What's wrong with my keyboard?

[throughthefireandtheflame]

Hi everyone!

I have Ducky YOTG that's served me well over the years, but last year it started acting up as can be seen below (all I did was press the ALT key a single time):


I initially tried finding someone to look into it for me, but the truth is no IT guy around my place will venture beyond swapping RAM modules… so I started doing research online and picked up some basics before diving into the board myself.

First, I found some sticky spots around the ALT area that resembled dried soda, so I started by scrubbing everything thoroughly with isopropyl alcohol.
With a multimeter, I then ran some continuity tests and checked all resistors near the affected area, but nothing really stood out to me.
Since I couldn't think of something else, I went ahead and replaced the MX itself because it looked bad, but that didn't do the trick either.
At this point, I feel like I did what I could do confidently, I just don't have the knowledge to pinpoint the issue any further.

Here's the board:


Here's how the ALT area looked before doing any maintenance:


And how it looks now after replacing two keys:


If anyone could give me directions on what to check or where to look for possible faults next, that'd be greatly appreciated!

[Rhienfo]

Maybe replace the diode? that's the only thing that I can think of that might have been an issue. I'm not very knowledgeable about pcbs so maybe there is something else I don't know about that could be the issue. But just throwing the idea out, just in case it works/leading you to fixing it.

[throughthefireandtheflame]

Thanks for the suggestion! All the diodes I tested with a multimeter didn't display any odd behavior so I didn't think much of it, but one of them may be damaged in a way that I couldn't catch. At this point, I might as well try to find replacement diodes and see if that can help with this issue. I'm just hoping the issue isn't somewhere further down the traces that can't be easily pinpointed. Might take a while to get parts, but I'll keep updating here in the event that I make any kind of progress.

[Rhienfo]

Thanks for the suggestion! All the diodes I tested with a multimeter didn't display any odd behavior so I didn't think much of it, but one of them may be damaged in a way that I couldn't catch. At this point, I might as well try to find replacement diodes and see if that can help with this issue. I'm just hoping the issue isn't somewhere further down the traces that can't be easily pinpointed. Might take a while to get parts, but I'll keep updating here in the event that I make any kind of progress.

All good, good luck with fixing it. Hope it goes all well.

[throughthefireandtheflame]

Took a few weeks, but all parts have finally arrived!

I didn't feel like it initially, but I ended up desoldering every single switch to remove the PCB from its frame.
Desoldering ended up being a good idea because I found more soda residues that were not visible before, plus it just gave me the opportunity to thoroughly clean everything else.

Here's the side of the board I couldn't see:

Here's the backside of the board, now desoldered.

Here's the frame (which I'm considering stripping to repaint it if the board works):

Here's a close-up of the cleaned ALT area:


I decided to test every single trace and diode—much more in-depth than what I initially tested—, but didn't find any issue either.
I'm starting to think my issue didn't require new parts… but those were cheap and can always come in handy!

Now, this might be a bit of a tricky question to answer, but how many (and which) switches would I need to solder back to the board to test the ALT key again? I know it works a bit like a grid, so just soldering the ALT key alone isn't enough, but I'd rather not spend time soldering more than necessary if the board is actually toast.

Again, any help would be appreciated!

[Axiom_]

That's alot of work you've put in.

Before moving on to soldering stuff back, I would double check to see if the diodes were properly measured.

Anything other than a low resistance reading in one direction and high reading in the other is a bad diode.

If that's been done correctly then proceed with the following:

Check the controller chip model and see if you can find a schematic for it. Knowing the schematic and logic of the controller chip is essential.

Otherwise, you need to complete the circuit before you could carry out a proper test. That will likely require you to solder everything back.

It may be a faulty controller. Since its hard to know the logic behind the pins, its hard to diagnose and fix.
Your best bet is to contact ducky for a replacement board if nothing works.

[throughthefireandtheflame]

Thanks for your input!

Every step has been a learning experience so far, I'm almost impressed I didn't break anything else.

Regarding diodes, I feel like I've tested every single one of them more than I care to count, so it doesn't appear to be the source of the issue (unless I'm awfully incompetent ).

