[FIXED] Help - Both pads lifted. Where to wire?

[SuperBaseball]

I'm not new to soldering but I got a cheap soldering iron on amazon after my old Radioshack one corroded and I couldn't find replacement heads.  I must have been running it too hot, because I accidentally lifted both pads while desoldering/resoldering my R Shift while doing a full rebuild of my Tex Yoda II. I tried to bridge the switch with some wire, and I thought I was successful, however I realized that while R Shift was now working, the Key Up event was never firing.  This effectively lead to a caps lock that can only be turned off with a power cycle.  The pads I

I really need some help, because this is an expensive board with no spare PCBs available. 

Below in picture 1 is the area in question.  Notice to the right are some unused PCB holes directly to the right, where it looks like a split shift could be (if the plate allowed for such).  When I bridge those two holes with tweezers, it fires shift, so I assumed that wiring the legs directly to those would be a fix.  Well when I connect some wires, the shift does fire as soon as I touch the wires, but nothing happens when I press the actual key.  I must need to make another set of connections somewhere, but I really don't know where. NOTE: I DID test the switch, and it works when wiring it up to a breadboard and having it trigger an LED to turn on.

If anyone could give me some assistance I'd really appreciate it.  I'm trying to understand how the circuit should work, but nothing I'm trying with jumper wires or soldering directly is working.

Thanks for reading.

Picture 1: Closeup of the lifted pads.  To the right you can see the split shift holes I'm referring to
Picture 2: This wiring configuration seems so dumb in retrospect, but it did work to turn the Shift on, but not back off. 

[suicidal_orange]

Some people shouldn't be allowed to design PCBs, the guy(?) who made this one belongs on that list.  What were they thinking putting the diodes in blocks with most of the traces on the other side, then making it black so what little could be seen cannot?  Madness.

My first thought (that I cannot see to confirm thanks to blackness and flux) is that the two pads closest together with different jumpers soldered to them are probably connected and hopefully on the visible side of the board - can you see anything?  If one of your damaged pads isn't fully detached and you've wired the switch up 'backwards' it could not work (not sure this makes sense - this stuff is weird to talk about)

It's just one dead switch, we will get it working even if it means covering a bigger picture in colourful dots to map out half the matrix of the board.

[SuperBaseball]

Some people shouldn't be allowed to design PCBs, the guy(?) who made this one belongs on that list.  What were they thinking putting the diodes in blocks with most of the traces on the other side, then making it black so what little could be seen cannot?  Madness.

My first thought (that I cannot see to confirm thanks to blackness and flux) is that the two pads closest together with different jumpers soldered to them are probably connected and hopefully on the visible side of the board - can you see anything?  If one of your damaged pads isn't fully detached and you've wired the switch up 'backwards' it could not work (not sure this makes sense - this stuff is weird to talk about)

It's just one dead switch, we will get it working even if it means covering a bigger picture in colourful dots to map out half the matrix of the board.

Thanks for the assurance.  I agree about the PCB.  If it weren't for the look, I'd probably have this fixed already.  Unfortunately the traces on top are obscured by the plate + plate foam I added.  I reached out to the designer about sending me a diagram because I can't even find a good picture of the top of this PCB.  Hopefully he gets back to me soon.

The pads are 100% gone.  I think I see what you mean about it being wired backwards.  I guess it's possible that I've melted down the pcb so much that one of the legs is actually making contact.  That actually would explain a lot if I'm understanding the electronics correctly.  How would I go about testing that?  I've been mostly testing through trial and error with 2 unsoldered (or soldered to the switch legs) wires, then just poking around.  I've done my research and found others with the same problem on this site, but their pcbs were easy to follow and nobody really had BOTH pads of the same switch lifted.  I'm getting very close to the point of hiring a professional to hopefully fix this mess I've made so any info is helpful and appreciated at this point. 

[Rubbermilitia]

I'm no expert, but I've had luck with a hack solution for lifted pads .. you could try using a razor blade to expose the copper traces, use conductive copper tape to act as new pads, apply some flux, and solder a joint between the trace and makeshift copper pad. Then solder the pins to the new pads very carefully. Obviously not a great fix, and I did it on a cheap board.

Best of luck, I've been in your situation more than once (helidox pcbs jfc) and it sucks big time.

Sent from my SM-G950U1 using Tapatalk

[SuperBaseball]

I'm no expert, but I've had luck with a hack solution for lifted pads .. you could try using a razor blade to expose the copper traces, use conductive copper tape to act as new pads, apply some flux, and solder a joint between the trace and makeshift copper pad. Then solder the pins to the new pads very carefully. Obviously not a great fix, and I did it on a cheap board.

Best of luck, I've been in your situation more than once (helidox pcbs jfc) and it sucks big time.

Sent from my SM-G950U1 using Tapatalk

I'm happy to explore an option like this, but I find it strange how simply wiring the switch to the pads for split shift doesn't work. When I simply hold wires from a switch outside of the case to those pads, I can trigger shift from that switch.  But when I do it from the switch that's already mounted (again, I tested this switch multiple times) I get nothing.

I ordered a multimeter and some non solid core wire, hopefully those will help.  Still no response from TEX on a picture of the PCB or solution.

[suicidal_orange]

First thing if you're still testing (you wont get a response on the weekend so why not...) - your straight jumper is definitely wrong.  You can see a trace going from the APP key to the split shift position and continuing upwards, then there's another trace going up next to it from where you soldered on \| so they shouldn't be connected together.

If the split shift works and is in the same position in the matrix (it usually would be, though this PCB is weird) and it doesn't work with the U jumper as-is does it work like this?



If there is a dodgy connection as part of the pad remains on the front of the PCB by connecting the other pin/pad you would be holding the switch which may not register until you un-press it, which you can't do.

Another option if they're not in the same place in the matrix would be removing the solder from the switch and insulating the holes (small paper tube in the hole or heatshrink on the switch pins?) using thin jumpers soldered to the switch before pushing it through the PCB.  These jumpers could then be attached to the split shift pads with the sides being irrelevant as it is definitely not connected to anything.

[SuperBaseball]

First thing if you're still testing (you wont get a response on the weekend so why not...) - your straight jumper is definitely wrong.  You can see a trace going from the APP key to the split shift position and continuing upwards, then there's another trace going up next to it from where you soldered on \| so they shouldn't be connected together.

If the split shift works and is in the same position in the matrix (it usually would be, though this PCB is weird) and it doesn't work with the U jumper as-is does it work like this?

(Attachment Link)

I can't thank you enough, swapping the wires fixed it!  So one of the legs must be making a connection to the trace, is that it?

Attaching a pic of my shoddy wire job and a pic of the final rebuild.

[suicidal_orange]

Yup, one of them is still attached at least sometimes but there's no harm in connecting it twice.

Fix looks good and glad to hear it's working.  Nice to see some colourful caps on a Yoda too, is there any hope for a matching backspace?

[SuperBaseball]

Yup, one of them is still attached at least sometimes but there's no harm in connecting it twice.

Fix looks good and glad to hear it's working.  Nice to see some colourful caps on a Yoda too, is there any hope for a matching backspace?

I plan to just keep buying blank OEM colors so I can keep mixing and matching without having to dremel any of my nicer keycaps so maybe!