

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-117.01%2C-16.83%7C-116.61%2C102.85 serves each with its own layer_image region; all of them lie in x -118.01..-115.61, y -14.26..99.43 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -117.01,-21.9|-111.54,-16.432 nets cross one gap in P4's negative-plane field at 2 places, the longest 0.57 mm (fastest: RTM_SYNC_CLK_N at 1 ns 10-90 %, suggested by its name)

      Along 'RTM_SYNC_CLK_N' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.57 mm under the track at (-112.3, -21.15). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 2 places on 2 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -117.01..-111.54, y -21.9..-16.43. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'RTM_SYNC_CLK_N' on L3 at (-112.3, -21.15), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'RTM_SYNC_CLK_P' on L3 at (-112.52, -20.92), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-117.01%2C-21.9%7C-111.54%2C-16.43 serves each with its own layer_image region; all of them lie in x -113.8..-111.01, y -22.15..-19.92 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -122.49,-22.04|-122.09,111.9516 nets cross one gap in P4's negative-plane field at 17 places, the longest 0.57 mm (fastest: LED_B at 100 ps 10-90 %, ASSUMED)

      Along 'FJTAG.TDI' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.57 mm under the track at (-122.29, 74.09). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 17 places on 16 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -122.49..-122.09, y -22.04..111.95. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'CLKS1.CLK_DIR' on L3 at (-122.29, 46.87), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'CLKS1.CLK_DIR' on L3 at (-122.29, 50.39), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FJTAG.TDI' on L3 at (-122.29, 74.09), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FJTAG.TDO' on L3 at (-122.29, 73.53), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FJTAG.TMS' on L3 at (-122.29, 72.9), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'I2C2.SCL' on L3 at (-122.29, 22.55), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'I2C2.SDA' on L3 at (-122.29, 22.93), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'RTM_TCK' on L3 at (-122.29, 92), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'RTM_TDO' on L3 at (-122.29, 92.37), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'LED_B' on L3 at (-122.29, 42.52), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'LED_G' on L3 at (-122.29, 42.02), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'LED_R' on L3 at (-122.29, 42.28), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • and 5 more

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-122.49%2C-22.04%7C-122.09%2C111.95 serves each with its own layer_image region; all of them lie in x -123.49..-121.09, y -22.15..97.45 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -24.28,82.44|-23.88,111.982 nets cross one gap in P4's negative-plane field at 2 places, the longest 0.41 mm (fastest: FPGA_CLK2_N at 1 ns 10-90 %, suggested by its name)

      Along 'FPGA_CLK2_P' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.41 mm under the track at (-24.06, 93.44). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 2 places on 2 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -24.28..-23.88, y 82.44..111.98. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FPGA_CLK2_P' on L3 at (-24.06, 93.44), 0.41 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FPGA_CLK2_N' on L3 at (-23.99, 93.68), 0.41 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-24.28%2C82.44%7C-23.88%2C111.98 serves each with its own layer_image region; all of them lie in x -25.28..-22.88, y 92.38..94.69 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -38.36,68.36|-23.88,82.844 nets cross one gap in P4's negative-plane field at 4 places, the longest 0.57 mm (fastest: IO0_33 at 100 ps 10-90 %, ASSUMED)

      Along 'FPGA_CLK2_N' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.57 mm under the track at (-26.49, 80.23). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 4 places on 4 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -38.36..-23.88, y 68.36..82.84. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FPGA_CLK2_N' on L3 at (-26.49, 80.23), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FPGA_CLK2_P' on L3 at (-26.27, 80.45), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'IO25_33' on L3 at (-24.73, 81.99), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'IO0_33' on L3 at (-24.92, 81.74), 0.53 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-38.36%2C68.36%7C-23.88%2C82.84 serves each with its own layer_image region; all of them lie in x -27.49..-23.73, y 78.94..83.27 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -44.52,-15.2|-38.9,-9.58'FPGA_ALERTN' crosses one gap in P4's negative-plane field at 1 place, the longest 0.57 mm (fastest: FPGA_ALERTN at 20 ns 10-90 %, suggested by its name)

      Along 'FPGA_ALERTN' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.57 mm under the track at (-41.16, -11.84). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. What cleared it: the hole at x -44.52..-38.9, y -15.2..-9.58. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FPGA_ALERTN' on L3 at (-41.16, -11.84), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-44.52%2C-15.2%7C-38.9%2C-9.58 serves each with its own layer_image region; all of them lie in x -42.16..-40.16, y -13.12..-10.56 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -48.3,6.66|-45.75,9.21'FPGA_ALERTN' crosses one gap in P4's negative-plane field at 1 place, the longest 0.57 mm (fastest: FPGA_ALERTN at 20 ns 10-90 %, suggested by its name)

      Along 'FPGA_ALERTN' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.57 mm under the track at (-47.59, 8.5). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. What cleared it: the hole at x -48.3..-45.75, y 6.66..9.21. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FPGA_ALERTN' on L3 at (-47.59, 8.5), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-48.3%2C6.66%7C-45.75%2C9.21 serves each with its own layer_image region; all of them lie in x -48.59..-46.59, y 7.21..9.78 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -48.3,61.35|-47.9,68.768 nets cross one gap in P4's negative-plane field at 8 places, the longest 0.57 mm (fastest: IO0_33 at 100 ps 10-90 %, ASSUMED)

      Along 'VADJ1_IO4' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.57 mm under the track at (-48.1, 64.92). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 8 places on 8 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -48.3..-47.9, y 61.35..68.76. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'VADJ1_IO4' on L3 at (-48.1, 64.92), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'IO25_33' on L3 at (-48.09, 65.66), 0.41 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FDDR.RST_N' on L3 at (-48.1, 68.05), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'IO0_33' on L3 at (-48.1, 66.16), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'IO13_33' on L3 at (-48.1, 67.17), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'PVADJ2_IO10' on L3 at (-48.1, 66.37), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'VADJ1_IO0' on L3 at (-48.1, 64.15), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'VADJ1_IO1' on L3 at (-48.1, 64.34), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-48.3%2C61.35%7C-47.9%2C68.76 serves each with its own layer_image region; all of them lie in x -49.3..-46.9, y 63.15..69.05 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -48.3,68.36|-37.96,68.76'FDDR.RST_N' crosses one gap in P4's negative-plane field at 1 place, the longest 0.56 mm (fastest: FDDR.RST_N at 20 ns 10-90 %, suggested by its name)

      Along 'FDDR.RST_N' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.56 mm under the track at (-41.33, 68.5). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. What cleared it: the hole at x -48.3..-37.96, y 68.36..68.76. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FDDR.RST_N' on L3 at (-41.33, 68.5), 0.56 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-48.3%2C68.36%7C-37.96%2C68.76 serves each with its own layer_image region; all of them lie in x -42.47..-40.08, y 67.36..69.76 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -48.3,8.81|-47.9,61.7512 nets cross one gap in P4's negative-plane field at 12 places, the longest 0.57 mm (fastest: RTM_LVDS.RTM_LVD2_N at 100 ps 10-90 %, ASSUMED)

      Along 'RTM_LVDS.RTM_LVD3_P' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.57 mm under the track at (-48.1, 51.23). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 12 places on 12 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -48.3..-47.9, y 8.81..61.75. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'I2C_CPU.SCL' on L3 at (-48.1, 33.85), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'I2C_CPU.SDA' on L3 at (-48.1, 33.35), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'RTM_LVDS.RTM_LVD3_N' on L3 at (-48.1, 50.91), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'RTM_LVDS.RTM_LVD3_P' on L3 at (-48.1, 51.23), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'RTM_SYNC_CLK_N' on L3 at (-48.1, 43.17), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'RTM_SYNC_CLK_P' on L3 at (-48.1, 43.49), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'RXD_USBB' on L3 at (-48.1, 34.31), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'PVADJ2_IO7' on L3 at (-48.1, 55.33), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'RTM_LVDS.RTM_LVD2_N' on L3 at (-48.1, 54.62), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'RTM_LVDS.RTM_LVD2_P' on L3 at (-48.1, 54.85), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'RTM_LVDS.RTM_LVD4_N' on L3 at (-48.1, 52.69), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'RTM_LVDS.RTM_LVD4_P' on L3 at (-48.1, 52.92), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-48.3%2C8.81%7C-47.9%2C61.75 serves each with its own layer_image region; all of them lie in x -49.3..-46.9, y 32.15..56.33 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -53,0.84|-38.88,1.24'FPGA_ALERTN' crosses one gap in P4's negative-plane field at 1 place, the longest 0.4 mm (fastest: FPGA_ALERTN at 20 ns 10-90 %, suggested by its name)

      Along 'FPGA_ALERTN' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.4 mm under the track at (-45.47, 1.04). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. What cleared it: the hole at x -53..-38.88, y 0.84..1.24. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FPGA_ALERTN' on L3 at (-45.47, 1.04), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-53%2C0.84%7C-38.88%2C1.24 serves each with its own layer_image region; all of them lie in x -46.47..-44.47, y -0.16..2.24 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -55.01,-3.82|-38.81,-3.42'FPGA_ALERTN' crosses one gap in P4's negative-plane field at 1 place, the longest 0.53 mm (fastest: FPGA_ALERTN at 20 ns 10-90 %, suggested by its name)

