Telecom PCB copy means taking a physical telecom board — a line card, a backplane, an RF power amplifier, a timing card, a DSLAM or E1/T1 interface module — and rebuilding the manufacturing and design files from the board itself. You get Gerbers, drill data, a stackup, a BOM, and optionally a schematic and netlist, so you can rebuild spares for equipment whose original files no longer exist. That’s the whole job.
Most people who need this are in one of two spots: a network still running gear the OEM discontinued years ago, or a product owner whose layout files walked out the door with a contractor or died on an unbacked workstation.
Why telecom boards end up needing this at all
Telecom hardware has an unusually long service life. A chassis installed in a central office or a remote cabinet can stay in service well past the point where the vendor’s support contract, the CAD files, and sometimes the vendor itself are gone. Meanwhile the boards inside keep failing at the rate electrolytic capacitors and hot power stages always fail.
Two things then collide. Spares dry up on the broker market, and the ones you do find are the exact parts most likely to be relabeled counterfeits. And the semiconductors on the original BOM — a specific framer, a specific line driver, a specific clock recovery part — have gone end-of-life, so even the OEM couldn’t build the board again without changes.
That’s the moment a telecom board copy stops being an academic exercise. You’re not trying to compete with anyone. You’re trying to keep a working network on the air for another five or ten years.
The cases we see most often
- Line cards and interface modules. Dense, four to twelve layers, lots of magnetics, transformers and protection components. Usually the most straightforward telecom boards to reproduce.
- Backplanes. Physically large, often thick, high layer count, mostly connectors and controlled-impedance differential pairs. Geometrically simple, electrically fussy.
- RF and power amplifier boards. Microstrip on PTFE or ceramic-filled laminate, where the copper geometry is the circuit. These need dielectric measurement, not just trace tracing.
- Power and rectifier boards. Heavy copper, wide planes, high-current paths where copper weight and thermal relief matter more than trace routing.
- Timing, sync and clock distribution cards. Sensitive to layout details you can’t see on a photograph — plane splits, guard traces, via placement under oscillators.
What actually happens to your telecom board on the bench
Here’s the order of work, so you know what you’re buying rather than reading a promise about quality.
- Intake and documentation. High-resolution scans of both sides before anything is touched, plus overall dimensions, board thickness measured with a micrometer at several points, connector positions, and mounting hole locations. If your board is a failed unit, we photograph the damage first — burnt areas destroy information, and we need to know what was already gone.
- Component identification. Every marking is read and cross-referenced. Telecom boards are full of house-numbered parts and custom magnetics, so this step is where honest work separates from guesswork. Anything ambiguous gets measured or curve-traced, and anything we can’t confirm is flagged as unconfirmed rather than filled in.
- X-ray, if there’s anything hidden. BGA-mounted framers, network processors and FPGAs on modern line cards need X-ray to see ball patterns and buried via structures before teardown.
- Desoldering and cleaning. Components come off, solder is removed, the bare board gets cleaned so the copper is legible under the soldermask.
- Layer separation and imaging. The stackup is taken apart layer by layer and each copper layer is imaged individually. On a twelve-layer backplane that’s twelve separate imaging and alignment passes, which is exactly why layer count drives cost more than board size does.
- Vectorizing and rebuilding. Images become real copper geometry with real pad stacks and drill data, aligned to a common origin. Trace widths, spacings and annular rings are recorded as measured, not as a designer would have drawn them.
- Verification. The rebuilt netlist is checked against continuity measured on the original board, connector pin by connector pin. Rebuilt Gerbers are compared against the intake scans. Layer-to-layer registration is checked against the drill file.
- Optional prototype and test. We build bare boards, assemble a unit, and compare it against your known-good card in your equipment or on a fixture you specify.
If you want the generic version of this workflow without the telecom specifics, the step-by-step walkthrough of the copy process covers it in more depth.
