Electric Vehicle PCB copy means taking a physical EV board — an AC charger controller, a battery management slave, a DC-DC converter, a traction inverter gate driver card — and recovering the manufacturing data set from it: layer-by-layer copper artwork, drill data, a verified BOM, and a schematic if you need one. You send a board, we send back files you can fabricate, assemble and repair from. That’s the whole service in one sentence.
The reason people ask for it is almost always the same: a fleet or a charging site is down, the original supplier is gone or won’t sell the spare, and there are no design files anywhere in the company.
Which EV boards actually come through the lab
EV electronics aren’t one board class, and the work changes a lot depending on which one is sitting on the bench.
- AC charging controllers and EVSE mainboards — mixed-signal, 4 to 6 layers, a control MCU, a relay or contactor driver stage, RCD/RCM sensing, PLC or CAN comms, and usually a metal-clad or thick-copper section for the mains path.
- BMS master and slave boards — dense, low-power, lots of precision analog. Cell-monitoring front ends with matched divider networks, isolated daisy-chain links, shunt or Hall current sensing. Component tolerance matters more here than anywhere else in an EV.
- Traction inverter driver and control cards — high-voltage isolation barriers, creepage and clearance slots milled into the board, gate resistors sized to a specific IGBT or SiC module. The layout is part of the safety design.
- DC-DC converters and on-board chargers — planar transformers made from copper on the inner layers, so the stackup and the exact turn geometry are the circuit, not just wiring.
- Auxiliary and body modules — thermal management controllers, PTC heater drivers, pump and fan controllers, charge-port lock and lighting drivers.
The heavy-copper power stages and the isolation geometry are what separate EV work from ordinary industrial boards. Getting a BMS slave 95 percent right is worthless — a mismatched divider puts every cell voltage reading off by an amount that never trips a fault but quietly ruins pack balancing.
What actually happens to your board
Intake and non-destructive capture
The board gets photographed and scanned on both sides at high resolution before anything touches it. Every component reference, marking line and date code is logged, including the ones that will be sanded off later. Conformal coating on EV boards is normal, so it gets identified first — some coatings lift with solvent, some need controlled abrasion, and choosing wrong destroys silkscreen you still need.
Potted modules get X-rayed before we decide whether to open them. If a charging controller has an epoxy-filled section over the mains stage, the X-ray tells us whether there’s anything under there worth recovering or just a common-mode choke and two Y-caps.
Component identification and BOM
Passives get measured, not guessed. Every resistor in a cell-sense divider is metered off-board, every electrolytic and film cap gets its value, voltage rating and ripple class recorded. Magnetics are measured for inductance and DCR, and where a transformer is custom we document the winding structure so it can be re-specified.
ICs are read from their markings where readable, and inferred from pinout and function where the top has been lasered or ground. Automotive-grade parts often carry house numbers that aren’t in any distributor catalog — we identify those by package, pin function tracing and behavior, and we flag them clearly rather than pretending we know.
Layer separation and net extraction
Outer layers are imaged directly. Inner layers are exposed by controlled grinding, one plane at a time, with imaging after every pass and registration holes to keep everything aligned. On a 6-layer charger board with two ground planes, that’s four separate grind-and-image cycles.
Every net is then verified with continuity checks against a second, intact sample. This is the step where mistakes hide, and it’s why a second board matters so much. If you only have one, we still work, but verification becomes electrical probing on a partially destroyed board instead of a clean cross-check.
Rebuild and verification
The extracted copper becomes a real CAD database — layers, stackup, drill table, netlist, footprints — not a bitmap traced over. Acceptability of the fabricated result is judged against IPC-A-600 and IPC-6012 class expectations, and where a power trace looked marginal in the original we check it against IPC-2152 current-carrying guidance and tell you rather than silently “improving” it. Isolation gaps and creepage distances are measured and preserved; on a high-voltage EV board those spacings often exist because of a safety approval, and shrinking one is a real hazard, not a layout tweak.
