Keyboard randomly locks up while PC is sleeping.

[fOmey]

Bug Description

I use RGBLIGHT_SLEEP. I've noticed if the RGB lights turn back on while the PC is sleeping (usually overnight).. I know the keyboard is stuck in this "frozen" state.

If I look under devices & printers the keyboard is labeled as: "Unknown USB Device (Device Descriptor Request Failed)". Only fix being removing the USB & re-plugging it in.

Everything was working fine up until what seems like a recent upstream sync, but looking at the commits I can't notice anything that would cause this? Coincidence perhaps?

I have my is keyboard connected right directly into the back of my PC, no hubs are being used.

Any ideas how I can debug further? The lock ups seem to be very intermittent, I'm struggling to come up with anything.. can't seem to reproduce them either.

System Information

• Keyboard: KDB75 REV2 board

• Operating system: Windows 10

• AVR GCC version: avr-gcc.exe (AVR_8_bit_GNU_Toolchain_3.6.1_1752) 5.4.0

• QMK Firmware version: 0.7.26

  • AutoHotKey
  • Karabiner
  • Other:

EDIT:

The issue still persists, I've noticed if I wake the PC up with the mouse however.. the keyboard doesn't always lock up into the "Unknown USB Device (Device Descriptor Request Failed)" state. At least when this happens I can avoid having to remove the USB cable to restart the keyboard...

[drashna]

That's really odd. Given the controller and everything, that's ... not typical.

My guess would be faulty hardware. Be it the USB port, the USB cable, or a short on the PCB.

But if you disable rgblight sleep, does the issue persist?

Also, there is a config.h setting that disables sleep outright. It may be worth setting that and see if that helps

[fOmey]

It is a strange issue, a very intermittent one also. It only happens every few days?

If it was a short of some kind, I'd imagine it would happen a lot more frequently.. I would be able to reproduce it, seems to be very set/specific circumstances that cause this to happen. Also while the PC is on the keyboard functions absolutely flawlessly.. zero issues, I don't think a short is the culprit.

It's as if the keyboard is detecting a false wake up signal, trying to initialize and breaking when its not receiving the expected result.. that's my guess.

I don't believe this to be related to RGB sleep at all.. I think it's just a side effect of the issue I described above that they're turned back on.

I'll test a brand new USB and see if the issue persists,

I've also modified some power settings on Windows to see if that has something to do with it.

[fOmey]

Purchased a brand new USB cable shortly after my last message, everything seemed fine up until 2 minutes ago when the issue returned.

God dam it, really thought it had fixed it for a second!

Argh, i'm going to dismantle the board and inspect the PCB.. give it another really good isopropyl clean in hopes a miracle happens and the issue disappears.

EDIT: Testing yet another USB (I noticed the replacement is the same brand).. fingers crossed.

[fOmey]

Well it's been 4 days and the issue has not reappeared since swapping the USB C cable out.

I've attached some images of the cable I was having issues with for record, just in case someone finds themselves in a similar situation in the future.. it was a generic braided USB C cable I sourced off eBay: