MICROSTEPPER DDS
OHWR0076 · revision R1
MICROSTEPPER DDS (OHWR0076 R1) is an open-hardware circuit board from the CERN Open Hardware Repository, designed in KiCad EDA: a 4-layer board of 100 × 100 mm with 221 components and 157 nets, published under CERN-OHL-S-2.0.
review 97% walked · 9 findings: 1 blocker · 3 major · 5 minor (13 items)
Open in the interactive viewerRead the full review

Design review
Probe recorded 9 findings: 1 blocker · 3 major · 5 minor (13 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.
DDS board: the 1.8 V LDOs (U701, U801, U1201, U1301) are set to 3.3 V - SET resistor is 33k on all ten LT3045s, over the AD9912's 2 V DVDD/AVDD absolute maximum
The four 1.8 V LDOs on the DDS board (U701, U801, U1201, U1301) are set to 3.3 V, so both AD9912s (U301, U401) get 3.3 V on supplies rated 2 V absolute maximum at first power-up.
All ten LT3045s are instances of one reused hierarchical sheet (
LT3045_LDO.kicad_sch). KiCad shares the sheet's values across instances, so every SET resistor is 33k. The sheet's own note isVOUT = 100uA * RSET.What each LDO regulates to
Quantity Value Source SET pin current 100 µA LT3045 datasheet, p.12: "SET sources a precision 100µA current" SET resistor, all ten instances 33k (R317, R417, R601, R701, R801, R901, R1101, R1201, R1301, R1401) BOM; schematic sheet LT3045_LDOOutput of every LDO 100 µA × 33 kΩ = 3.3 V derived AD9912 DVDD / AVDD operating range 1.71–1.89 V AD9912 datasheet, p.3 AD9912 DVDD / AVDD absolute maximum 2 V AD9912 datasheet, p.7 SET resistor needed for 1.8 V 18k (1.8 V / 100 µA) derived Which rails are wrong
3.3 V is right for PLL33, PLL_VVCO33, AVDD33_DAC and DVDD33. It is wrong for the four 1.8 V rails:
``
mermaid flowchart LR subgraph "Shared LDO sheet: RSET = 33k" U701["U701 / R701"] U1201["U1201 / R1201"] U801["U801 / R801"] U1301["U1301 / R1301"] end U701 -- "DVDD18 = 3.3 V, via FB301" --> A["U301 AD9912 (DDS A)<br/>DVDD pins 3, 5, 7<br/>abs max 2 V"] U801 -- "AVDD18 = 3.3 V" --> A2["U301 AVDD pins 11, 19, 23-26, 29, 30, 36, 42, 44, 45, 53"] U1201 -- "DVDD18 = 3.3 V, via FB401" --> B["U401 AD9912 (DDS B)<br/>DVDD pins 3, 5, 7"] U1301 -- "AVDD18 = 3.3 V" --> B2["U401 AVDD pins"]``DVDD18 also feeds the CLKMODESEL pull-ups R308 / R408.
Why it matters
Built as the BOM states, both DDS chips see 1.3 V over their absolute maximum on the core and analog supplies the moment the board is powered. Nothing in the schematic or BOM marks the 1.8 V instances as needing 18k. A builder following the documents fits 33k everywhere.
DDS board: J105 pin 9 passes raw DC_IN (documented up to 20 V) to the MKR IO adaptor, which feeds it through a 0R to the Arduino VIN rated 6 V max
DDS J105 pin 9 passes the raw DC_IN rail, documented up to 20 V, across the ribbon to the MKR IO adaptor, where R2 (0R) joins it to the Arduino MKR VIN, rated 5 V, maximum 6 V.