      Along 'FPGA_ALERTN' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.53 mm under the track at (-45.31, -3.66). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. What cleared it: the hole at x -55.01..-38.81, y -3.82..-3.42. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FPGA_ALERTN' on L3 at (-45.31, -3.66), 0.53 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-55.01%2C-3.82%7C-38.81%2C-3.42 serves each with its own layer_image region; all of them lie in x -46.47..-44.15, y -4.82..-2.42 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -55.48,20.82|-55.08,44.767 nets cross one gap in P4's negative-plane field at 7 places, the longest 0.57 mm (fastest: RXD_USBB at 500 ps 10-90 %, suggested by its name)

      Along 'FMCC2.CLK1_M2C_N' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.57 mm under the track at (-55.28, 39.95). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 7 places on 7 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -55.48..-55.08, y 20.82..44.76. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FMCC2.CLK1_M2C_N' on L3 at (-55.28, 39.95), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FMCC2.CLK1_M2C_P' on L3 at (-55.28, 40.27), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'I2C_CPU.SCL' on L3 at (-55.28, 26.67), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'I2C_CPU.SDA' on L3 at (-55.28, 26.17), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'RTM_SYNC_CLK_N' on L3 at (-55.28, 35.99), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'RTM_SYNC_CLK_P' on L3 at (-55.28, 36.31), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'RXD_USBB' on L3 at (-55.28, 27.13), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-55.48%2C20.82%7C-55.08%2C44.76 serves each with its own layer_image region; all of them lie in x -56.48..-54.08, y 24.97..41.47 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -59.3,44.36|-55.08,48.586 nets cross one gap in P4's negative-plane field at 6 places, the longest 0.57 mm (fastest: HB2.HB00_CC_N at 100 ps 10-90 %, ASSUMED)

      Along 'HB2.HB00_CC_N' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.57 mm under the track at (-55.56, 44.84). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 6 places on 6 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -59.3..-55.08, y 44.36..48.58. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HB2.HB00_CC_N' on L3 at (-55.56, 44.84), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB00_CC_P' on L3 at (-55.78, 45.06), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB06_CC_N' on L3 at (-56.57, 45.85), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB06_CC_P' on L3 at (-56.79, 46.07), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB13_N' on L3 at (-57.56, 46.84), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB13_P' on L3 at (-57.78, 47.06), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-59.3%2C44.36%7C-55.08%2C48.58 serves each with its own layer_image region; all of them lie in x -58.78..-54.56, y 43.56..48.35 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -61.32,-1.2|-60.92,36.593 nets cross one gap in P4's negative-plane field at 3 places, the longest 0.57 mm (fastest: FMCC2.CLK1_M2C_P at 1 ns 10-90 %, suggested by its name)

      Along 'FMCC2.CLK1_M2C_P' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.57 mm under the track at (-61.12, 34.43). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 3 places on 3 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -61.32..-60.92, y -1.2..36.59. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FMCC2.CLK1_M2C_P' on L3 at (-61.12, 34.43), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'RTM_SYNC_CLK_N' on L3 at (-61.12, 30.15), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'RTM_SYNC_CLK_P' on L3 at (-61.12, 30.47), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-61.32%2C-1.2%7C-60.92%2C36.59 serves each with its own layer_image region; all of them lie in x -62.32..-59.92, y 28.95..35.63 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -64.03,36.19|-60.92,39.36 nets cross one gap in P4's negative-plane field at 6 places, the longest 0.4 mm (fastest: HB2.HB00_CC_N at 100 ps 10-90 %, ASSUMED)

      Along 'HB2.HB06_CC_P' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.4 mm under the track at (-63.08, 38.35). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 6 places on 6 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -64.03..-60.92, y 36.19..39.3. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HB2.HB00_CC_N' on L3 at (-61.56, 36.83), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB00_CC_P' on L3 at (-61.72, 36.99), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB06_CC_N' on L3 at (-62.92, 38.19), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB06_CC_P' on L3 at (-63.08, 38.35), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB13_N' on L3 at (-63.5, 38.77), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB13_P' on L3 at (-63.66, 38.93), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-64.03%2C36.19%7C-60.92%2C39.3 serves each with its own layer_image region; all of them lie in x -64.8..-60.42, y 35.69..40.07 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -67.38,61.35|-47.9,61.7518 nets cross one gap in P4's negative-plane field at 18 places, the longest 0.42 mm (fastest: HB2.HB00_CC_N at 100 ps 10-90 %, ASSUMED)

      Along 'HB2.HB13_N' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.42 mm under the track at (-57.56, 61.42). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 18 places on 18 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -67.38..-47.9, y 61.35..61.75. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HB2.HB13_N' on L3 at (-57.56, 61.42), 0.42 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB13_P' on L3 at (-57.78, 61.42), 0.42 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB11_N' on L3 at (-59.56, 61.39), 0.42 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB11_P' on L3 at (-59.78, 61.39), 0.42 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB20_P' on L3 at (-62.55, 61.43), 0.42 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB00_CC_N' on L3 at (-55.56, 61.55), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB00_CC_P' on L3 at (-55.78, 61.55), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB06_CC_N' on L3 at (-56.57, 61.55), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB06_CC_P' on L3 at (-56.79, 61.55), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB14_N' on L3 at (-64.55, 61.55), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB14_P' on L3 at (-64.78, 61.55), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB17_CC_N' on L3 at (-60.56, 61.55), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • and 6 more

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-67.38%2C61.35%7C-47.9%2C61.75 serves each with its own layer_image region; all of them lie in x -65.78..-54.56, y 60.35..62.75 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -68.31,48.18|-58.9,48.5816 nets cross one gap in P4's negative-plane field at 16 places, the longest 0.4 mm (fastest: HB2.HB11_N at 100 ps 10-90 %, ASSUMED)

      Along 'FMCC2.CLK2_BIDIR_P' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.4 mm under the track at (-66.79, 48.38). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 16 places on 16 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -68.31..-58.9, y 48.18..48.58. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FMCC2.CLK0_M2C_N' on L3 at (-65.59, 48.38), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FMCC2.CLK0_M2C_P' on L3 at (-65.82, 48.38), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FMCC2.CLK2_BIDIR_N' on L3 at (-66.56, 48.38), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FMCC2.CLK2_BIDIR_P' on L3 at (-66.79, 48.38), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB11_N' on L3 at (-59.56, 48.38), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB11_P' on L3 at (-59.78, 48.38), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB14_N' on L3 at (-64.55, 48.38), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB14_P' on L3 at (-64.78, 48.38), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB17_CC_N' on L3 at (-60.56, 48.38), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB17_CC_P' on L3 at (-60.79, 48.38), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB18_N' on L3 at (-63.79, 48.38), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB18_P' on L3 at (-63.56, 48.38), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • and 4 more

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-68.31%2C48.18%7C-58.9%2C48.58 serves each with its own layer_image region; all of them lie in x -67.79..-58.56, y 47.18..49.58 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -68.88,24.98|-63.81,30.054 nets cross one gap in P4's negative-plane field at 4 places, the longest 0.4 mm (fastest: FMCC2.CLK1_M2C_N at 1 ns 10-90 %, suggested by its name)

      Along 'RTM_SYNC_CLK_N' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.4 mm under the track at (-65.05, 26.22). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 4 places on 4 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -68.88..-63.81, y 24.98..30.05. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FMCC2.CLK1_M2C_N' on L3 at (-67.03, 28.2), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FMCC2.CLK1_M2C_P' on L3 at (-67.19, 28.36), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'RTM_SYNC_CLK_N' on L3 at (-65.05, 26.22), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'RTM_SYNC_CLK_P' on L3 at (-65.21, 26.38), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-68.88%2C24.98%7C-63.81%2C30.05 serves each with its own layer_image region; all of them lie in x -68.33..-63.91, y 25.08..29.5 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -69.84,49.71|-69.44,56.75'FSPI.FLASH_Q/D1' crosses one gap in P4's negative-plane field at 1 place, the longest 5.9 mm (fastest: FSPI.FLASH_Q/D1 at 2 ns 10-90 %, suggested by its name)

      Along 'FSPI.FLASH_Q/D1' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 5.9 mm under the track at (-69.5, 50.79). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 1 of them run ALONG the bare copper, parallel to its edges, rather than across it (the longest 4.29 mm). What cleared it: the hole at x -69.84..-69.44, y 49.71..56.75. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FSPI.FLASH_Q/D1' on L3 at (-69.5, 50.79), 5.9 mm, running ALONG the bare copper for 4.29 mm, not across it; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-69.84%2C49.71%7C-69.44%2C56.75 serves each with its own layer_image region; all of them lie in x -70.5..-68.44, y 48.48..56.36 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -79.11,6.35|-78.71,17.64 nets cross one gap in P4's negative-plane field at 4 places, the longest 0.57 mm (fastest: FMCC2.CLK1_M2C_N at 1 ns 10-90 %, suggested by its name)

