Its a partially cloudy day today, and the device charges to 4.8-5v no problems.

I just got a new fuse for my multimeter amperage testing mode (be careful when switching from testing current to voltage!) So i did some tests on how much this thing consumes with different things going on.

• Device on, no LED:     1.3 mA
• Device on, 1 LED lit:     1.5 mA
• Device on, 2 LEDs lit:  1.7 mA
• Device on in PWR_DOWN_MODE, all gpio set to input,pull_down, ADC off: 0.03-0.04 mA, or 30-40 uA

Obviously my meter is not super precise testing down in the uA range, but thats good enough for me.

30-40 uA is quite a lot for attiny85 in deepsleep. I am guessing the extra current consumption is the voltage dividers on ADC2 & ADC3.

ADC3 is connected to VCC with a 20k resistor for the top half of the button resistor ladder. If that were going straight to ground at the voltage i tested (3.9v), that would be drawing about 0.4 mA, so i think most the consumption is ADC2.

ADC2 is for reading VCC voltage, and is connected to VCC through 100k resistor and ground with 22k resistor. This brings the voltage on ADC to maximum of 1v when VCC is at its absolute max of 5.5v.

At test voltage (3.9v) and 120kOhms, This voltage divider itself will draw ~0.03 mA. Gotta research the maximum resistor value i can still get an accurately enough reading with...

 I will experiment later with higher value resistors to see how low i can go without losing responsiveness with buttons and still get good voltage readings from VCC.

As it is I imagine if I wrote a program to have this device run some code then deepsleep for 1 minute in a loop and the solar panel was facing the sun it would run indefinitively. Not sure a good use case for that but the setting where it deep sleeps if power drops below 2v should give the user ~2 hours to get it back in the sun for a few minutes to recharge, and continue working or save program.

I have also tested putting this very close to an LED lamp and it will indeed charge. Could be useful in the long dark winter months...