A motor control PCB copy takes a physical drive board — VFD control card, BLDC/PMSM controller, servo amplifier, stepper driver — and produces manufacturable files from it: Gerbers, drill data, BOM, netlist, and schematic if you need one. You send the board (a working one plus a dead one is ideal), we strip, image, and trace it, then rebuild the design in CAD and verify the copy against the original net for net. If the MCU or DSP holds firmware you own and no longer have, that runs as a separate track alongside the board work.
Motor drives are not generic boards. Half the value in this kind of job is in the parts that a careless copy gets wrong: gate-drive loop geometry, shunt sense routing, isolation gaps, and the thermal copper under the power stage.
Why motor drive boards behave differently from ordinary boards
On a logic board, a trace either connects or it doesn’t. On a motor control board, how it connects decides whether the thing survives its first full-load start. Three things drive that.
First, the gate loop. The path from the driver output through the gate resistor to the IGBT or MOSFET gate, and back through the source/emitter return, has to keep its original length and its original return geometry. Stretch it by a few millimeters or move the return path to a different plane, and you change ringing, dv/dt, and shoot-through margin. So we record gate resistor values, package sizes, and the physical loop area rather than just the netlist entry.
Second, current sensing. Shunt resistors, their Kelvin connections, and the differential pair back to the amplifier are the reason a torque loop is stable. A copy that reroutes those two traces as ordinary signals will read noise as current. We capture shunt geometry, sense tap points, and the filter components around the op-amp exactly as they sit on the board.
Third, isolation. Most drives have a high-voltage side and a control side separated by opto-couplers, digital isolators, or a transformer. That physical gap is a design decision — the creepage and clearance distances under the isolator body, the slot in the copper, the absence of silkscreen or via stitching across the barrier. In the copy we measure that gap and preserve it. IPC-2221 spacing guidance is the reference we check the original against, and if the original board was already marginal for its rated bus voltage, we say so in the report instead of silently reproducing it.
What the engineer is actually looking at
When a drive board lands on the bench, the first pass is optical and electrical, not destructive. We photograph both sides at high resolution, log every marking including date codes and any conformal coating, and identify the power stage — discrete FETs, an IPM, or a six-pack module. Then we ring out the accessible nets, check the DC bus caps for bulge or ESR drift, and note anything that looks like a prior repair. Rework on the original board is one of the most common sources of a bad copy, because a hand-added jumper looks exactly like an intended net if nobody flags it.
How a motor control PCB copy runs, step by step
- Intake and imaging. Both sides scanned, dimensions and board thickness measured, mounting holes and connector datums located, silkscreen recorded before anything is removed.
- Component removal and identification. Parts come off under controlled heat, get bagged by reference designator, and are measured or read where the marking is sanded, coded, or a house number. Magnetics — current transformers, gate-drive transformers, common-mode chokes — get turns-ratio and inductance measurements, since a datasheet usually doesn’t exist for them.
- Layer separation. Outer layers are imaged directly. Inner layers on a four- or six-layer control card are exposed one at a time, imaged, and registered to the drilled hole pattern so nothing shifts. Power boards with 3 oz or heavier copper etch and delaminate differently than 1 oz FR-4, and that is planned for rather than discovered halfway through.
- CAD rebuild. Copper is vectorized into real polygons and traces, not traced pixel art. Drill sizes, annular rings, plane splits, and the thermal reliefs under the power devices all get rebuilt. Where the original used heavy pours for heat spreading, we keep the pour, not a trace of equal ampacity — the copper is a heatsink, not just a conductor. IPC-2152 is the sanity check on any trace we have to reconstruct through damage.
- Verification. The rebuilt netlist is compared against continuity measured on the physical board, net by net. Fabrication output is checked for the kind of defects IPC-A-600 and IPC-6012 care about — annular ring, spacing, plating — before anyone quotes a fab run.
- Optional build and test. Bare boards, assembly, and a powered bench check into a real motor or a dyno load if you want proof rather than files.
The same workflow underpins our broader instrumentation and test equipment board copy work, but motor drives add the power-stage and isolation checks above.
