Heavy copper PCB reverse engineering costs more and takes longer than the same board in 1 oz copper — usually because copper weight has to be measured rather than assumed, thermal reliefs and split planes are large and irregular, and layer separation on thick-copper laminate behaves differently under heat. For a 2-layer 6 oz busbar-style board with a modest part count, expect a few working days. A 6- or 8-layer power stage mixing 3 oz inner planes with fine-pitch signal layers runs meaningfully longer.
This page is written for the person who has the board in hand and needs a number and a date. Below is what drives both, what we need from you, and where these jobs get hard.
Why copper weight changes the job at all
On a standard board, the etched geometry is what you’re recovering. On a heavy copper board, the cross-section is part of the design intent. A 10 mm trace at 1 oz and a 10 mm trace at 6 oz are two completely different current ratings, and if you rebuild the board with the wrong copper weight it may pass continuity and still cook in service.
So the first thing an engineer does with a thick-copper sample is measure, not trace. Copper thickness gets checked by micrometer on an exposed edge or by cross-sectioning a coupon cut from a non-critical corner of the board. On plated-up outer layers, the base foil and the plating add together, and you get a number that isn’t a clean 2 oz or 3 oz. We record what we measure and note the nearest manufacturable weight, then sanity-check the result against the trace widths and the currents the circuit plausibly carries, using IPC-2152 as the reference for what those widths would support.
Etch factor is the second wrinkle. Thick copper etches with a visibly sloped sidewall — the top of the trace is narrower than the base. When you photograph or scan an outer layer and trace the visible edge, you are capturing the top of the trapezoid, not the copper the fabricator started with. Getting the Gerber right means compensating for that taper, and the compensation depends on how heavy the copper is.
Features you see on thick-copper boards and almost nowhere else
- Huge irregular pours that are really conductors, not ground fills, and have to be captured as shapes rather than routed traces.
- Thermal reliefs and spoke patterns sized for hand or wave soldering into a 4 oz plane — these change assembly behavior and must be reproduced, not redrawn “cleanly.”
- Mixed copper stackups, for example 3 oz on layers 1 and 4 with 1 oz internal signal layers, which means the layer separation strategy is different for each pair.
- Filled and plated slots, press-fit holes, and copper coins that need drill and slot geometry recovered dimensionally, not just topologically.
- Thick prepreg and resin-rich fill between heavy layers, which shifts the dielectric spacing and matters if any signal on the board is impedance-sensitive.
- Bus bars, screw terminals, and bolt-down lugs whose pad geometry and hole tolerance are mechanical parts of the design.
What actually drives the price
Nobody can quote a thick-copper board from a description alone, but the variables are predictable. In rough order of impact:
- Layer count and how the layers mix. A 2-layer 6 oz board is fast. A hybrid stackup where heavy planes sandwich fine signal layers is the expensive case, because it needs two different imaging approaches on one part.
- Board size. Heavy copper boards are often physically large — inverter decks, motor drives, welder controls. Area drives scan time, stitching, and dimensional verification.
- Component count and how exotic the parts are. Power modules, custom magnetics, current shunts and gate drive transformers take longer to identify than 0603 passives.
- Which deliverables you need. Gerber and drill only is one price. Gerber plus BOM is another. A full schematic with net names and functional blocks is the most work by a wide margin.
- Whether any programmable part is involved. If a controller on the board needs its code recovered as well, that’s a separate scope with its own feasibility question.
- Condition of the sample. Conformal coating, potting compound, corrosion, burned areas, or a board that already failed all add hours. Two identical samples — one to work on, one to keep intact — reduce cost.
If you’re comparing quotes, the thing to look at is not the number but whether the vendor priced the copper weight measurement at all. A quote built on the assumption of 1 oz copper will be cheap and wrong. Our general breakdown of what drives PCB reverse engineering cost and lead time applies here too, with copper weight added as a first-order variable.
