Why “Just Scan It” Is Never the Full Answer
If you search “how to clone a PCB board,” most results suggest scanning the board, importing the image into PCB software, and re-drawing the traces. That workflow captures maybe 30 % of what a real clone requires. A flatbed or optical scanner sees the top copper and silkscreen—nothing more. It cannot tell you what sits on inner layers, what firmware runs inside the MCU, or whether a passive is 100 nF or 10 nF. A professional shop treats scanning as step one of a much longer chain.
This article walks through the complete process, explains where automated tools hit their limits, and shows which steps demand human engineering judgment. Whether you are evaluating vendors or considering a DIY attempt, you will know exactly what separates a photograph from a verified 1:1 electrically equivalent clone.
How to Clone a PCB Board — The 10-Step Professional Workflow

The steps below follow the exact sequence a professional reverse-engineering lab uses. Skipping or reordering them almost always leads to rework. Each step builds on the data produced by the previous one, so the chain is only as strong as its weakest link.
Step 1: Incoming Inspection and Photography
Before anything is disassembled, the original board is documented. A shop photographs both sides at high resolution, records every label, date code, and revision mark, and logs the board dimensions with calipers. This documentation serves two purposes:
- Traceability — if a dispute arises later, the original state is on record.
- Reference during assembly — component orientation, conformal coating areas, and mechanical keep-outs are easier to read from a photo than from a Gerber file.
At this stage the engineer also decides whether the board can realistically be cloned given its type and complexity. Rigid-flex, ceramic substrates, and boards with custom ASICs all require different tooling and may change the quote significantly.
Worked Example — Incoming Inspection of an Industrial Motor Controller
Imagine a 6-layer motor-controller board measuring 120 mm × 80 mm. During incoming inspection the engineer notes: a conformal coating on the bottom side, two BGA packages (one MCU, one FPGA), three custom-wound inductors, and a date code of 2017. The conformal coating must be chemically stripped before desoldering. The 2017 date code warns that some passives may now be obsolete. These observations go straight into the project log and shape every subsequent step.
Step 2: Component Removal and BOM Creation
Every component is desoldered, catalogued, and stored. The bill of materials (BOM) is built from physical markings, cross-referenced against distributor databases. This is where a scanner falls flat: a 0402 capacitor looks identical to a 0402 resistor in a photograph. Only a meter reading or a part-number lookup reveals the truth.
Common BOM Challenges
| Issue | Why It Matters | How a Shop Handles It |
|---|---|---|
| Sanded or missing markings | Part value unknown | LCR meter, X-ray, or circuit analysis |
| Obsolete components | No stock available | Cross-reference to compatible substitute |
| Custom-wound magnetics | No datasheet exists | Turns-ratio measurement, core material ID |
| Protected ICs | Firmware locked inside | chip decryption feasibility assessment |
A complete BOM is the backbone of the clone. If even one value is wrong, the finished board may power up but behave erratically—or not at all.
Worked Example — Identifying a Sanded IC
On the motor-controller board, the main gate-driver IC has its markings sanded off. The engineer measures the pin count (48-pin LQFP), checks the surrounding circuit topology (bootstrap capacitors, high-side and low-side gate outputs), and narrows candidates to three gate-driver families. A curve-tracer test on the I/O pins confirms the device is an IR2136 equivalent. Without this detective work, the BOM would carry a blank entry and the clone would stall at assembly.
Step 3: Layer-by-Layer Imaging
With components removed, the bare board is scanned optically. For a two-layer board, top and bottom copper are captured with a calibrated high-DPI scanner or a CCD camera rig. For multilayer boards, the process is far more involved.
Delayering Multilayer Boards
Inner layers are invisible from the outside. To expose them, the shop uses one or more techniques:
- Chemical etching — selectively dissolving copper and dielectric to reveal the next layer.
- Mechanical milling — CNC-controlled surface removal, measured in microns.
- X-ray imaging — non-destructive, but lower resolution; useful for via mapping before destructive steps.
Each exposed layer is photographed, aligned to the previous layer using fiducial holes or vias, and then digitised. A 12-layer board means at least 12 aligned image sets, each requiring manual cleanup where etching was uneven. This is the most time-consuming phase and the one that images alone can never fully replace.
How Long Does Delayering Take?
As a rough guide, a skilled technician can delayer and photograph one layer per half-day for a board up to 200 mm × 150 mm. A 6-layer board therefore needs about three working days just for imaging, not counting digitisation. Boards with blind or buried vias add extra steps because each via type must be mapped to the correct layer pair. HDI boards with stacked microvias can double the total imaging time.
Step 4: Gerber and Netlist Extraction
The aligned images are imported into CAD software—typically Altium Designer, PADS, or a proprietary tool—and redrawn as proper copper artwork. Pad geometries, trace widths, clearances, and drill sizes are all measured and replicated.
Once the layout is complete, the software exports:
- Gerber files — the manufacturing artwork for each copper, mask, and silkscreen layer.
- Drill files — locations and diameters for every through-hole, via, and slot.
