OLED issues with nice!nano v2

[clefever]

On a paintbrush v5 using a nice!nano v2, I cannot get the OLED screen to display anything.

The OLED display was working fine after my initial build and flash of the firmware until I plugged the board in to USB to charge for a while, after which it stopped working.

This sounds a lot like #674 in the zmkfirmware/zmk repo where the display is not being reinitialized, but I was unable to get suggested fixes to work for me in my own forks. It's possible I didn't implement the fixes correctly or a correct combination of them.

Some other observations:

Upon powering the board on, VCC seems to jump to 3.3V for a split second, then go back down to 0V. I tried the v1 nice nano version which maintains 3.3V, but still no display. I'm thinking this might be part of the power-saving feature / sleep.

In the debug logs, I get messages about the display device not being found. I tried a fix in the previously mentioned issue to reinitialize i2c, but with no luck. Again, it's possible I missed a step in applying the patch and building in my own fork.

[mcrosson]

I have pushed an update that disables sleep for ardux to hopefully avoid this issue for users.

If/when the upstream issue gets addressed we can look at enabling it and/or possibly adding a compile flag to enable it easily but have it off by default.

Going to leave this issue open for feedback on preferred approach(es).

[ThSGM]

I can confirm that I have almost exactly the same issue as the original poster. Like other discord members we also tried flashing the stock firmware via https://github.com/zmkfirmware/zmk/actions/runs/2741574670

[mauza]

I'm experiencing the same symptoms with nice nano v2 and corne lp keyboard