The power path
``
mermaid flowchart LR A["J104<br/>'Input for LDOs', MAX 20 V"] --> B["D101, FL101, FL102"] B -- "DC_IN" --> C["Ten LT3045 inputs"] B -- "DC_IN" --> D["DDS J105.9"] D -- "10-way ribbon" --> E["MKR IO J2.9<br/>DC_5V"] E -- "R2 0R, fitted" --> F["Arduino MKR VIN<br/>5 V, max 6 V"] E -- "R1 200R" --> G["MKR IO LED D1"]``The two documents
Quantity Value Source J104 input limit "MAX 20 V" DDS schematic, page 1 J105.9 net DC_IN, with the IN, EN/UV and PGFB pins of U302, U402, U601-U901, U1101-U1401 and FL102.2 DDS netlist J2.9 net on OHWR0078 R1 MICROSTEPPER MKR IO DC_5V, J2 pinout matches DDS J105 pin for pin MKR IO netlist R2 on MKR IO 0R, fitted, not DNP, DC_5V to VIN MKR IO netlist Arduino VIN limit "Vin: 5 V, maximum 6 V"; "5-6V supply from DDS-board" MKR IO schematic LED current at 12 V D1 with only 200R, ~50 mA statement Why it matters
The DDS board itself only needs about 3.3 V plus the LT3045 dropout. Any supply above 6 V that the DDS note allows reaches the Arduino VIN directly, and puts DC_IN across the MKR IO LED D1 with only 200R.
- FiducialMissingmajorOpen
FiducialMissing: 2 boards×2
2 checked by the review by evaluation@esharp.se — 2 different reasons
At bottom, top
Each finding, with its own headline
bottomThe bottom side carries 22 SMT component(s) and no fiducial was found on it
The bottom side carries 22 SMT component(s) but no fiducial was found on it — a placement machine has no optical reference to align to. Probe found no pad labelled .pad_usage=g_fiducial/l_fiducial (basis pad_usage); no single-pad, netless, undrilled component whose 0.5-3 mm round pad has a concentric solder-mask opening at least 1.5x its diameter with no other copper inside it (basis geometry); and no component whose refdes starts FID or whose part or package name contains FIDUCIAL (basis footprint-name). If your library names fiducials something Probe does not recognise, mark this benign — disposition_finding(rule='FiducialMissing', entity='bottom', status='benign', reason=…) — the name match only ever silences this check, it never asserts anything. If the side really has no fiducial, say so with status='confirmed': that answers the finding and keeps it gating release.
topThe top side carries 180 SMT component(s) and no fiducial was found on it
The top side carries 180 SMT component(s) but no fiducial was found on it — a placement machine has no optical reference to align to. Probe found no pad labelled .pad_usage=g_fiducial/l_fiducial (basis pad_usage); no single-pad, netless, undrilled component whose 0.5-3 mm round pad has a concentric solder-mask opening at least 1.5x its diameter with no other copper inside it (basis geometry); and no component whose refdes starts FID or whose part or package name contains FIDUCIAL (basis footprint-name). If your library names fiducials something Probe does not recognise, mark this benign — disposition_finding(rule='FiducialMissing', entity='top', status='benign', reason=…) — the name match only ever silences this check, it never asserts anything. If the side really has no fiducial, say so with status='confirmed': that answers the finding and keeps it gating release.
- ReferencePlaneDiscontinuitymajorOpen
ReferencePlaneDiscontinuity: 4 boards×4
e.g. “'/SIO' crosses where the outline of the GND plane on IN2.CU.PWR is drawn back, at 1 place, the longest 0.88 mm (fastest: /SIO at 100 ps 10-90 %, ASSUMED)” · 4 different statements
At void:GND@IN2.CU.PWR:120.5,-80.25, void:GND@IN2.CU.PWR:126.78,-99.62, void:GND@IN2.CU.PWR:136.76,-103.33, void:routed@B.CU|IN2.CU.PWR
Each finding, with its own headline
void:GND@IN2.CU.PWR:120.5,-80.25'/SIO' crosses where the outline of the GND plane on IN2.CU.PWR is drawn back, at 1 place, the longest 0.88 mm (fastest: /SIO at 100 ps 10-90 %, ASSUMED)
Along '/SIO' on B.CU, the GND plane on IN2.CU.PWR has 0.88 mm under the track with no copper, at (120.5, -80.25), the same plane on both sides. No single drill's anti-pad and no landing explains it: a clearance is excused only when it holds a drill and is drawn round it, or the track ends in it on the net's own pad or via, or runs through its own via's clearance and no farther. The return current detours round the void. No hole cleared it: the outline of the pour at x 100.08..199.92, y -149.92..-50.08 is drawn back there. 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.
