Only connected side is working

mattdibi / redox-keyboard

Ergonomic split mechanical keyboard

MIT License

1.75k stars 163 forks source link

Only connected side is working #164

Closed joe-batt closed 1 year ago

joe-batt commented 1 year ago

So this is yet another "only one side working issues".

I'm using the serial connection (without resistors)
TRRS cable is working (tested on a Redox bought from Falbatech)
TRRS Pad connections seem OK using a multimeter
D0 Pin connection seems OK
both Pro Micros light up
when typing on the TRRS-connected half, every once in a while a character might actually be send
I am using the 3D printed case

When mixing and matching with the Falbatech Redox (using the same firmware) the issue persists, only the connected half is detected. The full Falbatech one is working as expected. What would be the best course of action to debug the issue, preferably without having to desolder the pro micros and key switches?

joe-batt commented 1 year ago

I did some further testing. I am pretty confident the solder joints should be ok. Every single Key is working and if I measure the resistance between D0 and D1 on both boards, in comes out to 1.5 Ohm. The Pro Micros are the USB C variant.

The new Keyboard uses the rev1 PCBs from Github and is without underglow, the Falbatech one is prebuilt with underglow. Both are using the exact same firmware using MASTER_LEFT. If I connect one half of the Falbatech board as slave, the underglow stays off. It does not matter which half of the Falbatech I use as slave or which of the new boards I use as master. The Falbatech Board uses 4.7K Ohm resistors on the left side, while I have none at all.

What's bothering me is that on the slave side, I have only 2.7V between VCC and GND.

joe-batt commented 1 year ago

The issue appears is definitely the the power transfer. When I force slave/master as described in #33 and power both sides, it's working as expected. In fact I am typing on the keyboard right now.

I'll still leave this open for now, as I am still open for advice how to rectify the original issue.

joe-batt commented 1 year ago

Turns out the problem existed between keyboard and chair. The TRRS sockets are pretty firm, the plug appears to click into place, but there is still 1mm or so left before it actually clicks into place, so the tip had barely any connection.