- Netlist — the logical connectivity map that says “pin 3 of U5 connects to pin 7 of R12.”
The netlist is critical. Without it, you have a pretty picture but no way to verify electrical correctness. The engineer compares the extracted netlist against the physical board using continuity tests on the original. Any mismatch flags a potential redraw error. For a deeper look at this conversion, see the guide on recovering a schematic from a recovered netlist.
Worked Example — Catching a Netlist Error
During netlist verification of the motor-controller clone, the engineer finds that net GND_SENSE on pin 14 of the current-sense amplifier shows no continuity to the Kelvin ground pad on the shunt resistor. Reviewing the layer-3 image reveals a hairline trace obscured by uneven etching. The trace is redrawn in CAD, the netlist re-exported, and the continuity test now passes. Without this check, the clone would have had a floating sense input—causing erratic current readings that might not surface until the motor was under full load.
Step 5: Schematic Reconstruction
Many buyers assume a clone means “just the Gerber files.” But a schematic is what makes the clone maintainable. It groups components into functional blocks—power supply, processor core, analog front-end—so that future engineers can troubleshoot or modify the design.
The schematic is drawn from the netlist, then annotated with component values from the BOM. Power rails are traced to identify voltage domains. Decoupling capacitors are associated with their ICs. Signal names are assigned based on datasheet pin functions. This step is pure engineering work; no scanner or AI tool does it automatically with production-grade reliability.
For a broader view of the entire reverse-engineering pipeline, the full workflow guide on our homepage breaks the process into its major phases.
Step 6: Firmware and Programmable-Device Recovery
A PCB clone is electrically dead without the software that runs on its ICs. Microcontrollers, FPGAs, CPLDs, and EEPROMs all contain code or configuration data that must be extracted.
MCU Firmware
If the microcontroller’s read-protection fuse is not set, firmware can be read out through JTAG, SWD, or a standard programmer. If the fuse is set, the shop must evaluate whether a protected microcontroller firmware recovery technique exists for that specific device family. Success rates vary widely by chip vendor and revision.
FPGA Bitstreams
FPGAs add another layer of complexity. The configuration bitstream may be stored in an external SPI flash (relatively easy to dump) or encrypted inside the device. Encrypted bitstreams require FPGA bitstream recovery techniques with varying odds of success depending on the device family and security features.
Serial EEPROMs and Flash
Calibration data, MAC addresses, and configuration tables often live in small serial memories. These are straightforward to read but easy to overlook. Missing calibration data can make a cloned board fail functional tests even though the hardware is perfect.
Step 7: Stackup and Material Specification
A Gerber file tells the fabricator where to put copper, but not what to build it on. The shop must specify:
- Dielectric material — FR-4, Rogers, polyimide, or a hybrid stackup.
- Layer count and copper weights — 1 oz, 2 oz, or heavier for power planes.
- Impedance targets — especially for high-speed differential pairs or RF traces.
- Surface finish — HASL, ENIG, OSP, or immersion silver.
For high-frequency or HDI boards, getting the stackup wrong can shift impedance by 10–15 %, enough to cause signal-integrity failures. A power supply board case study illustrates how magnetics and stackup interact in a real clone project.
How Stackup Affects Impedance — A Quick Calculation
Consider a 50 Ω single-ended microstrip on standard FR-4 (Dk = 4.3). If the dielectric height between the trace and the reference plane is 0.2 mm and the trace width is 0.30 mm, the impedance lands near 50 Ω. Switch to a cheaper laminate with Dk = 4.8 and the same geometry yields roughly 47 Ω — a 6 % shift. Widen the tolerance on dielectric thickness by ±0.05 mm and you could see impedance swing from 44 Ω to 53 Ω. For USB 2.0 this might be tolerable; for PCIe Gen 3 differential pairs it is not. The stackup specification must therefore include both material grade and thickness tolerances.
Step 8: Prototype Fabrication and Assembly

The Gerber files, drill files, and stackup specification go to a PCB fabricator. Bare boards come back, and components from the BOM are assembled—usually by pick-and-place machine for SMD parts and hand soldering for through-hole or odd-form components.
Firmware is programmed into each IC. Configuration files are loaded into EEPROMs. The assembled board is then visually inspected under magnification and, for BGA packages, X-rayed to confirm solder joint quality.
Step 9: Functional Validation
This is the step that separates a professional clone from a hopeful copy. The prototype is powered up and tested against the original board’s behaviour:
- Power-rail verification — every voltage domain measured under load.
- Signal-integrity checks — oscilloscope probes on critical buses (SPI, I²C, LVDS, USB).
- Functional test — the board is placed in its host system (or a test jig) and exercised through its normal operating modes.
- Thermal imaging — hot spots compared to the original to catch layout or component errors.
- Boundary conditions — tested at voltage and temperature extremes if the application demands it.
If discrepancies appear, the engineer traces the fault back through the schematic, layout, or BOM and corrects it. Sometimes a second prototype spin is needed. The goal is electrical equivalence, not just visual similarity.
