A single-sided PCB copy is usually the fastest board duplication job in the shop — one copper layer, one imaging pass, no layer separation, no X-ray needed for buried vias. But “fast” is not the same as “trivial.” The hard parts on a single-layer board are almost never the copper. They’re the wire jumpers, the sanded-off part markings, the through-hole electrolytics nobody makes anymore, and the fact that these boards are typically 20 to 40 years old.
If you’re holding a dead controller board from a machine that still has ten years of service life left, this page tells you what we do with it and what could go wrong.
What a single-sided board actually is, and why it matters to the copy
Copper on one face only, components on the other, holes drilled straight through. Almost always FR-1, FR-2, CEM-1, or paper phenolic rather than FR-4 — cheap, single-clad, punched instead of routed. You see them in appliance control boards, LED drivers, power supplies, older thermostats, motor speed controllers, ballasts, and a huge amount of 1980s and 1990s industrial gear.
Two consequences for a copy job. First, the routing has to snake around every component lead, so traces are wide, curvy, and often hand-drawn in the original artwork. Second, when the layout ran out of room, the designer dropped in a wire jumper. Those jumpers are part of the netlist. Miss one and the board doesn’t work.
The material substitution question comes up immediately
Most fabs today don’t want to run paper phenolic in small quantities. Copying a phenolic single-sided board onto single-clad FR-4 is normally an improvement — better dielectric strength, better mechanical stability, no charring around hot resistors. But it changes two things you have to check: board thickness (a lot of these boards are 1.2 mm or 1.5 mm, not 1.6 mm, and the enclosure card guides care) and dimensional behavior at the punched edges.
If the board sat in a fan-cooled cabinet at 60 °C for two decades on paper phenolic, tell us. If it lived on a heatsink, tell us that too — some single-sided LED and driver boards are actually aluminum-backed, which is a different process entirely.
What happens to your board in the lab, in order
- Intake and photography. Both sides, high resolution, flat lighting, plus close-ups of every marked component. Nothing is removed yet. If the board came in cracked or with lifted pads, that gets documented now so nobody argues later about who broke what.
- Jumper and modification survey. Every wire link, every field rework, every cut trace gets logged before desoldering, because once parts come off you can lose the evidence. Factory jumpers and someone’s 1997 field fix look identical until you look at solder appearance and wire type.
- Component identification and BOM capture. Values, tolerances, voltage ratings, package sizes, polarity, and manufacturer markings. Measured where markings are gone.
- Desolder and clean. Careful heat on old phenolic — pads on FR-2 lift much more easily than on FR-4. We work at lower iron temperature and longer dwell rather than blasting it.
- Scan and vectorize the copper. Solder mask and any silver-print carbon get read, the copper is scanned, and the artwork is redrawn as real vectors, not traced bitmap outlines. Pad shapes matter: a lot of single-sided designs use teardrop and oblong pads for hand-insertion tolerance.
- Drill and outline. Hole sizes measured with pin gauges, not guessed from the scan. Punched boards often have non-round mounting slots and V-notches that a router has to approximate.
- Netlist verification. The recovered netlist is checked back against the physical board with a continuity meter, point to point, including through every jumper. This is the step that catches the mistakes.
- Bare board and, if you want it, a stuffed sample. Then side-by-side electrical comparison against the original.
The verification step is what separates a usable board from a nice-looking picture. IPC-A-600 gives you the vocabulary for judging the bare board that comes back — annular ring, plating, mask registration — and it’s worth agreeing on a class up front rather than after the panel arrives.
Where single-sided jobs get genuinely hard
Nobody publishes this part, so here it is plainly.
Jumper wires that were never on the artwork. Production floors add links. If your only sample has a rework that the rest of the fleet doesn’t have, you’ll copy a one-off. Two or three samples solve this instantly and cost you nothing but shipping.
