|
Post by lbendlin on Jan 22, 2019 5:44:57 GMT -8
It's fascinating to observe the behavior of the SunControl board , especially in the current conditions (2F for long periods of time). My goal is to be able to properly predict the charge level of the attached battery (10000mAh Li-ion), but it's really complex, especially when only considering the voltage readings of the INA3221. When the sun is shining and the solar panel is pushing charge into the battery I see some crazy spikes in the battery voltage, and they seem to become crazier with the recent temperature drops. I tried to account for that by looking at the waste energy (stuff that is produced by the solar panel but goes neither into the battery nor to the load) but it does not seem to be a linear relationship. I think what I'll do next is experiment with the Quad Power Management Board to temporarily disconnect the solar panel, take the battery voltage measurement, and then reconnect the solar panel (unless the battery is above 4.1V, and the charging current was down, for example) This article helped me to understand some of the complexity batteryuniversity.com/index.php/learn/article/charging_lithium_ion_batteries(I am not fully convinced that the thermistor is doing its job. The battery is in the weatherbox housing, with a Zener heating, but that still gives only about 19F average temperature. Yet the INA3221 claims that the battery is charging at 800mA ...)
|
|
|
Post by SDL on Jan 23, 2019 16:24:29 GMT -8
|
|
|
Post by lbendlin on Feb 28, 2019 7:22:44 GMT -8
I went for the method of switching the solar panels off completely, waiting half a second, and then measuring the battery voltage again. That seems to work well enough so far. The light green line is the reported voltage with solar panels on, the dark green line with solar panels off. From 9am to 12pm it was too cold for charging, so the panels were off during that time.
|
|
|
Post by SDL on Feb 28, 2019 12:02:49 GMT -8
Lutz,
I thought you told us that you had SunControl shutting off the power to the load too in this case. Is this not the case?
BP
|
|
|
Post by lbendlin on Feb 28, 2019 12:46:01 GMT -8
That was when I used the thermistor. Now I am not using the thermistor to control the charging but the Raspberry Pi, the INA3221 and the QPM. The Pi monitors the conditions (mainly temperature and battery current), and then switches on the solar panels selectively to only allow charging in the "good" temperature range (with the ability to control this by the degree rather than having to rely on the thermistor's guesswork) and to stop attempting to charge when the battery is already full.
Of course this also means I lose control over that when the Pi is not powered on, and there's a risk that the setup attempts to charge the battery outside the "good" temperature range.
|
|