Thick copper PCB reverse engineering usually costs more and takes longer than a signal board of the same layer count, and the reason is physical: heavy copper planes swallow detail in X-ray, resist clean etch-back during layer separation, and force us to measure copper weight and dielectric thickness instead of assuming them. Budget on layer count, board size, copper weight, and whether you need current-carrying capacity verified. Send photos and dimensions and we’ll quote in a day or two.
Why heavy copper changes the math
On a normal 1 oz board, an engineer can shine light through a thin FR-4 core and trace inner layers optically, or read them from a good X-ray shot. At 3 oz, 6 oz, or 10 oz, that stops working. A 6 oz plane is roughly 210 µm of solid copper — it’s opaque to X-ray at the settings you’d use for signal traces, and turning the power up just blooms the image.
So the workflow shifts toward destructive layer separation on a sacrificial board, plus careful cross-sectioning. That’s more lab time, and it usually means we need a second sample. If you only have one board and it has to survive, say so up front, because it changes the approach and the price.
The second complication is geometry. Heavy copper etches with visibly sloped sidewalls — the trapezoid is not a rounding error at 6 oz. Measuring a trace at the top surface can understate its real cross-section. We measure at the base and at the top and record both, because a fabricator building your new board needs to know which number the design intent was.
Copper weight is a deliverable, not a footnote
On a motor drive, a welding inverter, or a busbar-style power board, copper weight is the design. Get it wrong and the board still works on the bench and cooks in the field. We determine it three ways and cross-check: cross-section measurement under a calibrated microscope, weight-per-area on a cut coupon of known dimensions, and a sanity check against the trace widths and the currents the circuit clearly carries. IPC-2152 is the reference we use for that last check — if a 12 mm trace only makes sense at 4 oz for the current the shunt tells us is flowing, that’s a useful confirmation.
What actually drives the price
We don’t publish a rate card because the spread between a two-layer 4 oz LED driver board and an eight-layer mixed-copper inverter control board is enormous. These are the variables that move the number:
- Layer count and stackup mix. Boards that mix 1 oz signal layers with 6 oz power layers are the expensive case — you get the layer-separation cost of a heavy copper board plus the trace-density cost of a signal board.
- Copper weight itself. More copper means slower, more careful removal and more cross-sections to characterize the stackup honestly.
- Board size. Power boards are often physically large. Area drives scanning passes, stitching work, and the number of coupons we section.
- Component count and package types. A board that’s mostly TO-247s, bus terminals, and film caps is faster to identify than one with a fine-pitch controller and a dozen unmarked passives.
- Thermal features. Copper coins, embedded slugs, thick internal planes tied through arrays of thermal vias, and edge plating all need documenting, and none of it falls out of a plain Gerber extraction.
- Deliverable set. Gerbers and drill files only, versus Gerbers plus a full schematic, netlist, BOM, and stackup report. The schematic is the labor-heavy part.
- Whether firmware is involved. If a locked MCU on the board also needs code recovery, that’s a separate scope with its own feasibility question.
Two things that don’t add much: silkscreen recovery on power boards (there usually isn’t much) and impedance control (rarely relevant on the heavy layers, though it can matter on the signal layers of a mixed board).
How long a thick copper board takes
Honest ranges, not guarantees. A two-layer 3–4 oz board with under 60 components — a typical LED driver board reverse engineering job — is a matter of days for Gerbers, a bit longer if you want the schematic too. A four-layer board with two heavy inner planes moves into a week-plus territory because the planes have to be separated and imaged individually. Six to eight layers with mixed copper weights and a dense control section is meaningfully longer again, and the sectioning work is on the critical path.
| Board profile | Main time sink | Relative lead time |
|---|---|---|
| 2-layer, 3–6 oz, power only | Copper weight verification, large-area scan stitching | Shortest |
| 4-layer, heavy inner planes | Layer separation, plane-to-plane via mapping | Moderate |
| 4–6 layer mixed 1 oz / 6 oz | Two different imaging workflows on one board | Long |
| Heavy copper with copper coins or embedded slugs | Cross-sectioning to document the insert geometry | Long, and needs a sacrificial sample |
| Any of the above plus locked-MCU firmware recovery | Separate feasibility assessment on the chip | Runs in parallel, but sets the floor |
One scheduling note that catches people out: shipping. If the sample is coming from North America or Europe, transit and customs can easily exceed the engineering time on a two-layer board. Get the sample moving before you finalize the PO.