I did some research on the microcontroller, which turns out to be an HT32F1655 from Holtek.
Luckily, its datasheet is public and it contains detailed schematics (most of it goes way over my ahead though, but page 22 looks interesting): https://pdf1.alldatasheet.com/datasheet-pdf/download/1013860/HOLTEK/HT32F1655.html

Here's the bad boy:


Nothing looks out of the ordinary in the area, but it could be something internal.
I'll probably sit down and solder everything back to conduct more tests when I got time, though at that point I'm wondering if I shouldn't just make it hot swappable.

Ducky support was my first stop on this long journey, but I got ghosted.
Last time I contacted them for a spare part of this model, they did answer but said they were out of it, so I wouldn't hold my breath for a spare PCB.
Guess I could always look for a donor from this era, but I don't know if standard editions share the same board.

In any case, I'll keep updating. I may not be able to repair it, but maybe it'll help someone else down the line.

[Axiom_]

Thanks for your input!

Every step has been a learning experience so far, I'm almost impressed I didn't break anything else.

Regarding diodes, I feel like I've tested every single one of them more than I care to count, so it doesn't appear to be the source of the issue (unless I'm awfully incompetent ).

I did some research on the microcontroller, which turns out to be an HT32F1655 from Holtek.
Luckily, its datasheet is public and it contains detailed schematics (most of it goes way over my ahead though, but page 22 looks interesting): https://pdf1.alldatasheet.com/datasheet-pdf/download/1013860/HOLTEK/HT32F1655.html

Here's the bad boy:

Show Image

Nothing looks out of the ordinary in the area, but it could be something internal.
I'll probably sit down and solder everything back to conduct more tests when I got time, though at that point I'm wondering if I shouldn't just make it hot swappable. :rolleyes:

Ducky support was my first stop on this long journey, but I got ghosted.
Last time I contacted them for a spare part of this model, they did answer but said they were out of it, so I wouldn't hold my breath for a spare PCB.
Guess I could always look for a donor from this era, but I don't know if standard editions share the same board.

In any case, I'll keep updating. I may not be able to repair it, but maybe it'll help someone else down the line. :thumb:

It sucks that stuff just gets thrown out as e-waste all the time. Kudos to you for trying to diagnose and fix it.

If the pcb traces are fine, diodes and switches are all fine then it'll most likely be a faulty microcontroller. There's also a bunch of capacitors, resistors and stuff on that PCB. You may want to test those too.

The real problem lies in the proprietary firmware installed on the controller. If ducky can't send you a replacement with pre-installed firmware then you'll need find a way to get a copy of it and flash it onto the controller yourself.

On that topic, apparently the pok3r keyboard uses the same controller. There's been some interest in flashing custom firmware onto the controller here: https://geekhack.org/index.php?topic=72262.0 but it seems like progress has stalled.

There's also talk of flashing QMK firmware onto the pok3r here: https://github.com/pok3r-custom/pok3r_re_firmware/issues/23

It seems like ChaoticEnigma  is very knowledgeable of Holtek controllers as seen here: https://github.com/pok3r-custom/pok3r_re_firmware/issues/4

You could try to get in touch with them or check that github repo for anything useful.

Due to the proprietary nature of the whole system, it'll take a while and quite a bit of testing to know how the circuit's logic is setup. Finding or creating a circuit diagram would be a good start.

If its too much of a hassle, perhaps you could try buying a second hand ducky keyboard for parts.


All in all, best of luck with your efforts.

[throughthefireandtheflame]

You're right, I should definitely test more than the diodes and resistors.
I just felt like I wasn't doing a good job testing chips initially, plus they were all away from any visible issue, so I kinda left those alone.

Looking up the Vortex model you mentioned turned out to be interesting, because it appears to be similar to mine beyond just the controller: https://geekhack.org/index.php?topic=72262.msg1765976#msg1765976
The board itself looks similar in the way it's been designed and manufactured, so any work done on this model could help understanding what's going on on the YOTG!