Step 10: Production File Package
Once the prototype passes validation, the shop delivers a complete file package:
- Gerber and drill files (manufacturing-ready)
- Schematic (PDF and native CAD format)
- BOM with manufacturer part numbers and approved alternates
- Pick-and-place centroid file
- Firmware binaries and programming instructions
- Stackup drawing and impedance requirements
- Test procedure (if contracted)
This package lets any competent contract manufacturer replicate the board at volume. It is the deliverable that turns a single working prototype into a manufacturable file set ready for production.
What a Scanner Actually Contributes — and Where It Stops
To be clear, scanning is not useless. It is the fastest way to capture outer-layer geometry, and modern software can auto-trace simple two-layer boards with reasonable accuracy. But here is what a scanner cannot do:
| Capability | Scanner | Professional Clone Shop |
|---|---|---|
| Outer copper imaging | Yes | Yes |
| Inner layer recovery | No | Yes (delayering) |
| Component identification | Partial (silkscreen only) | Full (meter + database) |
| Netlist extraction | No | Yes |
| Schematic reconstruction | No | Yes |
| Firmware recovery | No | Yes (where feasible) |
| Stackup specification | No | Yes |
| Functional validation | No | Yes |
If your board is a simple two-layer design with no programmed ICs, a scan-and-redraw approach may work. For anything more complex, the scanner is just the starting line.
When Cloning Leads to Redesign
Sometimes a straight clone is not the best outcome. Obsolete connectors, discontinued memory chips, or outdated bus interfaces may force partial redesign. A good shop will flag these issues early and offer a path to modernise the board without losing its original behaviour. The clone becomes the baseline; the redesign updates only what needs changing.
Protecting Your Own Boards from Cloning
Understanding how cloning works also helps if you are on the defensive side. Knowing each step reveals which countermeasures are effective and which are merely inconvenient. A separate guide covers twelve measures ranked by how well they resist reverse engineering.
Quick Self-Assessment: Is Your Board Ready to Clone?
Before contacting a shop, answer these questions to speed up the quoting process. The interactive checker below estimates your project’s complexity tier based on your answers:
- Do you have a working sample, or only a dead board?
- How many layers does the board have (or your best guess)?
- Are there any programmable ICs (MCU, FPGA, CPLD, EEPROM)?
- Do you need just the Gerber files, or a full schematic and BOM?
- Is the goal a one-off replacement or volume production?
- Are any components visibly obsolete or custom?
Having clear answers to these six questions helps a shop return an accurate quote in days instead of weeks.
Frequently Asked Questions About PCB Board Cloning
How much does it cost to clone a PCB board?
Cost depends primarily on layer count, board size, and the number of programmable ICs. A simple 2-layer board with no firmware might cost a few hundred dollars. A 10-layer board with multiple protected MCUs, BGA packages, and impedance-controlled stackup can run into several thousand. The biggest cost drivers are delayering time (each additional layer pair adds roughly half a day of lab work) and firmware recovery (chip decryption fees vary by device family). Request a quote with your board in hand — photos of both sides and a rough layer-count estimate let a shop price the job accurately.
Can I clone a PCB board at home without professional equipment?
For a 2-layer through-hole board with no programmed ICs, a motivated hobbyist with a flatbed scanner, KiCad, and a multimeter can produce a usable clone. You scan both sides at 1200 DPI, import the images as board outlines, redraw traces, and verify continuity against the original. The moment you encounter multilayer construction, BGA footprints, or read-protected microcontrollers, home equipment hits a wall. Delayering requires controlled chemical baths or precision CNC milling, and firmware extraction from a locked MCU demands specialised silicon-level tools. For boards beyond two layers, professional services save more time and money than they cost.
How long does the entire PCB cloning process take?
A typical 4-layer board with moderate component density and unprotected firmware takes two to three weeks from sample receipt to validated prototype. The timeline breaks down roughly as follows: incoming inspection and BOM (2–3 days), delayering and imaging (3–4 days), CAD redraw and netlist verification (3–5 days), firmware extraction (1–3 days), fabrication and assembly (5–7 days), and functional validation (1–2 days). Boards with more layers, protected ICs, or exotic materials can stretch the schedule to six weeks or more. Rush services are available at most shops but typically carry a 30–50 % surcharge.
Is it legal to clone a PCB board?
Legality depends on jurisdiction and intent. In most countries, reverse-engineering a product you own for interoperability, repair, or replacement-part manufacturing is permitted under fair-use or right-to-repair provisions. Cloning a competitor’s patented design for commercial sale is a different matter entirely and may infringe patents, copyrights on firmware, or trade-dress protections. Always consult an intellectual-property attorney before producing cloned boards for resale. The technical ability to clone a board does not automatically grant the legal right to do so.
Final Thought
Learning how to clone a PCB board is really learning the full reverse-engineering chain: photography, delayering, BOM extraction, netlist verification, firmware recovery, stackup specification, and functional validation. Each link in that chain requires different equipment and different expertise. A scanner handles the first link. A professional lab handles all of them—and proves the result works before handing you the files.
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