Sanded, painted, or lasered-off part markings. Common on the one custom IC or the one transistor the OEM wanted to protect. Passives can be measured. An unmarked 8-pin device can’t be, not from the outside. We can often pin it down from the circuit function, supply pins, and package, but “often” is not “always.”
Conformal coating and potting. Coating comes off with solvent and patience. Epoxy potting over a single-sided board is the worst-case scenario — the copper underneath is thin and the mechanical removal risk is real.
Obsolete through-hole parts. Single-sided boards are old, so their BOMs are full of dead part numbers: specific TO-92 transistors, axial electrolytics in odd can sizes, PTC thermistors, mains-rated X and Y capacitors with safety approvals. Substitutes exist, but a safety-critical cap has to be replaced with a properly approved equivalent, not just something with the same capacitance.
A locked microcontroller on the board. Very common on appliance and motor control boards. Copying the copper gets you a bare board; it does not get you the code. That’s a separate firmware question, and it’s the one that decides whether the project is a few days or a few weeks.
Single-sided compared with the boards you’d expect it to be compared with
| Aspect | Single-sided | Double-sided | Four layers and up |
|---|---|---|---|
| Copper imaging | One pass, no registration issues | Two passes, front-to-back alignment | Delamination or grinding per layer, careful registration |
| Main risk | Missed jumper wires, unidentifiable parts | Plated via mapping | Inner-layer damage, blind and buried vias |
| Typical vintage | 1980s–2000s, cost-driven | Any era | Modern, dense |
| Time driver | Component ID and BOM sourcing | Copper work plus BOM | Layer separation and imaging |
| Non-destructive? | Usually yes, board survives | Usually yes | Often destroys the sample |
That last row is the practical advantage. A single-layer board almost always comes back to you intact, which matters if it’s the only spare you have. The trade-off comparison is laid out in more detail in our breakdown of single-layer versus multilayer copy work, if you’re deciding how to handle a mixed set of boards from one machine.
What to send us, and what you get back
Send this and you’ll have a real answer instead of a range:
- Photos of both sides, whole board in frame, on a dark non-reflective background, camera square to the board. Then close-ups of any chip markings.
- Board dimensions and thickness, plus a note on the substrate color if you know it (brown/tan usually means phenolic or FR-2).
- The marking on the main IC — full line, including the date code, even if it looks like nonsense.
- One line on deliverables: Gerbers only, Gerbers plus BOM, schematic as well, or working assembled boards.
- Quantity and use case. Five repair spares and five thousand production units are different projects.
- Whether the board has a programmed MCU and whether you have the hex file.
Back from us: a feasibility read, a clear statement of anything we can’t identify from photos, a lead time, and a price. Standard handover for a single-layer job is Gerber set, drill file, board outline with slots and notches, silkscreen, a BOM with measured values, and the jumper list as an explicit document rather than something buried in the artwork. Our page on what the final file package contains spells out each file if you need to check it against your fab’s requirements.
Cost and time drivers on a one-layer board
Copper work is cheap here. What moves the number is component count, how many parts have illegible markings, whether you want a schematic on top of Gerbers, whether obsolete parts need sourcing, and whether firmware is in scope. Board size barely matters. A 300 mm × 100 mm single-sided board with 40 parts is easier than a 60 mm square one with 120.
Time follows the same logic. Copper and drill on a simple single-layer board is quick. Waiting on a substitute for a discontinued mains-rated capacitor is not. Firmware recovery, if needed, sits on its own schedule entirely.
The legal side, briefly
The legitimate uses are the ones we actually see: keeping out-of-production industrial equipment running, rebuilding a board whose design files vanished with a former supplier, migrating away from an obsolete part, failure analysis after a field return. You need to hold the rights to the design or the authorization to have it reproduced — that’s your call to make, and every project runs under NDA on our side. If you want the longer version, our explanation of what PCB copy covers goes into the distinction between duplicating your own board and duplicating someone else’s.
FAQ
Can you copy a single-sided PCB without removing the components?
Partly. Photos and a mask scan get us most of the copper on a single-layer board because it’s all
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