What you get, and which tier fits your situation
| Deliverable set | What’s in it | When it’s the right choice |
|---|---|---|
| 1:1 copy package | Gerbers, drill files, stackup, BOM with reference designators, pick-and-place, silkscreen | You need identical replacement boards for equipment already in the field |
| Copy + netlist | Everything above plus a verified netlist and connectivity report | You want to run your own DRC or hand the design to another fab |
| Full reverse engineering | Editable schematic with functional blocks labeled — power stage, gate drive, sensing, protection, control | You plan to modify, re-rate, or requalify the drive |
| Copy + firmware track | Board files plus recovery of code from an MCU or DSP you own | The controller is programmed and you’ve lost the source or hex |
The firmware question on motor control boards
A drive board is inert without its control code. The commutation tables, current-loop gains, and fault thresholds live inside the MCU or DSP, so a perfect PCB copy of a programmed controller board still needs a plan for the code.
Legitimate reasons this comes up constantly in motor control: the original engineer left and the repository went with him, the toolchain was for a compiler nobody licenses anymore, or the controller itself is obsolete and you need the binary to port the design forward. Failure analysis on a burned drive is another — reading what the fault log stored before the stage let go tells you more than the debris does.
Feasibility depends entirely on the part and how it was protected. TMS320 C2000 devices are everywhere in servo and inverter designs and their password/security-module behavior varies a lot by generation. STM32 parts differ sharply between RDP Level 1 and Level 2. Older 8051-class and PIC parts in low-cost fan and pump drives have simpler lock-bit schemes. We assess the specific marking on your board and tell you honestly whether the answer is likely, unlikely, or needs a sacrificial sample first — never a guaranteed success rate. Related work sits on our firmware and chip service pages.
You need to hold the rights to the design you send us. Everything runs under NDA, boards and files stay in-house, and we don’t reuse a customer’s design for anyone else.
Where a motor control board copy gets hard, and where it sometimes fails
Honest constraints, because you will hit at least one of these:
- Potted or encapsulated modules. Intelligent power modules and potted gate-drive bricks can be opened, but the internal die layout is a separate job from copying the carrier board. Sometimes the practical answer is sourcing the module, not reproducing it.
- Custom magnetics. Current transformers and gate-drive transformers wound to an OEM spec need measurement and a wind-to-match partner. We can characterize them; matching them exactly takes longer than the rest of the board.
- Thick copper and metal-core power boards. Heavy copper etches unevenly during separation and IMS/aluminum-backed boards need a different approach than FR-4. Doable, slower, and worth telling you up front.
- Severely burned boards. If a short vaporized copper across several inches, that region has to be inferred from the surrounding topology and the function of the circuit. A second, undamaged sample removes the guesswork entirely — this is the single biggest thing you control.
- Obsolete controllers and drivers. Sometimes the honest recommendation is a partial redesign around a current part rather than a 1:1 clone that you can’t build in volume.
- Conformal coating. Common on drives in dirty or humid environments. It comes off, but it adds a step and can lift silkscreen with it, so we image before stripping.
What to send us to get a real quote
Guessing wastes both our time. Send these and you’ll get feasibility, lead time, and price rather than a range:
- Clear photos of both sides, whole board in frame, on a dark non-reflective background, taken straight on with even light. Add close-ups of the power stage and any hand-modified area.
- Board outline dimensions and layer count if you know it, plus board thickness.
- The marking on the MCU, DSP, or gate-driver IC, typed out — coating and glare make photos of chip tops unreliable.
- Motor type and ratings the board drives: bus voltage, phase current, BLDC vs induction vs stepper.
- One line on deliverables: replacement boards only, files only, files plus schematic, or files plus firmware.
- Whether you can ship a spare board and whether it’s known-good or failed.
Cost on this kind of work is driven by layer count, board area, component count, whether there are fine-pitch BGAs or an IPM, how much of the board is damaged, whether custom magnetics are involved, and whether firmware is in scope. Those variables move the number far more than any list price would suggest, which is why we quote per board.
FAQ
Can you copy a VFD control board without the schematic?
Yes — that’s the normal case. The schematic is an output of the process, not an input. We der
Working on a board like this?
Send the chip marking or two photos. You get feasibility, lead time and price within 24 hours, and the check costs nothing.
Get a free quote