      Along 'FMCC2.CLK1_M2C_P' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.57 mm under the track at (-78.91, 16.64). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 4 places on 4 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -79.11..-78.71, y 6.35..17.6. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FMCC2.CLK1_M2C_N' on L3 at (-78.91, 16.32), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FMCC2.CLK1_M2C_P' on L3 at (-78.91, 16.64), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'RTM_SYNC_CLK_N' on L3 at (-78.91, 12.36), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'RTM_SYNC_CLK_P' on L3 at (-78.91, 12.68), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-79.11%2C6.35%7C-78.71%2C17.6 serves each with its own layer_image region; all of them lie in x -80.11..-77.71, y 11.16..17.84 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -81.31,38.9|-63.63,39.312 nets cross one gap in P4's negative-plane field at 12 places, the longest 0.57 mm (fastest: HB2.HB11_N at 100 ps 10-90 %, ASSUMED)

      Along 'HB2.HB14_N' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.57 mm under the track at (-71.76, 39.1). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 12 places on 12 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -81.31..-63.63, y 38.9..39.3. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HB2.HB11_N' on L3 at (-64.32, 39.1), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB11_P' on L3 at (-64.64, 39.1), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB14_N' on L3 at (-71.76, 39.1), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB14_P' on L3 at (-72.08, 39.1), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB17_CC_N' on L3 at (-66.64, 39.1), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB17_CC_P' on L3 at (-66.96, 39.1), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB21_N' on L3 at (-68.11, 39.1), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB21_P' on L3 at (-67.79, 39.1), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB20_P' on L3 at (-69.23, 39.16), 0.43 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB20_N' on L3 at (-69.49, 39.1), 0.42 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB18_P' on L3 at (-70.16, 38.99), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB18_N' on L3 at (-70.38, 39.12), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-81.31%2C38.9%7C-63.63%2C39.3 serves each with its own layer_image region; all of them lie in x -73.28..-63.12, y 37.9..40.3 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -83.84,17.2|-78.71,22.338 nets cross one gap in P4's negative-plane field at 8 places, the longest 0.44 mm (fastest: HB2.HB00_CC_N at 100 ps 10-90 %, ASSUMED)

      Along 'HB2.HB11_P' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.44 mm under the track at (-82.63, 21.03). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 8 places on 8 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -83.84..-78.71, y 17.2..22.33. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HB2.HB11_P' on L3 at (-82.63, 21.03), 0.44 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB11_N' on L3 at (-82.27, 20.93), 0.43 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB13_P' on L3 at (-81.35, 19.77), 0.43 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB13_N' on L3 at (-81.17, 19.61), 0.42 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB00_CC_N' on L3 at (-79.95, 18.44), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB00_CC_P' on L3 at (-80.11, 18.6), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB06_CC_N' on L3 at (-80.53, 19.02), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB06_CC_P' on L3 at (-80.69, 19.18), 0.4 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-83.84%2C17.2%7C-78.71%2C22.33 serves each with its own layer_image region; all of them lie in x -83.78..-78.81, y 17.3..22.14 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -90.88,38.9|-80.91,48.8714 nets cross one gap in P4's negative-plane field at 14 places, the longest 0.57 mm (fastest: FLASH_D2 at 100 ps 10-90 %, ASSUMED)

      Along 'FSPI.DONE' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.57 mm under the track at (-88.24, 46.23). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 14 places on 14 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -90.88..-80.91, y 38.9..48.87. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'CLKS1.CLK3_BIDIR_P' on L3 at (-90.41, 48.4), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FLASH_D2' on L3 at (-87.56, 45.55), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FLASH_D3' on L3 at (-87.22, 45.21), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FMCC2.CLK0_M2C_N' on L3 at (-81.89, 39.88), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FMCC2.CLK0_M2C_P' on L3 at (-82.12, 40.11), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FMCC2.CLK2_BIDIR_N' on L3 at (-84.03, 42.02), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FMCC2.CLK2_BIDIR_P' on L3 at (-84.26, 42.25), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FSPI.DONE' on L3 at (-88.24, 46.23), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FSPI.FCS_B\' on L3 at (-87.9, 45.89), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FSPI.FLASH_Q/D1' on L3 at (-86.54, 44.53), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FSPI.FLASH_SI/D0' on L3 at (-86.88, 44.87), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'FSPI.FPGA_CCLK' on L3 at (-88.92, 46.91), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • and 2 more

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-90.88%2C38.9%7C-80.91%2C48.87 serves each with its own layer_image region; all of them lie in x -91.69..-80.61, y 38.88..49.4 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -90.88,48.47|-90.48,88.0343 nets cross one gap in P4's negative-plane field at 43 places, the longest 0.57 mm (fastest: HB1.HB04_N at 100 ps 10-90 %, ASSUMED)

      Along 'HB1.HB04_N' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.57 mm under the track at (-90.68, 68.99). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 43 places on 43 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -90.88..-90.48, y 48.47..88.03. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HB1.HB04_N' on L3 at (-90.68, 68.99), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB1.HB04_P' on L3 at (-90.68, 69.31), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'LA1.LA25_N' on L3 at (-90.68, 65.54), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'LA1.LA25_P' on L3 at (-90.68, 65.86), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'LA1.LA26_N' on L3 at (-90.68, 66.69), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'LA1.LA26_P' on L3 at (-90.68, 67.01), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'LA1.LA27_N' on L3 at (-90.68, 67.84), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'LA1.LA27_P' on L3 at (-90.68, 68.16), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB1.HB20_P' on L3 at (-90.7, 73.93), 0.56 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'CLKS1.CLK3_BIDIR_N' on L3 at (-90.68, 48.62), 0.52 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB1.HB10_P' on L3 at (-90.55, 70.39), 0.51 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB1.HB18_N' on L3 at (-90.76, 72.54), 0.49 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • and 31 more

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-90.88%2C48.47%7C-90.48%2C88.03 serves each with its own layer_image region; all of them lie in x -91.88..-89.35, y 47.62..78.59 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P4:hole -98.42,29.65|-68.48,30.0518 nets cross one gap in P4's negative-plane field at 18 places, the longest 0.63 mm (fastest: HB2.HB00_CC_N at 100 ps 10-90 %, ASSUMED)

      Along 'HB2.HB21_P' on L3, P4 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.63 mm under the track at (-75.53, 29.85). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 18 places on 18 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -98.42..-68.48, y 29.65..30.05. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HB2.HB21_P' on L3 at (-75.53, 29.85), 0.63 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB06_CC_N' on L3 at (-69.69, 29.85), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB06_CC_P' on L3 at (-70.01, 29.85), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB11_N' on L3 at (-73.23, 29.85), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB11_P' on L3 at (-73.55, 29.85), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB13_N' on L3 at (-70.84, 29.85), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB13_P' on L3 at (-71.16, 29.85), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB14_N' on L3 at (-81.01, 29.85), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB14_P' on L3 at (-81.33, 29.85), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB17_CC_N' on L3 at (-74.38, 29.85), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB17_CC_P' on L3 at (-74.7, 29.85), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • 'HB2.HB18_N' on L3 at (-78.19, 29.85), 0.57 mm; other side: (negative-layer field) on G3, 0.1 mm from the track against 0.11 mm to P4: 48% of the return on P4
      • and 6 more

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP4%3Ahole%20-98.42%2C29.65%7C-68.48%2C30.05 serves each with its own layer_image region; all of them lie in x -82.53..-67.47, y 28.6..31.05 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -116.12,44.03|-115.74,61.34 nets cross one gap in P6's negative-plane field at 4 places, the longest 0.54 mm (fastest: MISC1.PG_C2M at 100 ps 10-90 %, ASSUMED)

      Along 'FSPI.FPGA_CCLK' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.54 mm under the track at (-115.93, 44.66). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 4 places on 4 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -116.12..-115.74, y 44.03..61.3. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FSPI.FPGA_CCLK' on L5 at (-115.93, 44.66), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'MISC1.PG_C2M' on L5 at (-115.93, 47.35), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'MISC1.PG_M2C' on L5 at (-115.93, 47.54), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'MISC1.PRSNT' on L5 at (-115.93, 47.73), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-116.12%2C44.03%7C-115.74%2C61.3 serves each with its own layer_image region; all of them lie in x -117.12..-114.74, y 43.47..48.73 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -118.71,63.51|-118.33,96.15'FSPI.INIT_B' crosses one gap in P6's negative-plane field at 1 place, the longest 0.38 mm (fastest: FSPI.INIT_B at 2 ns 10-90 %, suggested by its name)

      Along 'FSPI.INIT_B' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.38 mm under the track at (-118.52, 93.41). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. What cleared it: the hole at x -118.71..-118.33, y 63.51..96.15. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FSPI.INIT_B' on L5 at (-118.52, 93.41), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-118.71%2C63.51%7C-118.33%2C96.15 serves each with its own layer_image region; all of them lie in x -119.71..-117.33, y 92.41..94.41 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -119.85,24.22|-119.47,40.6810 nets cross one gap in P6's negative-plane field at 10 places, the longest 0.38 mm (fastest: PLL_CTRL.PLL_VADJ2_DAC1_SYNC_N at 100 ps 10-90 %, ASSUMED)

