Help debugging why random keys won't work on an old pcb

[SpacemanToby]

Hey guys.

Does anyone have some tips or links to reading on how to debug old pcbs? I got a video writer the other day, and I wanted to replace the alps brown switches with greens so the unit could remain operational. After replacing them, a couple won't work. I've tested all the switches, and a couple of them I didn't even replace. The keys with a red box are ones that I didn't replace and stopped working and the ones with green boxes are the ones i did replace. The squares are the only keys not working. Sorry if what I'm saying is confusing.

[rickard.von.essen]

The first thing to do is to check that pushing the closes the circuit and opens when releasing it. Use a multimeter.

Sent from my SM-G935F using Tapatalk

[SpacemanToby]

That is is one thing I have done, and my multimeter beeps on the continuity setting when I press the key down.

[snacksthecat]

Do we know any info about the signalling for this keyboard? Might help in debugging issues.

Also secretly interested because I just got one for myself

Edit: Another thing to check; is there a common trace between the affected keys?

[SpacemanToby]

I've tried briefly to follow the trace, but that is tough to do. I'll have to spend more time on it.

[Sangdrax]

My bet is it's a double-sided PCB and you borked one of the topside connections when desoldering. Just seems likely you didn't apply enough heat when desoldering when I can see several lifted pads and jumpers on the bottom side and cold joints on the replacements.

Also, fix those cold joints on 60, 65 and 71. May not be the problem but if you use the thing, they'll give out after awhile.

[SpacemanToby]

Thanks for taking a look! 60, 65, and 71 look bad because alps mashes down the pins on some of the switches, and I accidentally messed up the solder pads that connect to the trace during the de-soldering process. I i had to expose some of the trace copper and wired the pin directly to the trace. Those keys actually work just fine. I don't doubt I've screwed something up though. But some of the switches not responding I didn't even touch.

[Sangdrax]

I can see several physical topside wire jumpers just looking at the bottom of the board. I imagine there's a couple of actual traces for the really long stuff. That's where I think the broken connection is.

But unfortunately like the snacks said, I think you're going to have to break out a graphics program and a multimeter in continuity mode to map the matrix and find the break.

What I mean by cold joint is you can tell the wire on your jumpers wasn't heated quite enough to adhere right or didn't use enough flux. Just give em another once over with some flux and the iron and they should be as strong as the regular solder pads.

[SpacemanToby]

Anyone have a good guide on how to map out keyboard matrixes? I've done a lot of handwiring, but I have next to no experience with pcbs.

[Sangdrax]

Take a picture of the PCB. Mark every switch contact with a little circle or something. Then overlay colored lines over the top of traces and follow them out one at a time. I usually assign every row a different color and make columns a single color. You'll need the multimeter for when the traces hop to the other side of the board or a jumper is used. It will probably take a few hours of work.

In your case I would mark the problem switches and just map out those rows and columns by themselves. Should be a lot easier work than the whole enchilada.

[SpacemanToby]

Thanks for the help! I was able to figure out that all the problem keys were in the same column, and there was one with a bad solder. I resoldered that pin and it works again! Such a cool machine, I'm glad I didn't break it.

[Sangdrax]

Congrats! It's a cool machine. Actually, as a linear fan, I'd say you even improved it.

[green-squid]

Hey, glad your problem was fixed! There is something fascinating about these big word processors, and their weird place between computer and typewriter.

[SpacemanToby]

It's definitely a hit at the office. Many memos are getting printed and distributed. Also it's fun to feed different kind of paper through and try to line up what you're printing. I prints surprisingly well on paper towels.

[Hak Foo]

Mount a roll of Brawny behind the unit and it;s like an old dot-matrix printer with continuous-feed tractor paper!