Impedance, dielectrics and the part telecom makes hard
On a consumer board, tracing copper is most of the work. On telecom hardware, the copper is only half the answer. A 100-ohm differential pair on a backplane depends on trace width, spacing, dielectric thickness and the Dk of the laminate. Copy the copper onto the wrong material and the geometry is right while the impedance is wrong.
So we measure the dielectric layers, record copper weights per layer, identify the laminate family where we can from construction and glass style, and reproduce the stackup as a specification rather than a guess. Where the original used a low-loss material for an RF section, that gets called out explicitly. Where we can only narrow it to a class of material, we say that too, because you’d rather know than find out at test.
The same logic applies to RF sections. Microstrip filters, couplers and matching networks are distributed structures. We reproduce the etched geometry to measurement and note the substrate assumptions, and we recommend a prototype build with network-analyzer verification before you commit to a production run. For controlled-impedance work, design-rule practice from IPC-2221 and IPC-2222 and the current-capacity curves in IPC-2152 are what we sanity-check the reconstructed geometry against; board acceptability at fabrication is judged against IPC-A-600 and IPC-6012 class levels, and telecom infrastructure boards usually sit at Class 2 or Class 3 depending on the customer.
Which files you get, and how to choose the tier
Not every project needs everything. Deciding this up front saves real money.
| What you need to do | Deliverable set | Notes |
|---|---|---|
| Just build more of the same board | Gerbers, drill files, stackup, netlist, BOM, pick-and-place | Fastest and cheapest path. No schematic needed. |
| Replace an obsolete framer, PHY or regulator | Above plus schematic | You can’t substitute a part intelligently without seeing the circuit. |
| Change form factor or connector | Above plus editable native CAD | Re-layout needs a real design database, not just Gerbers. |
| Debug an intermittent failure in the field | Schematic plus netlist | Often no board build at all. Documentation is the product. |
| Assembled spares, not files | Files plus assembly and functional test | Worth reading up on where bare-board copy ends and assembly copy begins. |
A common question is whether the schematic is worth paying for on a telecom card. If the board’s job is fixed and the parts are still available, no. If you expect to substitute components over the next several years — which on discontinued telecom silicon you almost certainly will — yes, and it’s cheaper to capture it while the board is already apart than to come back later.
What to send us, and what comes back
To get a real feasibility answer and a real number instead of a range, send this:
- Clear photos of both sides of the board, whole board in frame, laid on a dark non-reflective background, shot straight down with even light. No angled phone shots through a chassis window.
- Board dimensions and thickness, plus layer count if you know it.
- Close-ups of the main IC markings — the framer, processor, FPGA, CPLD or MCU — including any house numbers.
- A photo of the connector edge and any shielded or potted areas.
- One line on what you actually need: files only, files plus prototypes, or production spares.
- Whether any programmable device on the board holds firmware you need recovered.
What comes back is a feasibility assessment naming the specific risks on your board, a lead time, a deliverable list, and a price. If something on the board looks like it will be a problem — a potted module, a laser-marked custom ASIC, a badly burnt section — you hear about it before you pay, not after.
Where firmware fits in
Many telecom cards carry an MCU or CPLD holding configuration and control code. If the hex file is lost, copying the hardware alone gives you a board that boots to nothing. Read protection and lock-bit state determine whether code readout is feasible at all, and it varies enormously by part family — an old 8051-class device with a simple security fuse is a different proposition from a modern ARM part at its highest read-protection level. We evaluate that separately and tell you honestly which category yours falls into. Never assume it’s a given.
Honest limits on telecom board copy work
Things that make a telecom PCB copy harder, slower, or occasionally not worth attempting:
- Severe burn damage. Carbonized laminate has destroyed the copper it used to carry. We can often infer the missing net from symmetry and function, but on a unique section we may not be able to. Send a second, working board if you have one.
- Potted or conformally coated assemblies. Removal is possible but risks the board, and it adds time. Two samples helps.
- House-numbered custom silicon. If the main device is an OEM-specific ASIC with no commercial equivalent, we can reproduce the board but nobody can supply the part. That usually turns the project into a redesign conversation instead.
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 quoteRelated reading