      Along 'PLL_CTRL.VADJ2_SI57X_SCL' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.38 mm under the track at (-119.66, 38.23). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 10 places on 10 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -119.85..-119.47, y 24.22..40.68. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FLASH_CSN' on L5 at (-119.66, 40), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'PLL_CTRL.PLL_VADJ2_DAC1_SYNC_N' on L5 at (-119.66, 37.09), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'PLL_CTRL.PLL_VADJ2_DAC2_SYNC_N' on L5 at (-119.66, 36.71), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'PLL_CTRL.PLL_VADJ2_DAC_DIN' on L5 at (-119.66, 37.28), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'PLL_CTRL.PLL_VADJ2_DAC_SCLK' on L5 at (-119.66, 36.9), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'PLL_CTRL.VADJ2_CLK_UPDATEN' on L5 at (-119.66, 37.85), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'PLL_CTRL.VADJ2_SI57X_OE' on L5 at (-119.66, 37.47), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'PLL_CTRL.VADJ2_SI57X_SCL' on L5 at (-119.66, 38.23), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'PLL_CTRL.VADJ2_SI57X_SDA' on L5 at (-119.66, 38.04), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'PLL_CTRL.VADJ2_SI57X_TUNE' on L5 at (-119.66, 37.66), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-119.85%2C24.22%7C-119.47%2C40.68 serves each with its own layer_image region; all of them lie in x -120.85..-118.47, y 35.71..41 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -119.85,40.3|-115.74,44.418 nets cross one gap in P6's negative-plane field at 8 places, the longest 1.85 mm (fastest: FLASH_D2 at 100 ps 10-90 %, ASSUMED)

      Along 'FSPI.PROGRAM_B' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 1.85 mm under the track at (-116.3, 43.81). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 1 of them run ALONG the bare copper, parallel to its edges, rather than across it (the longest 1.29 mm). 8 places on 8 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -119.85..-115.74, y 40.3..44.41. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FSPI.PROGRAM_B' on L5 at (-116.3, 43.81), 1.85 mm, running ALONG the bare copper for 1.29 mm, not across it; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'FLASH_D2' on L5 at (-118.45, 41.7), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'FLASH_D3' on L5 at (-119.13, 41.02), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'FSPI.DONE' on L5 at (-117.77, 42.38), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'FSPI.FCS_B\' on L5 at (-118.11, 42.04), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'FSPI.FLASH_Q/D1' on L5 at (-119.47, 40.68), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'FSPI.FLASH_SI/D0' on L5 at (-118.79, 41.36), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'FLASH_CLK' on L5 at (-119.66, 40.34), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-119.85%2C40.3%7C-115.74%2C44.41 serves each with its own layer_image region; all of them lie in x -120.85..-114.74, y 39.34..45.37 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -121.89,25.95|-121.49,55.2922 nets cross one gap in P6's negative-plane field at 22 places, the longest 0.57 mm (fastest: FLASH_D2 at 100 ps 10-90 %, ASSUMED)

      Along 'PLL_CTRL.PLL_VADJ2_DAC2_SYNC_N' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.57 mm under the track at (-121.69, 35.72). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 22 places on 22 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -121.89..-121.49, y 25.95..55.29. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'PLL_CTRL.PLL_VADJ2_DAC1_SYNC_N' on L5 at (-121.69, 36.26), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'PLL_CTRL.PLL_VADJ2_DAC2_SYNC_N' on L5 at (-121.69, 35.72), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'PLL_CTRL.PLL_VADJ2_DAC_DIN' on L5 at (-121.69, 36.53), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'PLL_CTRL.PLL_VADJ2_DAC_SCLK' on L5 at (-121.69, 35.99), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'PLL_CTRL.VADJ2_CLK_UPDATEN' on L5 at (-121.69, 37.33), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'PLL_CTRL.VADJ2_SI57X_OE' on L5 at (-121.69, 36.79), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'PLL_CTRL.VADJ2_SI57X_SCL' on L5 at (-121.69, 37.87), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'PLL_CTRL.VADJ2_SI57X_SDA' on L5 at (-121.69, 37.6), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'PLL_CTRL.VADJ2_SI57X_TUNE' on L5 at (-121.69, 37.06), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'MISC1.PRSNT' on L5 at (-121.69, 46.16), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'MISC1.PG_M2C' on L5 at (-121.65, 45.93), 0.53 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'MISC1.PG_C2M' on L5 at (-121.61, 45.7), 0.5 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • and 10 more

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-121.89%2C25.95%7C-121.49%2C55.29 serves each with its own layer_image region; all of them lie in x -122.89..-120.49, y 34.52..47.36 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -124.52,93.21|-123.05,94.68'FSPI.INIT_B' crosses one gap in P6's negative-plane field at 1 place, the longest 0.57 mm (fastest: FSPI.INIT_B at 2 ns 10-90 %, suggested by its name)

      Along 'FSPI.INIT_B' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.57 mm under the track at (-123.69, 93.85). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. What cleared it: the hole at x -124.52..-123.05, y 93.21..94.68. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FSPI.INIT_B' on L5 at (-123.69, 93.85), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-124.52%2C93.21%7C-123.05%2C94.68 serves each with its own layer_image region; all of them lie in x -124.97..-122.41, y 92.85..94.85 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -127.05,30.68|-126.65,53.6219 nets cross one gap in P6's negative-plane field at 19 places, the longest 0.57 mm (fastest: FLASH_D2 at 100 ps 10-90 %, ASSUMED)

      Along 'FLASH_D2' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.57 mm under the track at (-126.85, 42.44). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 19 places on 19 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -127.05..-126.65, y 30.68..53.62. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FLASH_D2' on L5 at (-126.85, 42.44), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'FLASH_D3' on L5 at (-126.85, 41.48), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'FSPI.DONE' on L5 at (-126.85, 43.4), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'FSPI.FCS_B\' on L5 at (-126.85, 42.92), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'FSPI.FLASH_Q/D1' on L5 at (-126.85, 41), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'FSPI.FLASH_SI/D0' on L5 at (-126.85, 41.96), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'FSPI.FPGA_CCLK' on L5 at (-126.85, 44.37), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'FSPI.PROGRAM_B' on L5 at (-126.85, 43.89), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'PLL_CTRL.PLL_VADJ2_DAC_DIN' on L5 at (-126.85, 31.37), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'PLL_CTRL.VADJ2_CLK_UPDATEN' on L5 at (-126.85, 32.17), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'PLL_CTRL.VADJ2_SI57X_OE' on L5 at (-126.85, 31.64), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'PLL_CTRL.VADJ2_SI57X_SCL' on L5 at (-126.85, 32.71), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • and 7 more

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-127.05%2C30.68%7C-126.65%2C53.62 serves each with its own layer_image region; all of them lie in x -128.05..-125.65, y 30.17..47.06 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -130.55,-16.5|-130.15,27.582 nets cross one gap in P6's negative-plane field at 3 places, the longest 0.57 mm (fastest: PLL_CTRL.PLL_VADJ2_DAC2_SYNC_N at 100 ps 10-90 %, ASSUMED)

      Along 'VTTI' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.57 mm under the track at (-130.35, -10.6). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 3 places on 2 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -130.55..-130.15, y -16.5..27.58. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'PLL_CTRL.PLL_VADJ2_DAC2_SYNC_N' on L5 at (-130.35, 27.06), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'VTTI' on L5 at (-130.35, -10.6), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'VTTI' on L5 at (-130.35, 0.22), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-130.55%2C-16.5%7C-130.15%2C27.58 serves each with its own layer_image region; all of them lie in x -131.55..-129.15, y -11.8..28.26 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -130.55,-29.24|-117.42,-16.12 nets cross one gap in P6's negative-plane field at 2 places, the longest 0.4 mm (fastest: UPDATEN at 100 ps 10-90 %, ASSUMED)

      Along 'UPDATEN' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.4 mm under the track at (-128.02, -18.63). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 2 places on 2 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -130.55..-117.42, y -29.24..-16.1. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'CLK_UPDATEN' on L5 at (-128.18, -18.48), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'UPDATEN' on L5 at (-128.02, -18.63), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-130.55%2C-29.24%7C-117.42%2C-16.1 serves each with its own layer_image region; all of them lie in x -129.32..-126.88, y -19.77..-17.33 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -130.55,27.18|-126.65,31.082 nets cross one gap in P6's negative-plane field at 2 places, the longest 5.52 mm (fastest: PLL_CTRL.PLL_VADJ2_DAC1_SYNC_N at 100 ps 10-90 %, ASSUMED)

      Along 'PLL_CTRL.PLL_VADJ2_DAC_SCLK' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 5.52 mm under the track at (-128.57, 29.11). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 2 of them run ALONG the bare copper, parallel to its edges, rather than across it (the longest 4.98 mm). 2 places on 2 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -130.55..-126.65, y 27.18..31.08. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'PLL_CTRL.PLL_VADJ2_DAC_SCLK' on L5 at (-128.57, 29.11), 5.52 mm, running ALONG the bare copper for 4.89 mm, not across it; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'PLL_CTRL.PLL_VADJ2_DAC1_SYNC_N' on L5 at (-128.71, 29.24), 5.52 mm, running ALONG the bare copper for 4.98 mm, not across it; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-130.55%2C27.18%7C-126.65%2C31.08 serves each with its own layer_image region; all of them lie in x -131.55..-125.65, y 26.13..32.3 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -131.53,57.7|-131.13,92.322 nets cross one gap in P6's negative-plane field at 3 places, the longest 11.01 mm (fastest: MISC1.PG_M2C at 100 ps 10-90 %, ASSUMED)

