SCK 2.1 Solar Panel / Battery weird readings when first charging

[david1]

I have attached the following a 5v 10w solar panel with a usb output to my kit;
https://www.amazon.co.uk/gp/product/B07FMVKHLY/ref=ppx_yo_dt_b_asin_title_o00_s00?ie=UTF8&psc=1

It is working great with the kit charging quickly and running right through the night. However, when light starts to appear and the light sensor gets to about 500lux the battery charge seems to suddenly drop by about 30% before starting to charge again. The following graph should help explain;

This is showing Battery (dark orange) v Light (yellow) over about 3 days.

Has anyone got any ideas what is going on? It doesn’t make much sense to me to get a large battery drop just as it starts charging.

I’ve been playing around with the API as well. Very well done SC team. Keep up the great work.
https://draksck.1mb.site

[lz.junks]

The first question is if that solar panel has a built in diode. I did not examine the SCK circuit, maybe there is a diode there. Could it be that you are observing a “fight” between the battery and the solar panel?
The question to the SC team - how to properly add solar panel?

[david1]

I never thought of that. There is an led in the solar panel but it still seems a lot of drop in battery charge in such a small amount of time.

At 05:00 the capacity reported at 53% which had been steadily dropping all night but by 05.03 it showed at 9%. It almost looks like it wasn’t reporting properly.

I’ll unplug the solar panel and see what if the readings follow the same decline and then plug it in again tomorrow morning.

[victor]

Hello David1,

Sorry for the late response.

To calculate battery percent we use lookup tables, we use one table while discharging and a different one during charge. This tables are calculated with a stable power source.

We should make some tests to be sure, but my best guess is that the solar panel starts providing power with a lower voltage than the expected 5 volts, and since we are charging the battery we switch to calculate the percentage based on the charging lookup table.

If you look at the point where the panel stops providing power the problem seems to exists but in the opposite direction.

I think the best way to solve this will be to take in account not only the started or stopped charging notification but also check wich voltage we are receiving to calculate the percentage.

Thanks for posting your findings!
We will put this in the TODO list!

Cheers
Victor

[david1]

Hi Victor,

Thank you for that It certainly makes sense. I think also the panel does not work quite as I expect which is most probably due to the low cost. If the panel is plugged in overnight, and not charging, then the power usage is about 30% to 40% higher than if it is not plugged in which suggests it is actually drawing some power.

I’ll keep playing around and see what happens.

Just out of interest what is the maximum panel voltage as some people are talking about 12v but I stuck with 5v with a USB socket as it was an easy starter option?

[victor]

Hi David,

Acording to the datasheet of the charger IC to be on the safe side the voltage should be under 6.2v:

So a 5V panel is the way to go.