If you order assembled samples, they get powered up on a bench and driven through the functions we can exercise without a vehicle — CAN traffic, PWM outputs, relay switching, isolated ADC readings on a simulated cell stack.
What to send us, and what comes back
Nothing here is complicated, and getting it right the first time removes a round of email.
- Photos of both sides, whole board in frame, laid flat on a dark non-reflective surface, even lighting, no flash glare across the silkscreen.
- Board dimensions in millimeters, plus overall thickness if you can measure it, and layer count if you know it.
- The main chip marking — every line of text on the MCU or the main controller, transcribed exactly. If it’s illegible, say so.
- One line on deliverables: Gerbers only, Gerbers plus BOM, full schematic, or files plus assembled samples.
- What it does and how it failed, if the goal is repair. “Charger stops at 2 kW and throws a ground fault” narrows the analysis enormously.
Back from us: a written feasibility assessment naming the specific risks on your board, a lead time, and a price. That comes before any commitment, and it says what we can’t guarantee as plainly as what we can. If you want the difference between recovering artwork and recovering a full functional understanding spelled out, our note on how a copy differs from a clone covers it.
Deliverable tiers and what each one is good for
| Deliverable set | Contains | Right when |
|---|---|---|
| Fabrication data | Gerbers, drill files, stackup, solder mask and paste layers | You have the BOM and just need bare boards rebuilt |
| Fabrication + BOM | Above, plus verified BOM with measured passive values and package data | You’ll assemble locally or hand the package to a contract manufacturer |
| Full reverse engineering | Above, plus netlist and editable schematic | You need to modify the design, swap an obsolete part, or debug a field failure |
| Files + firmware | Above, plus recovered firmware image from the controller | The board is useless without its code and the source is lost |
| Turnkey samples | Assembled, tested boards built from the recovered data | You want working spares, not a CAD package |
The firmware question on EV boards
A copied charger mainboard with an empty MCU is a paperweight. Where the controller is locked and the original hex file is gone, firmware recovery becomes part of the job — reading out the code so it can be reprogrammed into replacement units of the same part.
Feasibility depends entirely on the part and how it was configured at end of line. STM32 devices sitting at RDP Level 1 are a different proposition from Level 2, which permanently disables debug access. MSP430 parts hinge on whether the JTAG fuse was blown and whether a BSL password was set. Classic 8051 and AVR derivatives depend on which lock bits were programmed. Automotive-qualified MCUs with secure boot, on-chip encryption or a hardware security module are frequently not feasible at all, and we say that up front instead of taking your money to find out.
We do this for people who own the design and lost the files, for out-of-production equipment that still has years of service life, for obsolescence migration when the original part no longer exists, and for failure analysis. You need to hold the rights to the design you’re asking about, and you’re responsible for compliance with applicable IP law; every project runs under NDA, and if you’re checking the ground rules first, the short answer on the legality of copying a board you own is worth two minutes.
Where this gets hard, and where we’ll say no
Honest constraints, because you’d rather hear them now:
- One heavily damaged sample. A board with a blown power stage that took the surrounding copper with it leaves gaps. We can often infer them from the circuit topology, but “often” isn’t “always.”
- Custom magnetics and planar transformers. The copper geometry we can copy exactly. The core material and gap we characterize by measurement and inference, and there is genuine uncertainty in matching a proprietary ferrite grade.
- House-numbered automotive ASICs. If a BMS front end is a custom part built for one OEM, no amount of reverse engineering makes it purchasable. The recovered files are still useful for repair; new-build isn’t on the table without a redesign.
- Safety-approved high-voltage boards. We reproduce the isolation geometry faithfully, but a copied board carries no certification. Re-approval for anything touching a vehicle’s HV bus is on you.
- Locked secure-boot controllers. Sometimes the answer is simply no.
Lead time follows the same logic as cost. A two-layer auxiliary controller is a matter of days. A 6-layer BMS master with fine-pitch BGAs, blind vias and conformal coating is meaningfully longer because every inner layer needs its own imaging c
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