- '/SIO' on B.CU at (120.5, -80.25), 0.88 mm over no copper; 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/77/reference-planes?group=void%3AGND%40IN2.CU.PWR%3A120.5%2C-80.25 serves each with its own layer_image region; all of them lie in x 119.5..121.5, y -81.69..-78.81 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.
void:GND@IN2.CU.PWR:126.78,-99.62'/SIO' crosses where the outline of the GND plane on IN2.CU.PWR is drawn back, at 1 place, the longest 1.48 mm (fastest: /SIO at 100 ps 10-90 %, ASSUMED)
Along '/SIO' on B.CU, the GND plane on IN2.CU.PWR has 1.48 mm under the track with no copper, at (126.78, -99.62), the same plane on both sides. No single drill's anti-pad and no landing explains it: a clearance is excused only when it holds a drill and is drawn round it, or the track ends in it on the net's own pad or via, or runs through its own via's clearance and no farther. The return current detours round the void. No hole cleared it: the outline of the pour at x 100.08..199.92, y -149.92..-50.08 is drawn back there. 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.
- '/SIO' on B.CU at (126.78, -99.62), 1.48 mm over no copper; 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/77/reference-planes?group=void%3AGND%40IN2.CU.PWR%3A126.78%2C-99.62 serves each with its own layer_image region; all of them lie in x 125.44..128.25, y -100.96..-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.
void:GND@IN2.CU.PWR:136.76,-103.33'Net-(R305-Pad1)' crosses where the outline of the GND plane on IN2.CU.PWR is drawn back, at 1 place, the longest 2.86 mm (fastest: Net-(R305-Pad1) at 100 ps 10-90 %, ASSUMED)
Along 'Net-(R305-Pad1)' on B.CU, the GND plane on IN2.CU.PWR has 2.86 mm under the track with no copper, at (136.76, -103.33), the same plane on both sides. No single drill's anti-pad and no landing explains it: a clearance is excused only when it holds a drill and is drawn round it, or the track ends in it on the net's own pad or via, or runs through its own via's clearance and no farther. The return current detours round the void. 1 of them run ALONG the bare copper, parallel to its edges, rather than across it (the longest 1.41 mm). No hole cleared it: the outline of the pour at x 100.08..199.92, y -149.92..-50.08 is drawn back there. 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.
- 'Net-(R305-Pad1)' on B.CU at (136.76, -103.33), 2.86 mm, running ALONG the bare copper for 1.41 mm, not across it over no copper; 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/77/reference-planes?group=void%3AGND%40IN2.CU.PWR%3A136.76%2C-103.33 serves each with its own layer_image region; all of them lie in x 134.25..139, y -104.33..-102.07 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.
void:routed@B.CU|IN2.CU.PWR9 nets on B.CU cross the tracks routed through the GND pour on IN2.CU.PWR at 15 places, the longest 27.79 mm; 4 run ALONG the cut round another net's track (fastest: /CS1 at 100 ps 10-90 %, ASSUMED)
The GND pour on IN2.CU.PWR is cleared round tracks routed through it, and tracks on B.CU, which take that pour as their reference, cross them: for the width of each cut the return has no plane under it and detours round the cut. 15 place(s) on 9 net(s). 4 of the crossings run ALONG the cut round another net's track rather than across it, parallel to its edges (the longest 25.94 mm): over or right beside that track, broadside to it as well as off the plane. One decision covers the layer pair: whether these layers may take a pour their own tracks cut as the reference, or the routing or the stackup should give the tracks a plane that is not cut where they cross. No crossing has a plane on the track's other side, so the whole return meets the break.
