Best way of checking connections

[AGmurdercore]

Hey guys, I am assisting a friend with his keyboard, which has a full column not working - F2, 3, E, D, X, and Space. The keyboard itself is Redragon Indrah with Outemu Blue switches. It doesn't look like anything has been spilled on it and there are no visible breaks in the connections. The switches are working, as well as the diodes, I have checked them with multimeter.

The question i have is what is the best way to check from where the problem is coming so that I can fix/handwire the column? I would like to try hardwiring the column right away to see if this will help, but I am not sure which pin is connected to the column and which to the row. Any tips/tricks are appreciated!

Photo of the PCB here

[nevin]

1st check continuity between the switch leg (leg closest to diode, you can see the trace connecting them) & the diode on the affected switches
2nd check continuity between the other leg and the switch above/below (non diode leg).
- not sure what the matrix looks like on this board but it's probably the 1st or last in that column with an issue. simple jumper should fix it. i've had this happen a couple times over the years. no obvious trace lift/break, but obvious connection loss.
- also look around the controller and make sure you don't see anything obvious.

[AGmurdercore]

1st check continuity between the switch leg (leg closest to diode, you can see the trace connecting them) & the diode on the affected switches
2nd check continuity between the other leg and the switch above/below (non diode leg).
- not sure what the matrix looks like on this board but it's probably the 1st or last in that column with an issue. simple jumper should fix it. i've had this happen a couple times over the years. no obvious trace lift/break, but obvious connection loss.
- also look around the controller and make sure you don't see anything obvious.

There are no obvious breaks anywhere on the bottom of the PCB, I can't look at the other side at the moment (obviously since the switches are still here) but what bothers me is that there are no direct connections to switches above/below one another. Tested the connections between a few random switches, some that I know are working, and some that i know are not, none of them seem to have direct connection to the switch below/above on the leg without diode soldered next to it. It is bugging me so much that I consider desoldering all switches only to take a look at the other side of the pcb and figure what is going on. Would be helpful if I can avoid that on the other hand so all suggestions are welcome.

[nevin]

that's a full size right? who knows how they wired the matrix. probably some silly/crazy stuff. especially if it has full/faked NKRO.

- yes, looks like the column traces are on the other side of the pcb. but i've had a couple trace issues that you couldn't see but were obvious connection issues and fixed with a simple jumper

- possible the common connection is in rows instead of column?
    - check continuity between non diode switch legs in the row (instead of column). there should be a common connection between switches somewhere.

- reflow any solder point that is not well domed. the dome around the switch leg should not sink in. just means there's not a lot of solder in that hole.

after reflowing, check board with computer again.

- larger higher rez image of the entire pcb?

[AGmurdercore]

- possible the common connection is in rows instead of column?
    - check continuity between non diode switch legs in the row (instead of column). there should be a common connection between switches somewhere.

You were right! The non-diode legs are connected in rows, no idea why I didn't think of checking that initially, but thank you for the suggestion!

The keyboard is full size, yes, as for a better image I will take to take one later on and post it as I am unable to do so right now.

Also the joints look ok-ish, there is too much solder on some but the connection should be working regardless of the inconsistency in the quality.

[nevin]

Also the joints look ok-ish, there is too much solder on some but the connection should be working regardless of the inconsistency in the quality.

reflow them anyway. could be as simple as that, because the common traces are on the other side of the pcb. always better to have too than not enough, especially when the connection is on the other side of the pcb (other side of the hole).

once you do that & check again, you should be able to find where the lost connection is and then we can make a bridge/jumper of needed.

[AGmurdercore]

I am starting to feel a little desperate. I had some time and checked the connections on the faulty column, there are connections from diode-end to diode-end and also there is connection from each of the diodes to the controller. I am attaching some more pictures of the board and I am open for any suggestions on what might be the issue here and how to resolve it.

1, 2, 3, 4, 5, 6

[AGmurdercore]

Just tester the keyboard without actually doing anything to it, it's working. Will re-heat some joints now and test again after that for a few days to see how it will handle, but i am baffled with what is going on

[suicidal_orange]

Did you check for connections between the ends of the diode not connected to the switches?  From the problem I guess they would connect across the rows but in the pic it looks like they go up...  Whichever they don't do try shorting the non diode pin of a nearby switch to the diode pin of a dead one until you get the dead key pressed, then solder a jumper between that pad on the other switch and the non-diode pin on a dead one.  Hopefully one jumper will be enough.

If this makes no sense I'll fire up paint when I get home, it's not the easiest thing to discuss in text.

[AGmurdercore]

Did you check for connections between the ends of the diode not connected to the switches?  From the problem I guess they would connect across the rows but in the pic it looks like they go up...  Whichever they don't do try shorting the non diode pin of a nearby switch to the diode pin of a dead one until you get the dead key pressed, then solder a jumper between that pad on the other switch and the non-diode pin on a dead one.  Hopefully one jumper will be enough.

If this makes no sense I'll fire up paint when I get home, it's not the easiest thing to discuss in text.

A drawing would be welcome! The board seems to be working for now, I am typing from it. Will test it for a day or two to see it it will fail or something, but from the tests I did with the multimeter everything was looking ok.