Testing my PCB prototype. Weird phenomenon ...

[Eszett]

Hi! While testing my PCB with AquakeyTest.exe I encountered some weird phenomenon: I'm shorting each switch footprint's pads with tweezers to see in Aquakeytest.exe whether each switch actuates properly or not. Here is what I got for shorting the keys W, E, R:



"W" actuates W+A, "E" actuates E+S, "R" actuates R+D, ie. the next switch below in the same column is actuated too. It happens not always, but most of the time. How can that be? Do you know this phenomenon and what is the cause? and how can I solve it? Update: When I change the tweezers by a jumper cable, the phenomenon is gone, and there is no "double actuation" anymore.

[TalkingTree]

Taking the wildest of guesses, I would think of a faulty diode.
Perhaps a close up picture of the PCB area could help.

[Eszett]

Hi tree man! Your guess gets me to an idea. I could replace the diodes on the PCB (they were aseembled by the china fab, maybe some cheap noname ones) by some of my Mini-Melf diodes I have lying around here! That should give more information..

[Findecanor]

Check the traces and diodes with probes on a multimeter. That should allow you to pinpoint where the fault is.

[TalkingTree]

I could replace the diodes on the PCB

Sure, at least the ones on W, E and R.
Keep me posted, I'm curious.

[Eszett]

Findecanor & TalkingTree: Ye, will apply the multimeter tomorrow, then I can report.