My solar tester is coming along nicely.

It started as a thought when I was using a tabletop charger/tester after buying a bunch of 18650 cells for my meshcore builds. If I have a bunch of solar panels and charger modules, why not use them?

My electronics design skills are pretty nonexistent so I admit, I used llm to draft a hardware design and iterated on it for a bit while simultaneously learning how it should work. After becoming reasonably confident it should work, I took a plunge and started to learn KiCAD. Drawing the schematic, even designing custom parts and footprints to possibly make a proper PCB. All that was taking a long time though, so eventually I settled on using a prototype PCB (didn't finish the KiCAD design).

After getting all the missing parts (it's only 3 transistors and a bunch of resistors apart from the MCU, display and sensors) I started to lay them out on the board according to the schematic. That took a while but in the end I had a plan (some painted traces are colored and will probably be invisible here).

It's a ProMicro Supermini nrf52840 clone at the core (chosen for low power consumption). System power is from a 18650 cell charged with a small solar panel through a CN3791 module. Test circuit is separate for easier control and measurement, although it leads to some duplicated parts (panel and charger).

Test cell is charged by a separate panel through another charger module. Both the test panel and charger are pluggable and can be swapped for different parts. Two INA226 sensors measure current and voltage at the charger input and output. This allows to calculate panel and charger efficiency. After full charge the cell is discharged through a 10W 6R8 resistor to measure its capacity. After that, it's charged again to a preset storage voltage (3.6v).

Soldering it took maybe two days. I needed to cut some PCB traces, learn how to solder SMD SOT-23 transistors, bodge several wires and be careful to not ruin anything. In the end the hardware pretty much worked on the first boot! The firmware is not done yet, but I have the display working, sensors providing good values (some calibration still needed), MOSFETs correctly switching charge and discharge circuits, and a rotary switch for mode selection.

Tomorrow I'll solder the charger and panel connections to test charging. The discharge path was tested today. Then completing the firmware, polish, maybe some shoddy enclosure if I feel like it. Exciting!