Tips for reading old hand-routed PCBs that nobody talks about

Oct 10, 2026 Last reply: 4 days ago 4 Replies
Jump to replies (4)

Spent years repairing old industrial gear, so let me explain those wiggly traces you see on boards from the 80s and 90s. They were drawn by hand. Literally. The layout person taped black crepe tape onto big sheets of mylar at 2x or 4x scale, and cutting curved runs with a sharp blade was often faster than making neat right angles. Every bend was a seperate little patch of tape, so curves saved time.



The other reason is that 90-degree corners were genuinely frowned upon back then. Old etching processes coud underetch the inside of a sharp corner and leave a sliver of copper, so the rule was 45 degrees or curves. Your straight-traces-with-right-angle-bends board from the same era is actualy the odd one for its time. If you ever need to trace out one of these boards, here is what worked for me. Put the board on a lightbox or hold it up to a window so the copper shows thru, then follow one net at a time with a colored pencil on a photocopy. Curves are friendlier than they look, they almost always take the shortest path around an obstacle.



Last tip, count the solder joints on the underside before desoldering anything. Hand-assembled boards often have reworked joints that never matched the drawing, and that will send you chasing ghosts for hours.


This matches what I found fixing arcade boards through the 90s. I spent a whole winter tracing a bootleg board where every net wandered in long lazy arcs, and it finally clicked when I saw the tape layout artwork in a binder the shop owner kept. Curves everywhere because someone was cutting tape by hand with a swivel knife, and the corners cost them patches and patience. The colored pencil trick works, but my version was a highlighter on a photocopy, one net per pass, marking both sides of the board so I could follow vias through. I got burned once chasing what looked like a direct connection between two chips, turned out a trace ran under a socket leg and the drawing was wrong anyway. Also watch for cut traces that were reconnected with jumper wire on the solder side. On those old boards a repair like that is common, and it will make your traced schematic lie to you unless you flip the board and check every joint as you go.

I'd push back on the underetching bit though. I worked summers at a small fab in the late 90s and we ran right-angle corners all day without slivers of copper. That acid trap story was mostly shop folklore that got repeated untill everyone believed it. Pull datasheets or app notes from big manufacturers from that era, plenty of thier reference boards have crisp 90s. The tape layout explanation holds up fine on its own, no need to dress it up with etching chemistry that was never really the problem.

That said, the tip about reworked joints saved me more than once, so no arguement there.

the scale thing trips people up too. when i apprenticed at a controls shop in the late 80s we did artwork at 4x, so a trace drawn a sixteenth of an inch wide on the mylar came out about 15 mil on copper. folks measuring the tape art and guessing at current capacity were always off by half. also the tape came in set widths, like a little dispenser rack of them on the drafting table, wich is why every trace on a board has the same thickness no matter how lazy the curve is. the layout lady there could cut a curve freehand faster than i could draw one on paper.

Update: the fab worker calling the acid trap thing folklore has me second-guessing myself, because I heard it from an old-timer and just ran with it. Curious if anyone here has actually seen a sliver like that on a scrapped board, or if I've been repeating the same shop story. Also going back and re-measuring some tape widths with the 4x scale thing in mind, since my current ratings were probably way off.

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required