Solar tester update: soldered solar panel and charging module connectors today. Charging path was not switching on correctly, turned out I missed one connection in the PCB design. After adding that the switching MOSFET turned the charging module on as it should.

Still, there was a problem: the test cell was not getting any meaningful current. After a lot of multimeter measurements and swapping the panel with a lab power supply, I concluded that the CN3791 charger module was the culprit.

These Chinese boards come in flavors that differ with the panel operating voltage. I always ordered the 6v variants (that are actually set to 5v). This particular specimen turned out to be a 12v one though! I guess these are cheap for a reason... I found one more set to 12v, I'll need to go through all of them and double check. You can tell the voltage setting by values of a couple resistors on the board.

After using a correctly tuned charger and doing some more voltage and current calibration: it works!

Next: implementing actual charge/measure cycle.