Need Electrical help (Soldering / Wiring)

[PuYang]

Spoiler!

Hi, sorry, not sure what to categorize this under. I am at my wit's end!

I am building a Hand-Wired keyboard, and am nearing the final stage, but ran into a few issues.

I'm looking for feedback from people who know more about electronics, soldering, and wiring, etc, than I do. (I have 0 knowledge actually). Or have done hand-wired keyboards before (super niche, I know, so general electronic feedback is probably what I'm asking for).

-----

For a little bit of context:

1a. Each key switch has 2 pins. One pin is wired along a ROW. And the other pin along a COLUMN. For example, keys QWERTY are on the same Row. 1QAZ are on the same Column. (These have been SOLDERED)

1b. A Diode is SOLDERED to one of the pins that leads to the ROW. (Switch > Diode > Row wire). No Diode needed for the other pin (Switch > Column wire)

2. The ROW and the COLUMN wires are then wired to the micro controller. (I did NOT SOLDER these wires yet). Each wire (in my entire build) are wrapped around their joint for 'decent wire on wire contact'. But I haven't soldered this step yet.

3. The micro controller has been flashed and confirmed working (bridging / creating a circuit on the inputs on the controller with a external wire simulates a keypress properly)

I decided to test my keyboard before Soldering the Row/Columns to the controller. With the wires all making decent contact, I figure solder TECHNICALLY isn't needed? (correct me if I'm wrong). As a closed circuit is being made??

Ok, hopefully the above explains decently where I am at.


NOW THE ISSUE:

Plugging in the keyboard, everything works except for a few keys (T, 5, 7). And some keys work inconsistently (Y, DEL). (Pressing the keyboard switch doesn't trigger the keypress).

Things I have tested / troubleshoot (For the T key):

1. Bridging or creating a circuit with an EXTERNAL wire on the Row and Column wires for the T key, does work and simulates pressing the T key.

2. Other keys in the same Row and Column as the T key are working. (Although 5 which is not working is on the same Column as T. But GB are on the same Column and do work properly).

3. I opened the switch to inspect for faulty switch. The inside of the switch is fairly simple, 2 metal... leafs? Contact each other to create a circuit when the key is pressed. Visibly, there are no issues.

4. I tried bridging from inside the switch using an external wire, DOESN'T WORK. But bridging the switch from below DOES WORK.


Some photos of the build if necessary:

Spoiler!

https://imgur.com/a/3uM1ntl

Apologies, attachments here kept failing... so IMGUR it is :S

The photo shows how the board is wired from below, with the T switch highlighted. Again, bridging the ROW and COLUMNS for the T switch works and simulates the key press. But pressing the actual switch does not.

-----


I'm hoping all of this made some sense xD... apologies for the long post.

My question now is, if the switch is not faulty, and I CAN make the T key register by bridging along the Row / Column wires, why is it not working when I press the switch? Is my soldering job affecting it?

As far as I know (in my super limited electronics knowledge), a circuit IS MADE when I press the switch. And if it works by bridging it from below, why wouldn't it work when I press the switch??? I mean, it can't get any simpler than that right? Metal touch metal = circuit = does it's job...

Would re-soldering the switch (Switch, Diode, Row // Switch, Column) help? Why would it help? If it is already completing a circuit now, what difference would it make?

Is there something I am not understanding? My solder joints look ... decent? In my opinion. I can't think of a reason why a 'bad looking' solder joint wouldn't work either. Since it's still completing a circuit regardless no??? Or am I totally wrong?


Any help would be appreciated! Thank you so much in advance!


EDIT:
Ok, so I did a bit more trouble shooting, and have narrowed down the issue a bit more.

When I bridge the faulty switch (T) at the switch pin BEFORE the diode, and then somewhere on the column wire, it does not work.

Keys that are not faulty can be bridged BEFORE the diode and anywhere along their columns.

This means something is wrong with my Switch Pin --> Diode.

I will have to try resoldering that joint later tonight. My question now is, what would be the logical cause for this? Can a bad solder joint actually make this not work at all? (Prevent closing a circuit?) I don't think the solder joint looks bad at all, it looks like any other joint I've done. But I'll redo it regardless.

EDIT2:

Ok... apologies to anyone who spent time reading my post... I re-soldered (heated up, moved the pre-existing solder around a bit, applied a tiny bit of new additional solder) to my T switch, and now it works.... o___O so, conclusion is, something was indeed wrong with my solder joint, even though visually it looked fine, and a closed circuit SHOULD have been achieved regardless.... I'm still confused, and none of this makes sense to me personally. But I am glad it works now!

Thanks again!

[underscore]

If the keys that aren't faulty can be bridged before the diode, and work, what are the diodes for?

Do you have a multimeter with a diode test function?

Is there anything that the faulty keys have in common (ie using the same row or column)?

Are you sure there isn't a short somewhere? That's a lot of exposed metal in close proximity.

Also I'd really suggest getting solder on all the connections. Alternatively use test leads with alligator clips so you're getting some pressure on there. Loosely twisted wires are a recipe for problems.

[PuYang]

Hi, thanks for reading / replying. Apologies, as I sorta resolved the issues.

The diodes for each switch > Row is to guarantee you can press any number of keys and any combination of keys without issue (as you'd want to in a keyboard. I think they call it N-Key Rollover for keyboards). TECHNICALLY, you don't need a diode on EVERY single switch. But then, you'd have to figure out which combination of keys will require it... so it's much easier to just add a diode to each switch to guarantee all combos work. I don't know the details, but something about diodes makes the current flow only in one direction is supposed to resolve that issue?

There are no shorts! I put in a lot of time and effort to ensure all wires are bent and clearing eachother. Probably the most time spent in this entire build is designing and bending rigid the wires xD

I will be soldering the remaining connections tonight or tomorrow. But as of now, I have fixed the issues I was experiencing! I simply reheated the existing solder joints on the switches that weren't working and added a bit more fresh solder to the joints and now the keys all work properly.

My DEL key that I mentioned as somewhat inconsistent did turn out to be a faulty switch, as I soldered / re-soldered its joints a few times and it still acted funny. I then de-soldered it and removed it. Put in a new switch and it now works correctly!

Thanks to everyone who read through my long post. I apologize, as I made it in haste / frustration. Then solved the issues a few hours later. Probably could have saved you guys the trouble of reading it all xD

[underscore]

Derp, yeah preventing backfeeding makes sense. I was thinking resistors for some reason.

Glad you got it sorted!

[PuYang]

Wasn't sure if I wanted to bump this, as this is resolved, but thought I would post the completed project for anyone curious:

https://imgur.com/a/CWTRpz6

Still working on my custom keycaps (Spiderman / Venom, etc). I have a Carnage, Gwenom, and Scream that are unpainted right now. They will be on the far left row (currently F1, F2, F3, F4)

I am quite happy overall with the result of the wiring. Everything is in the open and exposed, with the gold-toned brass wires looking quite blingy.

Thanks for reading!