If your POS terminal board is out of production and the original design files are gone, a POS terminal PCB copy rebuilds a manufacturable file set — Gerbers, drill data, netlist, BOM and, when you need it, a schematic — from the physical board you still have. That covers the copper, the stackup and the parts list. It does not cover the firmware or the security keys inside the payment controller, and on a POS board that distinction matters more than on almost any other product.
This page is written for the person holding a dead or unsupportable terminal: a service engineer, a fleet owner with 400 units in the field, or a product manager whose contract manufacturer folded and took the CAD with it.
Why POS boards end up with no files at all
Payment terminals have an unusually short commercial life and an unusually long service life. A model gets certified, ships for three or four years, gets superseded, and then sits in retail lanes and fuel forecourts for another decade because replacing the whole estate means re-certifying the software and retraining staff.
By the time a spare board is needed, the design house has moved on, the layout engineer has left, and the ODM in Shenzhen that built it has changed names twice. What survives is a working unit and maybe a failed one. That’s enough to work from.
The other common trigger is a single obsolete part. A magnetic stripe read head amplifier, a specific SIM/SAM socket, an old GPRS module, a thermal printer controller — one end-of-life line item can stall production of an otherwise healthy board. Reconstructing the files gives you something you can actually edit and re-spin.
What a POS terminal board actually looks like under the hood
POS main boards are typically four to eight layers, FR-4, with fine-pitch parts and a mix of technologies packed into a small outline. When we plan a copy, these are the features that drive the work:
- An application processor or MCU under a BGA or QFN. Ball pitch is often 0.5 mm or 0.4 mm, which means X-ray is the only way to map the escape routing without destroying the part.
- A secure element or dedicated crypto controller. Sometimes a separate package, sometimes integrated. Either way it usually sits inside a tamper mesh.
- Tamper detection hardware. Mesh traces on outer layers, dome switches under the keypad, light sensors, sometimes a battery-backed tamper latch. These traces look like ordinary signals and are easy to mis-trace if you don’t know what they are.
- Keypad and display flex connectors. ZIF footprints with tight pin pitch and controlled-length runs to the LCD.
- Card reader front ends. Contact smart card slot, magnetic stripe head, contactless antenna loop — the antenna is usually a copper coil on the board or a mated coil FPC, and it is impedance and geometry sensitive.
- RF and comms. Cellular, Wi-Fi or Bluetooth modules with matching networks and keep-out zones, plus a printer and power section that runs at much higher current than the digital side.
That mix is why we treat these as high-density boards rather than routine two-sided work. The approach we use on telecom boards transfers well here, because both share the RF plus fine-pitch digital combination in a small footprint.
How the POS terminal PCB copy runs, step by step
1. Intake and non-destructive capture
The board is photographed, dimensioned and X-rayed before anything is removed. X-ray gives us via locations, buried structures, and BGA ball patterns while the assembly is still intact. We log every component marking, including the ones under shields, and note anything with a scratched-off or laser-blanked top mark.
2. De-assembly and bare board recovery
Components come off under controlled heat, sorted and bagged by reference so nothing gets confused later. The bare board is cleaned, solder mask and silkscreen are captured at high resolution, then the board goes to layer separation — inner layers are exposed one at a time and imaged individually.
3. Copper reconstruction
Each imaged layer is vectorized into real copper geometry: traces with widths, pads with real shapes, planes with their actual voids and thermal reliefs. We measure dielectric thicknesses and copper weights so the stackup in the output is the stackup that was built, not a guess. On the RF section and the contactless antenna, geometry gets copied dimension-for-dimension because a 5% change in a coil or a matching stub changes behavior.
4. Netlist and electrical verification
This is the step that separates a usable file set from a pretty picture. We extract a netlist from the reconstructed layers, then verify it against the physical board with continuity measurements — pad to pad, connector to connector, node by node on the nets that matter. Discrepancies get resolved by going back to the board, not by assuming the CAD is right.