      Along 'MISC1.PG_M2C' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 11.01 mm under the track at (-131.45, 82.81). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 1 of them run ALONG the bare copper, parallel to its edges, rather than across it (the longest 10.78 mm). 3 places on 2 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -131.53..-131.13, y 57.7..92.32. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'MISC1.PG_M2C' on L5 at (-131.45, 82.81), 11.01 mm, running ALONG the bare copper for 10.78 mm, not across it; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'MISC1.PRSNT' on L5 at (-131.33, 74.33), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'MISC1.PRSNT' on L5 at (-131.33, 89.98), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-131.53%2C57.7%7C-131.13%2C92.32 serves each with its own layer_image region; all of them lie in x -132.53..-130.13, y 73.13..91.18 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -131.53,91.92|-128.77,94.68'FSPI.INIT_B' crosses one gap in P6's negative-plane field at 1 place, the longest 0.57 mm (fastest: FSPI.INIT_B at 2 ns 10-90 %, suggested by its name)

      Along 'FSPI.INIT_B' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.57 mm under the track at (-129.6, 93.85). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. What cleared it: the hole at x -131.53..-128.77, y 91.92..94.68. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FSPI.INIT_B' on L5 at (-129.6, 93.85), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-131.53%2C91.92%7C-128.77%2C94.68 serves each with its own layer_image region; all of them lie in x -130.88..-128.32, y 92.85..94.85 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -17.49,16.22|-17.09,44.446 nets cross one gap in P6's negative-plane field at 6 places, the longest 0.4 mm (fastest: FCLKA_N at 1 ns 10-90 %, suggested by its name)

      Along 'TCLKA_N' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.4 mm under the track at (-17.29, 24.95). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 6 places on 6 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -17.49..-17.09, y 16.22..44.44. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FCLKA_N' on L5 at (-17.29, 29.78), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'FCLKA_P' on L5 at (-17.29, 29.55), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKA_N' on L5 at (-17.29, 24.95), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKA_P' on L5 at (-17.29, 24.73), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKB_N' on L5 at (-17.29, 27.2), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKB_P' on L5 at (-17.29, 26.98), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-17.49%2C16.22%7C-17.09%2C44.44 serves each with its own layer_image region; all of them lie in x -18.49..-16.09, y 23.73..30.78 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -21.41,12.3|-17.09,16.622 nets cross one gap in P6's negative-plane field at 2 places, the longest 0.57 mm (fastest: TCLKC_N at 1 ns 10-90 %, suggested by its name)

      Along 'TCLKC_N' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.57 mm under the track at (-20.55, 13.16). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 2 places on 2 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -21.41..-17.09, y 12.3..16.62. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'TCLKC_N' on L5 at (-20.55, 13.16), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKC_P' on L5 at (-20.78, 12.94), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-21.41%2C12.3%7C-17.09%2C16.62 serves each with its own layer_image region; all of them lie in x -21.78..-19.55, y 11.65..14.44 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -24.17,5.4|-21.12,8.452 nets cross one gap in P6's negative-plane field at 2 places, the longest 0.54 mm (fastest: TCLKD_N at 1 ns 10-90 %, suggested by its name)

      Along 'TCLKD_P' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.54 mm under the track at (-22.15, 7.42). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 2 places on 2 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -24.17..-21.12, y 5.4..8.45. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'TCLKD_N' on L5 at (-21.93, 7.64), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKD_P' on L5 at (-22.15, 7.42), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-24.17%2C5.4%7C-21.12%2C8.45 serves each with its own layer_image region; all of them lie in x -23.15..-20.93, y 6.15..8.91 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -24.83,81.32|-24.45,97.92 nets cross one gap in P6's negative-plane field at 2 places, the longest 0.54 mm (fastest: I2C_CPU.SCL at 10 ns 10-90 %, suggested by its name)

      Along 'I2C_CPU.SDA' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.54 mm under the track at (-24.64, 89.57). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 2 places on 2 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -24.83..-24.45, y 81.32..97.9. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'I2C_CPU.SCL' on L5 at (-24.64, 88.94), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'I2C_CPU.SDA' on L5 at (-24.64, 89.57), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-24.83%2C81.32%7C-24.45%2C97.9 serves each with its own layer_image region; all of them lie in x -25.83..-23.45, y 87.75..90.76 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -31,15.53|-30.62,23.824 nets cross one gap in P6's negative-plane field at 4 places, the longest 0.38 mm (fastest: TCLKC_N at 1 ns 10-90 %, suggested by its name)

      Along 'TCLKC_N' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.38 mm under the track at (-30.94, 23.45). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 4 places on 4 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -31..-30.62, y 15.53..23.82. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'TCLKC_N' on L5 at (-30.94, 23.45), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKC_P' on L5 at (-30.81, 23.23), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKD_N' on L5 at (-30.81, 21.7), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKD_P' on L5 at (-30.81, 21.48), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-31%2C15.53%7C-30.62%2C23.82 serves each with its own layer_image region; all of them lie in x -32..-29.62, y 20.48..24.45 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -42.28,12.36|-41.9,26.126 nets cross one gap in P6's negative-plane field at 6 places, the longest 0.38 mm (fastest: TCLKA_N at 1 ns 10-90 %, suggested by its name)

      Along 'TCLKC_N' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.38 mm under the track at (-42.09, 23.45). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 6 places on 6 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -42.28..-41.9, y 12.36..26.12. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'TCLKA_N' on L5 at (-42.09, 24.95), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKA_P' on L5 at (-42.09, 24.73), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKC_N' on L5 at (-42.09, 23.45), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKC_P' on L5 at (-42.09, 23.23), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKD_N' on L5 at (-42.09, 21.7), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKD_P' on L5 at (-42.09, 21.48), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-42.28%2C12.36%7C-41.9%2C26.12 serves each with its own layer_image region; all of them lie in x -43.28..-40.9, y 20.48..25.95 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -42.8,23.44|-30.62,35.626 nets cross one gap in P6's negative-plane field at 6 places, the longest 0.54 mm (fastest: FCLKA_N at 1 ns 10-90 %, suggested by its name)

      Along 'TCLKA_N' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.54 mm under the track at (-32.13, 24.95). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 6 places on 6 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -42.8..-30.62, y 23.44..35.62. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FCLKA_N' on L5 at (-36.96, 29.78), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'FCLKA_P' on L5 at (-36.73, 29.55), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKA_N' on L5 at (-32.13, 24.95), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKA_P' on L5 at (-31.91, 24.73), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKB_N' on L5 at (-35.96, 28.78), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKB_P' on L5 at (-35.73, 28.55), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-42.8%2C23.44%7C-30.62%2C35.62 serves each with its own layer_image region; all of them lie in x -38.22..-30.64, y 23.73..30.78 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -42.81,35.23|-42.41,72.752 nets cross one gap in P6's negative-plane field at 2 places, the longest 0.4 mm (fastest: PVADJ2_IO8 at 2 ns 10-90 %, suggested by its name)

      Along 'PVADJ2_IO8' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.4 mm under the track at (-42.61, 67.59). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 2 places on 2 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -42.81..-42.41, y 35.23..72.75. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'PVADJ2_IO8' on L5 at (-42.61, 67.59), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'VADJ1_IO5' on L5 at (-42.61, 67.78), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-42.81%2C35.23%7C-42.41%2C72.75 serves each with its own layer_image region; all of them lie in x -43.81..-41.41, y 66.59..68.78 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -50.58,40.15|-50.18,64.766 nets cross one gap in P6's negative-plane field at 7 places, the longest 14.64 mm (fastest: HB2.HB01_N at 100 ps 10-90 %, ASSUMED)

      Along 'HB2.HB01_N' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 14.64 mm under the track at (-50.38, 58.39). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 4 of them run ALONG the bare copper, parallel to its edges, rather than across it (the longest 13.62 mm). 7 places on 6 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -50.58..-50.18, y 40.15..64.76. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HB2.HB01_N' on L5 at (-50.38, 58.39), 14.64 mm, running ALONG the bare copper for 13.62 mm, not across it; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB01_P' on L5 at (-50.27, 51.45), 5.42 mm, running ALONG the bare copper for 4.64 mm, not across it; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB02_N' on L5 at (-50.24, 43.18), 2.69 mm, running ALONG the bare copper for 1.93 mm, not across it; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB02_P' on L5 at (-50.47, 43.18), 2.63 mm, running ALONG the bare copper for 1.94 mm, not across it; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB01_P' on L5 at (-50.49, 63.66), 0.7 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB03_P' on L5 at (-50.48, 45.59), 0.45 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB03_N' on L5 at (-50.4, 45.39), 0.44 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-50.58%2C40.15%7C-50.18%2C64.76 serves each with its own layer_image region; all of them lie in x -51.58..-49.18, y 40.52..64.75 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -64.85,40.15|-50.18,40.5516 nets cross one gap in P6's negative-plane field at 16 places, the longest 0.57 mm (fastest: HB2.HB01_N at 100 ps 10-90 %, ASSUMED)