How long a heavy copper board takes
Lead time tracks the same drivers. These are typical ranges, not commitments — every quote we send names a date for that specific board.
| Board class | Typical scope | Relative lead time |
|---|---|---|
| 2-layer, 4–6 oz, low part count | Gerber, drill, BOM | Shortest — a few working days |
| 4-layer, uniform 3 oz | Gerber, drill, BOM, netlist | Roughly double the 2-layer case |
| 4–6 layer hybrid (heavy outer, thin inner) | Gerber, drill, BOM, schematic | Noticeably longer; two imaging strategies |
| 8+ layer power board with BGAs or modules | Full package plus verification build | Longest; plan in weeks, not days |
| Any of the above with firmware recovery | Add code readout | Runs in parallel, gated by chip feasibility |
Two things stretch a schedule more than customers expect. One is a single sample with no spare — every destructive step then has to be conservative. The other is asking for a schematic on a board where the power topology is unusual; understanding a resonant converter or a multi-phase interleaved stage takes real engineering time, not just tracing. If you need speed above all, tell us which deliverable you actually need first and we’ll sequence it. A rapid turnaround path exists, but it’s honest work compression, not magic.
What we need from you to quote it
Send these five things and you’ll get feasibility, lead time, and price back rather than a request for more information:
- Clear photos of both sides, whole board in frame, on a dark non-reflective background, with even light and no flash glare on the solder mask.
- Board dimensions — length, width, and total thickness. Total thickness plus layer count tells us a lot about the stackup before the board arrives.
- Copper weight if you know it, or a photo of an exposed copper edge with a caliper or ruler in frame if you don’t. Any fab note or old drawing fragment helps.
- The markings on the main controller and power devices — a close, in-focus shot of the top of each chip.
- One line on which deliverables you need: manufacturing files only, files plus BOM, or a full schematic. And whether you’ll want boards built afterward.
What comes back is a written scope: whether we consider the job feasible, what we can and can’t guarantee about the stackup, the deliverable list, a lead time, and a price. If something looks risky, that’s in the reply too.
What you receive at handover
For a typical heavy copper job: Gerber layers with copper weight annotated per layer, drill and slot files with plated/non-plated called out, a stackup drawing with measured dielectric thicknesses, a BOM with manufacturer part numbers, and — if scoped — the schematic in your CAD format plus PDF. Fabrication notes cover the things a house building thick copper needs to hear: minimum trace and gap achievable at that weight, surface finish, and mask thickness over tall copper, which is a real yield issue on 4 oz and up.
Where this gets hard, and where it sometimes doesn’t work
Straight answers, because you’ll hit these anyway:
Potted and coated boards. Heavy copper power boards are often potted for vibration or dielectric reasons. Removing potting without destroying the sample is case-by-case. Some compounds come off cleanly; some are effectively permanent. We tell you before we start, not after.
Delamination on thick copper. Layer separation relies on controlled heat and mechanical work. Heavy copper carries heat differently and the resin between thick layers is often more brittle after years of thermal cycling. On aged boards, expect some layers to come apart imperfectly, which means recovering a plane from two partial images. It’s doable. It costs time.
Burned or failed boards. If a trace vaporized, the original width is gone. We can infer it from the surviving geometry and from current-carrying requirements, but that becomes engineering judgment, and we’ll flag every inferred feature in the deliverables rather than quietly guessing.
Exact copper weight on plated outer layers. We report a measured thickness and the nearest standard weight. If your application needs the exact original foil-plus-plating breakdown, that requires cross-sectioning, which consumes a sample.
Unmarked or house-numbered power parts. Custom magnetics and relabeled controllers may not be identifiable from markings alone. Electrical characterization gets us to an equivalent, not always to the original part number.
One legal note, said once: we work on boards our customers have the right to work on. Typical jobs here are keeping out-of-production industrial equipment alive, rebuilding a design whose files were lost, or migrating away from an obsolete power controller. You confirm you hold the rights; we run the project under NDA and return your samples.
Frequently asked questions
Can you tell the copper weight from photos?
Not reliably. Photos tell us layer count hints, feature sizes, and part markings, which is enough to
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