Still, checking the threads and repos you listed, I feel awfully overwhelmed, so I will keep working on things I'm confident in doing first since I'm not completely out of ideas yet.
If all else fails, then I'll go down the controller route and try to get in touch with people who worked on similar models (by the way, you weren't kidding when you said the dude was knowledgeable )
Would always be interesting to have the firmware dumped, even if not to mod it, just to have it around because the only related thread is vastly outdated: https://geekhack.org/index.php?topic=78520.msg2001492;topicseen#msg2001492
I always had concerns about the firmware, because I believe most of the lighting options are (at least were back then) exclusive to the model, which'd all be lost if I were to replace the whole board with a different one…

Anyways, enough rambling. Back to testing!

[Axiom_]

You're right, I should definitely test more than the diodes and resistors.
I just felt like I wasn't doing a good job testing chips initially, plus they were all away from any visible issue, so I kinda left those alone.

Looking up the Vortex model you mentioned turned out to be interesting, because it appears to be similar to mine beyond just the controller: https://geekhack.org/index.php?topic=72262.msg1765976#msg1765976
The board itself looks similar in the way it's been designed and manufactured, so any work done on this model could help understanding what's going on on the YOTG!

Still, checking the threads and repos you listed, I feel awfully overwhelmed, so I will keep working on things I'm confident in doing first since I'm not completely out of ideas yet.
If all else fails, then I'll go down the controller route and try to get in touch with people who worked on similar models (by the way, you weren't kidding when you said the dude was knowledgeable )
Would always be interesting to have the firmware dumped, even if not to mod it, just to have it around because the only related thread is vastly outdated: https://geekhack.org/index.php?topic=78520.msg2001492;topicseen#msg2001492
I always had concerns about the firmware, because I believe most of the lighting options are (at least were back then) exclusive to the model, which'd all be lost if I were to replace the whole board with a different one…

Anyways, enough rambling. Back to testing!

Down the rabbit hole you go )

This is going to be quite a project (if you go down the custom firmware path).

Keep us posted.

[throughthefireandtheflame]

Time for updates!

On the upside, I got the frame powder coated and it turned out pretty good!


On the downside, after carefully inspecting and cleaning every thing I could, I soldered everything back together and… it shows the exact same faulty behavior.

I can only assume the microcontroller is the issue as others suggested, which at this point makes me feel I might as well swap the whole board if I can find one. I have zero hope of finding a YOTG board, but I know some models have similar boards in them (the Shine 5 if my memory serves me right—if you know of other compatible boards, let me know!), so I could always give that a try. I'm sad to know that the board's number won't match that of the frame anymore, but I just want to be able to use my keyboard again (being stuck with a spare K200 for a year sucks ).

I did get in touch with Ducky again in the meantime, and I must say I'm wholly unimpressed by their support. My model isn't even 10 years old and Ducky themselves don't have the firmware anymore! No wonder they didn't answer at all when I asked for spare parts. I did end up finding the firmware for my layout on some obscure website and managed to update it, but that did not help either.

Sorry the outcome wasn't the one we hoped for, but I learned a couple of things on the way and I'm not quite giving up yet.

[Rhienfo]

Time for updates!

On the upside, I got the frame powder coated and it turned out pretty good!

Show Image

On the downside, after carefully inspecting and cleaning every thing I could, I soldered everything back together and… it shows the exact same faulty behavior. :(

I can only assume the microcontroller is the issue as others suggested, which at this point makes me feel I might as well swap the whole board if I can find one. I have zero hope of finding a YOTG board, but I know some models have similar boards in them (the Shine 5 if my memory serves me right—if you know of other compatible boards, let me know!), so I could always give that a try. I'm sad to know that the board's number won't match that of the frame anymore, but I just want to be able to use my keyboard again (being stuck with a spare K200 for a year sucks :D).

I did get in touch with Ducky again in the meantime, and I must say I'm wholly unimpressed by their support. My model isn't even 10 years old and Ducky themselves don't have the firmware anymore! No wonder they didn't answer at all when I asked for spare parts. :rolleyes: I did end up finding the firmware for my layout on some obscure website and managed to update it, but that did not help either.

Sorry the outcome wasn't the one we hoped for, but I learned a couple of things on the way and I'm not quite giving up yet. :thumb:

Good luck on repairing it, hope it all goes well :)