      Along 'HB2.HB05_N' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.57 mm under the track at (-57.61, 40.35). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 16 places on 16 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -64.85..-50.18, y 40.15..40.55. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HB2.HB01_N' on L5 at (-56.46, 40.35), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB02_N' on L5 at (-52.88, 40.35), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB02_P' on L5 at (-53.2, 40.35), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB03_N' on L5 at (-54.99, 40.35), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB03_P' on L5 at (-55.31, 40.35), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB05_N' on L5 at (-57.61, 40.35), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB05_P' on L5 at (-57.29, 40.35), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB08_N' on L5 at (-62.74, 40.35), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB08_P' on L5 at (-62.42, 40.35), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB09_N' on L5 at (-61.16, 40.3), 0.47 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB09_P' on L5 at (-60.91, 40.29), 0.44 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB07_P' on L5 at (-60.16, 40.44), 0.41 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • and 4 more

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-64.85%2C40.15%7C-50.18%2C40.55 serves each with its own layer_image region; all of them lie in x -63.94..-51.68, y 39.15..41.55 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -64.85,69.35|-45.18,69.752 nets cross one gap in P6's negative-plane field at 2 places, the longest 9.72 mm (fastest: FDDR.A7 at 100 ps 10-90 %, ASSUMED)

      Along 'FDDR.A7' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 9.72 mm under the track at (-50, 69.43). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 1 of them run ALONG the bare copper, parallel to its edges, rather than across it (the longest 8.43 mm). 2 places on 2 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -64.85..-45.18, y 69.35..69.75. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FDDR.A7' on L5 at (-50, 69.43), 9.72 mm, running ALONG the bare copper for 8.43 mm, not across it; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'FDDR.BA2' on L5 at (-57.56, 69.59), 0.44 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-64.85%2C69.35%7C-45.18%2C69.75 serves each with its own layer_image region; all of them lie in x -58.56..-44.18, y 68.35..70.75 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -65.88,40.15|-64.86,41.172 nets cross one gap in P6's negative-plane field at 2 places, the longest 0.4 mm (fastest: LA2.LA15_N at 100 ps 10-90 %, ASSUMED)

      Along 'LA2.LA15_N' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.4 mm under the track at (-65.26, 40.55). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 2 places on 2 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -65.88..-64.86, y 40.15..41.17. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'LA2.LA15_N' on L5 at (-65.26, 40.55), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'LA2.LA15_P' on L5 at (-65.42, 40.71), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-65.88%2C40.15%7C-64.86%2C41.17 serves each with its own layer_image region; all of them lie in x -66.56..-64.12, y 39.41..41.85 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -69.87,66.32|-47.1,66.72'HB2.HB08_P' crosses one gap in P6's negative-plane field at 1 place, the longest 0.41 mm (fastest: HB2.HB08_P at 100 ps 10-90 %, ASSUMED)

      Along 'HB2.HB08_P' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.41 mm under the track at (-56.56, 66.41). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. What cleared it: the hole at x -69.87..-47.1, y 66.32..66.72. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HB2.HB08_P' on L5 at (-56.56, 66.41), 0.41 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-69.87%2C66.32%7C-47.1%2C66.72 serves each with its own layer_image region; all of them lie in x -57.56..-55.49, y 65.32..67.72 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -69.88,64.36|-50.18,64.7614 nets cross one gap in P6's negative-plane field at 15 places, the longest 1.46 mm (fastest: HB2.HB01_N at 100 ps 10-90 %, ASSUMED)

      Along 'HB2.HB01_N' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 1.46 mm under the track at (-52.1, 64.45). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 1 of them run ALONG the bare copper, parallel to its edges, rather than across it (the longest 0.91 mm). 15 places on 14 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -69.88..-50.18, y 64.36..64.76. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HB2.HB01_N' on L5 at (-52.1, 64.45), 1.46 mm, running ALONG the bare copper for 0.91 mm, not across it; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB12_N' on L5 at (-58.53, 64.69), 0.51 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB08_N' on L5 at (-56.78, 64.56), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB08_P' on L5 at (-56.56, 64.56), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB15_N' on L5 at (-62.79, 64.56), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB16_N' on L5 at (-61.57, 64.56), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB16_P' on L5 at (-61.8, 64.56), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB19_N' on L5 at (-60.79, 64.56), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB19_P' on L5 at (-60.56, 64.56), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'PLL_CTRL.PLL_VADJ2_DAC1_SYNC_N' on L5 at (-67.48, 64.56), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'PLL_CTRL.PLL_VADJ2_DAC2_SYNC_N' on L5 at (-67.29, 64.56), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'PVADJ2_IO8' on L5 at (-63.67, 64.56), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • and 3 more

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-69.88%2C64.36%7C-50.18%2C64.76 serves each with its own layer_image region; all of them lie in x -68.67..-50.54, y 63.36..65.76 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -71.7,9.47|-71.32,20.048 nets cross one gap in P6's negative-plane field at 8 places, the longest 0.54 mm (fastest: FCLKA_N at 1 ns 10-90 %, suggested by its name)

      Along 'TCLKC_P' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.54 mm under the track at (-71.51, 10.8). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 8 places on 8 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -71.7..-71.32, y 9.47..20.04. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FCLKA_N' on L5 at (-71.51, 18.13), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'FCLKA_P' on L5 at (-71.51, 17.82), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKA_N' on L5 at (-71.51, 14.47), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKA_P' on L5 at (-71.51, 14.15), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKB_N' on L5 at (-71.51, 16.31), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKB_P' on L5 at (-71.51, 15.99), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKC_N' on L5 at (-71.51, 11.12), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKC_P' on L5 at (-71.51, 10.8), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-71.7%2C9.47%7C-71.32%2C20.04 serves each with its own layer_image region; all of them lie in x -72.7..-70.32, y 9.61..19.32 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -71.71,5.48|-71.31,9.862 nets cross one gap in P6's negative-plane field at 2 places, the longest 0.57 mm (fastest: TCLKD_N at 1 ns 10-90 %, suggested by its name)

      Along 'TCLKD_P' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.57 mm under the track at (-71.51, 8.67). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 2 places on 2 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -71.71..-71.31, y 5.48..9.86. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'TCLKD_N' on L5 at (-71.51, 8.99), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKD_P' on L5 at (-71.51, 8.67), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-71.71%2C5.48%7C-71.31%2C9.86 serves each with its own layer_image region; all of them lie in x -72.71..-70.31, y 7.47..10.19 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -72.96,35.23|-42.41,35.6321 nets cross one gap in P6's negative-plane field at 21 places, the longest 0.57 mm (fastest: HB2.HB01_N at 100 ps 10-90 %, ASSUMED)

      Along 'HB2.HB05_N' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.57 mm under the track at (-62.53, 35.43). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 21 places on 21 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -72.96..-42.41, y 35.23..35.63. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HB2.HB01_N' on L5 at (-61.38, 35.43), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB01_P' on L5 at (-61.06, 35.43), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB02_N' on L5 at (-57.8, 35.43), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB02_P' on L5 at (-58.12, 35.43), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB03_N' on L5 at (-59.91, 35.43), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB03_P' on L5 at (-60.23, 35.43), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB04_N' on L5 at (-63.37, 35.43), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB04_P' on L5 at (-63.69, 35.43), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB05_N' on L5 at (-62.53, 35.43), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB05_P' on L5 at (-62.21, 35.43), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB07_N' on L5 at (-64.52, 35.43), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB07_P' on L5 at (-64.84, 35.43), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • and 9 more

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-72.96%2C35.23%7C-42.41%2C35.63 serves each with its own layer_image region; all of them lie in x -73.52..-56.6, y 34.23..36.63 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -73.18,28.45|-44.6,28.8318 nets cross one gap in P6's negative-plane field at 18 places, the longest 14.75 mm (fastest: HB2.HB01_N at 100 ps 10-90 %, ASSUMED)

      Along 'TCLKB_N' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 14.75 mm under the track at (-51.51, 28.78). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 2 of them run ALONG the bare copper, parallel to its edges, rather than across it (the longest 13.88 mm). 18 places on 18 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -73.18..-44.6, y 28.45..28.83. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'TCLKB_N' on L5 at (-51.51, 28.78), 14.75 mm, running ALONG the bare copper for 13.88 mm, not across it; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKB_P' on L5 at (-51.49, 28.55), 14.45 mm, running ALONG the bare copper for 13.83 mm, not across it; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'FCLKA_N' on L5 at (-61, 28.64), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'FCLKA_P' on L5 at (-60.69, 28.64), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB01_N' on L5 at (-68.17, 28.64), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB01_P' on L5 at (-67.85, 28.64), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB02_N' on L5 at (-64.59, 28.64), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB02_P' on L5 at (-64.91, 28.64), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB03_N' on L5 at (-66.7, 28.64), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB03_P' on L5 at (-67.02, 28.64), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB04_N' on L5 at (-70.16, 28.64), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB04_P' on L5 at (-70.48, 28.64), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • and 6 more

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-73.18%2C28.45%7C-44.6%2C28.83 serves each with its own layer_image region; all of them lie in x -73.97..-43.44, y 27.45..29.83 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -73.37,19.66|-71.32,21.72 nets cross one gap in P6's negative-plane field at 2 places, the longest 0.38 mm (fastest: HB2.HB02_N at 100 ps 10-90 %, ASSUMED)