5. Schematic and BOM
If you asked for a schematic, we build it from the verified netlist and the identified parts, organized by function — power, processor, card reader, keypad, comms, tamper — rather than as one giant sheet of nets. The BOM lists manufacturer part numbers where the marking allows it, and flags parts we could only identify by value and package.
6. Output and first article
Deliverables go out as Gerber RS-274X or ODB++, drill files, stackup documentation, netlist, BOM and pick-and-place. Board acceptability is judged against IPC-A-600 and IPC-6012 class expectations, and where we recalculate a trace for a rebuilt power section, IPC-2152 is the reference for current-carrying width. If you want proof before committing, we build a first article and functionally test it against your golden unit. Our standard deliverable list is spelled out here so there is no ambiguity at handover.
What to send us, and what comes back
Feasibility on a POS board is decided in minutes if you send the right things. Please include:
- Clear photos of both sides, on a dark background, whole board in frame. No cropping, no flash glare. A second set of close-ups on the processor and any shielded areas helps.
- Board dimensions — length, width, thickness — and layer count if you know it.
- The main chip markings, typed out as text, including the full suffix.
- Quantity and intent: spares only, a re-spin for an obsolete part, or a production run.
- One line on deliverables — Gerbers only, or Gerbers plus schematic plus BOM, or a full turnkey build.
What you get back from us: a feasibility statement that names anything we consider risky, a lead time, a price, and a list of what we would need physically shipped. Projects run under NDA, and we ask that you hold the rights to the design or have the owner’s authorization — POS hardware sits close to payment-brand IP and certification marks, so that’s not a formality.
Where POS terminal board duplication gets hard, and where it stops
Honest answer first: the copper and the parts list are almost always recoverable. The secure parts of the design frequently are not, and no amount of process will change that.
| Part of the board | How it usually goes |
|---|---|
| Power section, printer driver, keypad matrix | Straightforward. Well-marked parts, generous geometry, easy to verify. |
| Digital core, memory, fine-pitch BGA routing | Doable with X-ray plus layer separation. Adds time, not risk. |
| Contactless antenna and RF matching | Copy geometry exactly and it works. Redraw it loosely and it won’t. |
| Tamper mesh and tamper switch network | Traceable as copper, but the board will latch tamper on power-up and stay dead until the OEM’s authorized process clears it. |
| Secure element / crypto controller contents | Generally not recoverable. Keys are injected under an audited process and protected against readout by design. |
| Certified payment application and injected keys | Out of scope. These belong to the OEM and the payment scheme, not to the board. |
Three other things shift the answer. Potting or conformal coating over the secure area slows removal and sometimes damages what’s underneath. Blank-topped or relabeled chips mean identification by behavior and pinout instead of by marking, which takes longer and occasionally ends inconclusive. Missing a working reference unit is the biggest one — with only a dead board, we can reconstruct files but we cannot verify functionally, and you should treat the first article as a prototype rather than a drop-in spare.
Firmware on a POS board: what’s realistic
Non-secure firmware on peripheral controllers — a printer MCU, a display driver, a housekeeping 8-bit part — is a normal firmware recovery job, and feasibility depends on the specific part and how its lock bits, security fuse or read-out protection were configured. An MSP430 with the JTAG fuse blown behaves differently from an 8051 derivative with lock bits set, and an STM32 at RDP Level 1 is a different conversation from one at Level 2.
Anything inside the certified payment path is a different matter, and we don’t take those jobs. If your goal is to keep terminals running, the productive path is nearly always OEM-authorized firmware plus a reconstructed board, not an attempt to lift the payment application.
Copy as-is, or use the files to redesign?
For spares supporting an existing certified estate, copy as-is. Any change to the board can invalidate the certification you’re trying to preserve, so the goal is a board that is electrically and mechanically indistinguishable from the original.
For an obsolescence problem, the copy is a starting point. Once you have an editable stackup and netlist, swapping a dead-end regulator or a discontinued comms module is ordinary engineering work — footprint change, re-route, re-verify. Same logic applies if you want to shrink the board or move to a current MCU family. That’s where
Send the chip marking or two photos. You get feasibility, lead time and price within 24 hours, and the check costs nothing.Working on a board like this?
Related reading