- '/CS3' on B.CU at (155.51, -140.72), 27.79 mm, running ALONG the bare copper for 25.94 mm, not across it over copper of PLL_VVCO33, not a plane; other side: none: the signal layer is outermost
- '/CS2' on B.CU at (137.71, -83.83), 14.3 mm, running ALONG the bare copper for 12.04 mm, not across it over copper of /sheet_DDS_B/AVDD18, /sheet_DDS_B/DVDD33, not a plane; other side: none: the signal layer is outermost
- '/Sheet_PLL/Vtune' on B.CU at (121.83, -130.27), 8.05 mm, running ALONG the bare copper for 6.16 mm, not across it over copper of PLL33, not a plane; other side: none: the signal layer is outermost
- '/SIO' on B.CU at (121.91, -137.25), 3.66 mm over copper of PLL33, not a plane; other side: none: the signal layer is outermost
- '/CS2' on B.CU at (160.71, -113.66), 3.43 mm, running ALONG the bare copper for 1.75 mm, not across it over copper of /Sheet_DDS_A/DVDD18, not a plane; other side: none: the signal layer is outermost
- '/CS1' on B.CU at (128.08, -107.92), 2.04 mm over copper of /Sheet_DDS_A/DVDD33, not a plane; other side: none: the signal layer is outermost
- '/CS2' on B.CU at (160, -102.34), 1.8 mm over copper of /Sheet_DDS_A/AVDD33_DAC, not a plane; other side: none: the signal layer is outermost
- '/SIO' on B.CU at (122.87, -103.53), 1.78 mm over copper of /Sheet_DDS_A/DVDD33, not a plane; other side: none: the signal layer is outermost
- '/SIO' on B.CU at (123.53, -75.5), 1.6 mm over copper of /sheet_DDS_B/AVDD18, not a plane; other side: none: the signal layer is outermost
- 'Net-(R304-Pad1)' on B.CU at (137.63, -105.34), 1.52 mm over copper of /Sheet_DDS_A/DVDD18, not a plane, to the end of the track; other side: none: the signal layer is outermost
- 'Net-(R404-Pad1)' on B.CU at (134.95, -78.93), 1.51 mm over copper of Net-(C406-Pad1), not a plane, to the end of the track; other side: none: the signal layer is outermost
- '/SCK' on B.CU at (124.38, -76.5), 1.41 mm over copper of /sheet_DDS_B/AVDD18, not a plane; other side: none: the signal layer is outermost
- and 3 more
Every site is a row of this finding's evidence, keyed by net, layers and position. GET /api/odb/boards/77/reference-planes?group=void%3Arouted%40B.CU%7CIN2.CU.PWR serves each with its own layer_image region; all of them lie in x 119.62..169.86, y -141.72..-74.5 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.
DDS board: the ten LDO-output LED resistors (R318, R418, R602, R702, R802, R902, R1102, R1202, R1302, R1402) have no value; BOM and schematic say only 'R'
The ten LDO-output LED resistors on the DDS board have no value: BOM and schematic say only "R", so an assembler cannot fit them and the LED current cannot be checked.
Affected: R318, R418, R602, R702, R802, R902, R1102, R1202, R1302, R1402.
- Each LT3045 sheet instance puts an indicator LED (D301 etc.) from the LDO output to GND through this series resistor.
- Its value field is the bare symbol name "R".
- The BOM lists all ten as value "R", no MPN, footprint R_0805_2012Metric.
- Whether the LED loads the 1.8 V rails noticeably cannot be checked.
DDS schematic: Date is blank on all 14 sheets and Rev is filled only on the two AD9912 sheets
The DDS schematic has a blank Date on all 14 sheets, and Rev is filled only on the two AD9912 sheets, so a printed sheet cannot be tied to a board revision.