What we need from you, and what comes back
To quote thick copper PCB reverse engineering accurately, send this much and no more:
- Clear photos of both sides, whole board in frame, on a dark background, with the camera square to the board. Even lighting, no flash glare on the solder mask.
- Board dimensions in mm, plus overall thickness if you can measure it with calipers. Thickness is a strong hint about layer count and copper weight.
- Copper weight if you know it, or just say “looks like heavy copper, traces are visibly raised.”
- The main chip marking, typed out exactly as printed, including the line under the part number.
- One line on deliverables — Gerbers only, Gerbers plus schematic, or full package with BOM and stackup report.
- How many samples you can send, and whether any of them can be destroyed.
Back from us: a feasibility read (including anything we think will be hard), a lead time, and a price. If something in the photos looks like a problem — potting compound, a scraped-off part marking, a suspected embedded copper coin — we’ll flag it in the reply rather than after you’ve paid.
The deliverable that matters most on power boards
Ask for the stackup report as a named deliverable, not as an afterthought. It should state each layer’s copper weight, the dielectric between layers, the finished board thickness, the base and top trace widths on the heavy layers, and the drill/via structure. Without it, a fab house will quote your Gerbers at 1 oz and ship you a board that fails thermally. Gerbers describe geometry; they don’t describe copper weight. This is also where the acceptability language of IPC-6012 and IPC-A-600 becomes practical — plating thickness and annular ring criteria are what your new fabricator will actually be judged against.
Where thick copper gets hard, and where it sometimes doesn’t work
Straight answers on the limits:
- Single-sample destructive jobs. If we only have one board and it must survive, heavy inner layers may not be fully recoverable. We’ll tell you what we can and can’t get before starting.
- Potted or epoxy-encapsulated power modules. Common on inverters. Removal risks damaging the very copper we’re trying to measure. Feasibility is case by case.
- Heavily corroded or delaminated boards. Field-returned power boards often arrive with lifted pads and burned areas. We can usually infer the original copper from the surrounding geometry, but a burned-away section of plane is a judgment call, and we mark it as reconstructed rather than measured.
- Custom magnetics and planar transformers. Winding turns embedded in the copper layers are recoverable; the ferrite core part number and gap often are not, unless it’s marked.
- Unmarked or house-numbered controllers. We can identify function from the circuit, but exact part matching sometimes isn’t possible. If that part is also obsolete, the honest recommendation is a targeted redesign of that block rather than a like-for-like clone.
Related board types raise the same questions in different clothing — heavy copper shows up constantly in automotive board reverse engineering and in industrial instrumentation board recovery, and the cost drivers carry across.
On legitimacy: this work exists to keep equipment running. Recovering documentation for a board whose design files are gone, keeping an out-of-production machine serviceable, replacing an obsolete part, doing failure analysis on a board that burned. You’re responsible for holding the rights to the design you send us, and every project runs under NDA.
FAQ
Can you tell copper weight from photos alone?
No — we can guess, but we won’t quote a stackup on a guess. Photos of heavy copper often show visibly raised traces and generous fillets, which tells us we’re in 3 oz-plus territory, but the actual number comes from cross-sectioning or weighing a coupon once the board is in the lab. Board thickness measured with calipers helps narrow the estimate.
Do you need to destroy my board?
For inner heavy copper layers, usually yes on at least one sample. Two-layer thick copper boards can often be documented non-destructively. Four layers and up with heavy planes need layer separation, which is destructive. Send two or
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