      Along 'HB2.HB02_P' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.38 mm under the track at (-72.61, 20.94). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 2 places on 2 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -73.37..-71.32, y 19.66..21.7. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HB2.HB02_N' on L5 at (-72.45, 20.79), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB02_P' on L5 at (-72.61, 20.94), 0.38 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-73.37%2C19.66%7C-71.32%2C21.7 serves each with its own layer_image region; all of them lie in x -73.74..-71.31, y 19.65..22.08 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -73.37,21.32|-72.99,28.7712 nets cross one gap in P6's negative-plane field at 12 places, the longest 1.08 mm (fastest: HB2.HB01_N at 100 ps 10-90 %, ASSUMED)

      Along 'HB2.HB05_P' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 1.08 mm under the track at (-73.15, 24.37). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 1 of them run ALONG the bare copper, parallel to its edges, rather than across it (the longest 0.38 mm). 12 places on 12 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -73.37..-72.99, y 21.32..28.77. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HB2.HB05_P' on L5 at (-73.15, 24.37), 1.08 mm, running ALONG the bare copper for 0.38 mm, not across it; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB05_N' on L5 at (-73.06, 24.9), 0.67 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB01_N' on L5 at (-73.18, 22.96), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB01_P' on L5 at (-73.18, 22.64), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB03_P' on L5 at (-73.18, 21.81), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB04_N' on L5 at (-73.18, 25.63), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB04_P' on L5 at (-73.18, 25.94), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB07_N' on L5 at (-73.18, 26.78), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB07_P' on L5 at (-73.18, 27.09), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB09_N' on L5 at (-73.18, 28.24), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB09_P' on L5 at (-73.18, 27.92), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB03_N' on L5 at (-73.14, 21.53), 0.45 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-73.37%2C21.32%7C-72.99%2C28.77 serves each with its own layer_image region; all of them lie in x -74.37..-71.99, y 20.37..29.43 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -75.79,35.26|-72.56,35.663 nets cross one gap in P6's negative-plane field at 3 places, the longest 0.57 mm (fastest: HB2.HB10_N at 100 ps 10-90 %, ASSUMED)

      Along 'HB2.HB10_N' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.57 mm under the track at (-74.3, 35.46). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 3 places on 3 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -75.79..-72.56, y 35.26..35.66. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HB2.HB10_N' on L5 at (-74.3, 35.46), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB10_P' on L5 at (-74.62, 35.46), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB12_N' on L5 at (-72.7, 35.6), 0.43 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-75.79%2C35.26%7C-72.56%2C35.66 serves each with its own layer_image region; all of them lie in x -75.82..-71.69, y 34.23..36.66 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -75.9,60.3|-75.5,67.86 nets cross one gap in P6's negative-plane field at 6 places, the longest 0.54 mm (fastest: HB1.HB02_N at 100 ps 10-90 %, ASSUMED)

      Along 'FPGA_I2C_SDA' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.54 mm under the track at (-75.73, 62.48). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 6 places on 6 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -75.9..-75.5, y 60.3..67.8. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FPGA_I2C_SDA' on L5 at (-75.73, 62.48), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB1.HB02_N' on L5 at (-75.7, 64.67), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB1.HB02_P' on L5 at (-75.7, 64.45), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB1.HB03_N' on L5 at (-75.7, 63.67), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB1.HB03_P' on L5 at (-75.7, 63.45), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'VADJ1_IO5' on L5 at (-75.7, 61.53), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-75.9%2C60.3%7C-75.5%2C67.8 serves each with its own layer_image region; all of them lie in x -76.9..-74.5, y 60.53..65.67 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -75.9,72.35|-64.45,72.754 nets cross one gap in P6's negative-plane field at 4 places, the longest 0.4 mm (fastest: HB1.HB16_N at 100 ps 10-90 %, ASSUMED)

      Along 'FPGA_I2C_SCL' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.4 mm under the track at (-71.53, 72.55). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 4 places on 4 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -75.9..-64.45, y 72.35..72.75. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FPGA_I2C_SCL' on L5 at (-71.53, 72.55), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'FPGA_I2C_SDA' on L5 at (-71.77, 72.55), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB1.HB16_N' on L5 at (-72.79, 72.55), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB1.HB16_P' on L5 at (-72.56, 72.55), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-75.9%2C72.35%7C-64.45%2C72.75 serves each with its own layer_image region; all of them lie in x -73.79..-70.53, y 71.35..73.75 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -76.35,35.26|-75.39,36.222 nets cross one gap in P6's negative-plane field at 2 places, the longest 0.56 mm (fastest: HB2.HB19_N at 100 ps 10-90 %, ASSUMED)

      Along 'HB2.HB19_P' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.56 mm under the track at (-75.52, 35.39). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 2 places on 2 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -76.35..-75.39, y 35.26..36.22. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HB2.HB19_P' on L5 at (-75.52, 35.39), 0.56 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB19_N' on L5 at (-75.68, 35.55), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-76.35%2C35.26%7C-75.39%2C36.22 serves each with its own layer_image region; all of them lie in x -76.82..-74.25, y 34.27..36.69 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -76.35,35.82|-75.95,41.174 nets cross one gap in P6's negative-plane field at 4 places, the longest 0.57 mm (fastest: HB2.HB15_N at 100 ps 10-90 %, ASSUMED)

      Along 'HB2.HB15_N' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.57 mm under the track at (-76.15, 38.12). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 4 places on 4 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -76.35..-75.95, y 35.82..41.17. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HB2.HB15_N' on L5 at (-76.15, 38.12), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB15_P' on L5 at (-76.15, 37.8), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB16_N' on L5 at (-76.15, 36.65), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB16_P' on L5 at (-76.15, 36.97), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-76.35%2C35.82%7C-75.95%2C41.17 serves each with its own layer_image region; all of them lie in x -77.35..-74.95, y 35.45..39.32 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -76.35,40.77|-65.48,41.1712 nets cross one gap in P6's negative-plane field at 12 places, the longest 0.57 mm (fastest: HB2.HB10_N at 100 ps 10-90 %, ASSUMED)

      Along 'HB2.HB15_P' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.57 mm under the track at (-72.98, 40.97). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 12 places on 12 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -76.35..-65.48, y 40.77..41.17. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HB2.HB10_N' on L5 at (-69.53, 40.97), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB10_P' on L5 at (-69.85, 40.97), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB12_N' on L5 at (-67.46, 40.97), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB12_P' on L5 at (-67.14, 40.97), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB15_N' on L5 at (-73.3, 40.97), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB15_P' on L5 at (-72.98, 40.97), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB16_N' on L5 at (-71.83, 40.97), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB16_P' on L5 at (-72.15, 40.97), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB19_N' on L5 at (-71, 40.97), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB19_P' on L5 at (-70.68, 40.97), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'LA2.LA16_N' on L5 at (-65.99, 40.97), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'LA2.LA16_P' on L5 at (-66.31, 40.97), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-76.35%2C40.77%7C-65.48%2C41.17 serves each with its own layer_image region; all of them lie in x -74.5..-64.79, y 39.77..42.17 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -76.9,0.98|-76.52,4.312 nets cross one gap in P6's negative-plane field at 2 places, the longest 0.54 mm (fastest: TCLKD_N at 1 ns 10-90 %, suggested by its name)

      Along 'TCLKD_N' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.54 mm under the track at (-76.71, 3.79). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 2 places on 2 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -76.9..-76.52, y 0.98..4.31. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'TCLKD_N' on L5 at (-76.71, 3.79), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKD_P' on L5 at (-76.71, 3.47), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-76.9%2C0.98%7C-76.52%2C4.31 serves each with its own layer_image region; all of them lie in x -77.9..-75.52, y 2.28..4.98 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -77.23,28.38|-72.99,28.773 nets cross one gap in P6's negative-plane field at 3 places, the longest 0.54 mm (fastest: HB2.HB08_N at 100 ps 10-90 %, ASSUMED)

      Along 'HB2.HB08_N' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.54 mm under the track at (-75.93, 28.58). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 3 places on 3 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -77.23..-72.99, y 28.38..28.77. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HB2.HB08_N' on L5 at (-75.93, 28.58), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB08_P' on L5 at (-75.62, 28.58), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'LA2.LA15_N' on L5 at (-76.73, 28.64), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-77.23%2C28.38%7C-72.99%2C28.77 serves each with its own layer_image region; all of them lie in x -77.99..-74.42, y 27.38..29.77 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -77.76,58.44|-75.51,60.694 nets cross one gap in P6's negative-plane field at 4 places, the longest 0.54 mm (fastest: HA1.HA06_N at 100 ps 10-90 %, ASSUMED)

      Along 'HA1.HA18_N' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.54 mm under the track at (-76.76, 59.44). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 4 places on 4 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -77.76..-75.51, y 58.44..60.69. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HA1.HA18_N' on L5 at (-76.76, 59.44), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HA1.HA18_P' on L5 at (-76.54, 59.66), 0.54 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HA1.HA06_N' on L5 at (-75.78, 60.43), 0.52 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HA1.HA06_P' on L5 at (-75.7, 60.66), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-77.76%2C58.44%7C-75.51%2C60.69 serves each with its own layer_image region; all of them lie in x -78.03..-74.5, y 58.44..61.66 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -78.33,4.03|-74.55,7.814 nets cross one gap in P6's negative-plane field at 4 places, the longest 0.4 mm (fastest: TCLKC_N at 1 ns 10-90 %, suggested by its name)