- Title blocks carry title, size and sheet Id n/14 throughout.
- "Date:" is empty on every sheet.
- "Rev:" is empty on sheets 1-4, 6-9 and 11-14. Only sheets 5 and 10 (AD9912 DDS) state "Rev: 2020-01 (draft)".
- The PCB silk states "Design Anders Walin, 2020-12-08".
DDS schematic on file prints one instance's designators on every reused sheet: both AD9912 pages show U401/R401-R413, all ten LDO pages show U302/R317/R318/D301/TP301
The DDS schematic PDF prints one instance's designators on every reused sheet: both AD9912 pages show U401/R401-R413, and all ten LDO pages show U302/R317/R318/D301/TP301, so a page cannot be tied to the physical part it shows.
Where it is
Reused sheet Instances Sheets in the PDF Designators printed on all of them AD9912.kicad_sch 2 (Sheet_DDS_A, sheet_DDS_B) 5, 10 U401, R401-R413 LT3045_LDO.kicad_sch 10 3, 4, 6-9, 11-14 U302, R317, R318, C328-C330, D301, TP301 The board places U301/U401, U302, U402, U601-U901, U1101-U1401 and their own resistors.
Why it matters
A reader cannot tell from a page which physical LDO or DDS it shows, for example that sheet 8 is U801/R801 on AVDD18.
DDS silkscreen: superimposed designators at R401/R402/R403, R301/R302/R303 and at the LDO column (U1101 over C1101, U1301 against C1301)
Superimposed designators on the DDS top silk at R401/R402/R403, R301/R302/R303 and in the LDO column (U1101 over C1101, U1301 against C1301) make these positions ambiguous at assembly and rework.
- R403, R402, R401 are vertical and placed end to end, so "R403R402R401" reads as one string, the R of each overlapping the last digit of the next.
- The same arrangement repeats at R303/R302/R301 beside U301.
- In the LDO column the horizontal U1101 designator crosses the vertical C1101 designator, and U1301 touches C1301. Same pattern at the other LT3045 instances.
Text height is ~0.8-1.2 mm with ~0.12 mm stroke, so size is at the floor; the problem is the collisions.
DDS board has no fiducials although it carries 0.5 mm-pitch LFCSP-64 (U301, U401) and LFCSP-32 (U201) parts
The DDS board has no fiducials, although it carries 0.5 mm-pitch leadless parts (U301, U401 in LFCSP-64; U201 in LFCSP-32) and is populated on both sides.
Probe finds no fiducial by pad usage or refdes, and none is visible on the top copper/silk render.
- U301, U401: AD9912, LFCSP-64 9x9 mm, 0.5 mm pitch
- U201: ADF4351, LFCSP-32 5x5 mm, 0.5 mm pitch
- Ten MSOP-12 0.65 mm pitch LDOs
- Components on both sides, 22 on the bottom
Main parts
| Part number | Maker | Description | Where |
|---|---|---|---|
| ADF4351 | U201 | ||
| BNX025 | FL102 | ||
| AD9912 | U301, U401 | ||
| ADT2-1T-1P | TR301, TR401 | ||
| LFCN-400 | FL1001, FL501 |
About this design
- Design
- MICROSTEPPER DDS — OHWR0076 R1
- Designed in
- KiCad EDA
- Board
- 4 copper layers, 100 × 100 mm, 221 components, 157 nets
- Licence
- CERN-OHL-S-2.0
- Source
- https://gitlab.com/ohwr/project/microstepper at commit
46ae03051ab5 - Attribution
- CERN Open Hardware Repository, project ohwr/project/microstepper (commit 46ae03051ab5b384582b0fa5100d2e5447509ebe, 2025-04-08). Licence: CERN-OHL-S-2.0. Source: https://gitlab.com/ohwr/project/microstepper. 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)