Mouse modifier key not reliably applied

[maxbethke]

Hello,

since today Im experiencing a really annoying behaviour with my UHK60 v2.

Im using a keymap which has the mouse layer toggle as secondary action on the "caps" key, which also has Escape as primary key

The problem now is that this 'secondary action' seems not to the applied reliably anymore.

Symptoms

I hold ESCAPE and press Y or N to scroll up and down. Now

• Sometimes it continues scrolling long after I released Y or N
• Mid scroll, the MOUSE current layer switches back to BASE, which means I starting typing Ys or Ns, which in turn resets my scroll state in most programs

I also noticed that the cursor, using the trackball module, sometimes lags behind. Also sometimes when I type some keys are lagging. It feels like the keyboard disconnects from time to time for a second-or-so. This might also reset the current layer, which might explain the scrolling problems.

[maxbethke]

I just updated my agent and both halves to firmware v17.0.0, which did not help

[kareltucek]

Hi!

• Any idea since which firmware version has this been exhibiting?
• Is only the mouse key affected? If so, have you tried to reseat or swap the switch?
• What is your typing behavior settings?

[mondalaci]

Just moved this issue to the firmware repo since it's a firmware issue, assuming it's not hardware.

[maxbethke]

• Any idea since which firmware version has this been exhibiting?

16.2.0 -> 17.0.0

• Is only the mouse key affected? If so, have you tried to reseat or swap the switch?

No, it seems like any key is affected as in sometimes the keys lag behind for a second(ish), but on the mouse key, its particularly noticeable for me because that is the only key that I frequently hold down for longer periods of time (for scrolling)

I pulled an reinserted the switch today anyways and lets see if that does anything.

• What is your typing behavior settings?

I dont remember playing with this, although I increased the key-repeat-rate in my OS

[kareltucek]

Would you mind giving a test to 16.3.0 ( https://github.com/UltimateHackingKeyboard/firmware/releases )?

There was a major transition from one usb stack to another, which might have caused issues like this.

[maxbethke]

To my surprise, it seemed to have disappeared after I reseated the Mouse/Caps key, but now after a few days (in which I hardly used the keyboard), it seems to be back. Correlation rather than causation?

I just flashed 16.3.0 onto all devices and will test it for a few more days... is what I wanted to say, but during the writing of this sentence, it already spammed keys and reset the cursor location twice... so probably I dont need a few days of testing to say that that wasnt the solution.

[maxbethke]

There is another thing id like to rule out first though... until now, it seems this happens when I use my new desk treadmill. This might also just be correlation, but ill uninstall my feet and lay the keyboard flat on the desk for a while. I guess when using feet for tenting while using the palm rest, the board might be susceptible to pressure applied to the palm rests?

[kareltucek]

Correlation rather than causation?

Probably as it doesn't explain any lagging. Although it might also be two unrelated problems.

Still, be sure to bend the pins a little so that they end up tensioned against the side of the hotswap socket.

---

Ad 16.3.0, that supports the usb stack hypothesis. To be sure, please retest 16.2.0.

---

Please do test the treadmill hypothesis. I had the same problem (badly misbehaving uhks) when using treadmill. It got solved by moving the treadmill onto a different power line.

[maxbethke]

Wild, as soon as I step of the treadmill and switch it off, the symptoms seem to disappear...
Let me test this a little longer and try another wall socket.

[mondalaci]

The UHK 60 is sensitive to electromagnetic interference (EMI). Please see the "Electromagnetic interference issues" section of the https://uhk.io/repair-and-modding/troubleshooting-repeated-and-missing-keystrokes page, where treadmills are explicitly mentioned. I'd be happy to extend this guide with your findings, if you have any.

[maxbethke]

I am still running some tests, but right now I observe the following for my particular setup:

• The EMI is emitted wirelessly, not over the power grid.
• The spiral cord connecting the two halves seems to act as antenna, as the disturbance is significantly lower when I connect the two halves directly together

[mondalaci]

If everything else fails, replacing the bridge cable with a straight cable may solve this issue. It should be as short as possible.

[maxbethke]

If everything else fails, replacing the bridge cable with a straight cable may solve this issue. It should be as short as possible.

Thank you, this is a very good point. Having a straight cable will also make it a lot easier to apply a ferrite to it, which would have been my next solution approach. Sadly I was sick the last week, which is why I could not make any progress with the tests.

I probably have some telephone cable flying around and some old VGA cable I can salvage the ferrite from for testing. I will get back once I tried it.

[maxbethke]

I guess getting a UHK 80 would also be a solution since its fully wireless, even thought I dont like the extra footprint.
Is there a UHK 60 with wireless capabilities planned?