      Along 'TCLKC_P' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.4 mm under the track at (-76.41, 5.89). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 4 places on 4 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -78.33..-74.55, y 4.03..7.81. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'TCLKC_N' on L5 at (-76.57, 6.05), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKC_P' on L5 at (-76.41, 5.89), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKD_N' on L5 at (-75.51, 4.99), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKD_P' on L5 at (-75.35, 4.83), 0.4 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-78.33%2C4.03%7C-74.55%2C7.81 serves each with its own layer_image region; all of them lie in x -77.71..-74.21, y 3.69..7.19 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -78.33,7.41|-77.93,23.118 nets cross one gap in P6's negative-plane field at 18 places, the longest 0.57 mm (fastest: HB2.HB01_N at 100 ps 10-90 %, ASSUMED)

      Along 'FCLKA_N' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.57 mm under the track at (-78.13, 11.51). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 18 places on 18 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -78.33..-77.93, y 7.41..23.1. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'FCLKA_N' on L5 at (-78.13, 11.51), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'FCLKA_P' on L5 at (-78.13, 11.2), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB01_N' on L5 at (-78.13, 18.01), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB01_P' on L5 at (-78.13, 17.69), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB02_N' on L5 at (-78.13, 15.1), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB02_P' on L5 at (-78.13, 15.42), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB03_N' on L5 at (-78.13, 16.54), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB03_P' on L5 at (-78.13, 16.86), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB04_N' on L5 at (-78.13, 20.67), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB04_P' on L5 at (-78.13, 20.99), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB05_N' on L5 at (-78.13, 19.22), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'HB2.HB05_P' on L5 at (-78.13, 18.9), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • and 6 more

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-78.33%2C7.41%7C-77.93%2C23.1 serves each with its own layer_image region; all of them lie in x -79.33..-76.93, y 6.43..23.34 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -78.39,58.44|-77.37,58.82'HA1.HA10_P' crosses one gap in P6's negative-plane field at 1 place, the longest 1.37 mm (fastest: HA1.HA10_P at 100 ps 10-90 %, ASSUMED)

      Along 'HA1.HA10_P' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 1.37 mm under the track at (-77.69, 58.66). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 1 of them run ALONG the bare copper, parallel to its edges, rather than across it (the longest 0.8 mm). What cleared it: the hole at x -78.39..-77.37, y 58.44..58.82. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'HA1.HA10_P' on L5 at (-77.69, 58.66), 1.37 mm, running ALONG the bare copper for 0.8 mm, not across it; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-78.39%2C58.44%7C-77.37%2C58.82 serves each with its own layer_image region; all of them lie in x -79.44..-76.07, y 57.66..59.66 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    • gap:P6:hole -78.57,-6.31|-9.34,-5.914 nets cross one gap in P6's negative-plane field at 4 places, the longest 0.9 mm (fastest: TCLKC_N at 1 ns 10-90 %, suggested by its name)

      Along 'TCLKC_N' on L5, P6 is a negative layer (copper everywhere inside the profile), and its field is cleared for 0.9 mm under the track at (-9.84, -6.26). The layer does not state the nets either side, so whether this is a split between two planes or a void in one is not said; either way the return current has to go round it. 4 places on 4 net(s) cross it, listed below, the longest first. What cleared it: the hole at x -78.57..-9.34, y -6.31..-5.91. One decision covers the void: whether these tracks may cross it, or the plane should be filled under them. Every crossing has a plane on the track's other side at most twice as far away as the broken one, so at most 48% of the return meets the break and the rest follows the track on that plane (to first order each plane beside a stripline carries a share of the return inversely proportional to its distance from the track). The finding is the lower tier (info).

      • 'TCLKC_N' on L5 at (-9.84, -6.26), 0.9 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKD_N' on L5 at (-12.04, -6.11), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKD_P' on L5 at (-12.36, -6.11), 0.57 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6
      • 'TCLKC_P' on L5 at (-10.4, -6.22), 0.37 mm; other side: (negative-layer field) on G4, 0.1 mm from the track against 0.11 mm to P6: 48% of the return on P6

      Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/34/reference-planes?group=gap%3AP6%3Ahole%20-78.57%2C-6.31%7C-9.34%2C-5.91 serves each with its own layer_image region; all of them lie in x -13.56..-8.62, y -7.31..-4.91 mm (board frame, Y up).

      Read from the copper as drawn on the reference layer, along each track's centreline at exact breakpoints. The reference is the copper beside the track, never the layer's type or name. Whether it matters depends on each signal's edge rate, which nothing on file states for these nets: a slow static line crosses a split harmlessly, a fast edge does not.

    All 142 in the full review

Main parts

Part numberMakerDescriptionWhere
1571983-4 or GDH04S04TYCO4 Ways SPST DIP SwitchSW1
39-28-1043MOLEXConnector 4 MaleJ4
570BBC000121DGSILICON LABSAny-Rate I2C Programmable XO Oscillator Si570 SerieOSC4
74HC4053DIC73
87832-1420MOLEXConnector 14 MaleJ6
8SLVD1204-33NLGI2:4, LVDS Output Fanout Buffer, 3.3VIC75
AD8402ARZ11k dual digital potentiometerIC11
ADN4604ASVZ16x16 CML crosspoint switchIC8
AT24MAC602-XHM-TSerial 2K EEPROM+ 48bit MACIC27
ATS-50310G-C2-R0heat sink BGA 31x31(54.9)mmHS1
CDCM61004RHBTTEXAS INSTRUMENTSFour Output, Integrated VCO, Low-Jitter Clock GeneratorIC50
goldpin 1.27 2x5PBTD10SJP2
HDT0001C&KPush-Button Switch SPDTPB1
HSME-C120AVAGO TECHNOLOGIESLED Green Avago C120LED2
HSMR-C120AVAGO TECHNOLOGIESLED Blue Avago C120LED1
LD1117DT33CTRST MICROELECTRONICS3-Terminal Regulator 1=GND 2=IN 3=OUTIC74
LF VCXO026156IQD FREQUENCY PRODUCTS3.3V ±100ppm 15pF HCMOS Tri-State Surface Mount Voltage Controlled Oscillator VCXO (CFPV-45 Serie)OSC2
LP5951MF-3.0/NOPBNATIONAL SEMICONDUCTOR150mA Micropower Low-Dropout CMOS Voltage RegulatorIC51
LPC1764FBD100ARM Cortex-M3 procesorIC1
MAX6642ATT90±1°C, SMBus- Remote/Local Temperature Sensor with Overtemperature AlarmIC18
MCP79410-I/SNRTC with 1k EEPROM and RAMIC9
OVSRRGBCC3OPTEK / TT ELECTRONICSRight Angle Tricolor (RGB) Surface Mount LED, Body 3.2x1mmLD1
PCA9517DGKLEVEL-TRANSLATING I2C BUS REPEATERIC42
PCA9547BS8-channel I2C-bus multiplexer with resetIC80
PLTD10SGoldpin 2.54mm, male headerJP1
PT7M7811 or STM811SWvoltage supervisorIC58
Q0M032D6-SMDCrystal oscillator 32k (Maritex)X3
QSDG8M000-TXCminiature crystal oscillator (Maritex) QSDG8M000-TXCX1
RFSM250 or MF-SM250-2-992.50A/15VF3
SCANSTA111MTEnhanced SCAN Bridge Multidrop Addressable IEEE 1149.1 (JTAG) PortIC6
SR60 batterySR60 6.8x2.15mm Battery 1.5V, 15mAhBAT1
SR60 Keystone 2998SR60 Battery 1.5V 17mAh (Farnell: 1854967)BH1
STM809SWX6F2.93V Reset circuitIC3
TACTR75H84B160Push buttonS1
TL431QDBZT2.5V ADJ Voltage referenceIC54
TPS51200DRCTTEXAS INSTRUMENTSSink/Source DDR Termination RegulatorIC44
TPS53353DQPTHigh-Efficiency 20-A Synchronous Buck Converter with Eco-mode™IC14
TS34312.5V ADJ Voltage referenceIC45
VM53S3-25.000-2.5/-30+75MERCURY ELECTRONICSOscillator 4 pins 1=VC 2=GND 3=OUT 4=+VsOSC5
XC7K325T-2FFG900CXILINXKINTEX-7, FPGA, 900-Ball BGA, Commercial Grade, Pb-FreeIC79

About this design

Design
AFCK — OHWR0032 R1
Designed in
Altium Designer
Board
14 copper layers, 180.7 × 148.6 mm, 1406 components, 1416 nets
Licence
CERN-OHL-1.1
Source
https://gitlab.com/ohwr/project/afck at commit 6ec91dd9ba53
Attribution
CERN Open Hardware Repository, project ohwr/project/afck (commit 6ec91dd9ba53a2c270d1d91fcc4e6bd6f61cff9e, 2025-04-09). Licence: CERN-OHL-1.1. Source: https://gitlab.com/ohwr/project/afck. This is a snapshot of the copy published on the OHWR GitLab pages (R1), not CERN's released record. CERN's source of truth is EDMS (https://edms.cern.ch), and the current release may be newer.
Reviewed with
Probe by E# (E-Sharp AB)