Guru
Joined: 07/09/2016
Location: United KingdomPosts: 2171
| Posted: 09:23am 13 Jun 2025 |
morning Pete. You well?
I agree that solar panels are "just" photo-diodes and so nothing flows into them, but it just doesn't sit right with me not to have some blocking which I can control - it probably says more about my trust issues over mass-produced, anonymous chinese stuff than anything. If the panels were 50V+ jobbies, the 0.3V of the barrier is immaterial, but I am using a 6V one (bright sun) so Vf of D2 represents a considerable loss of potential.
I like your idea of connecting through a light-controlled switch, and it certainly scratches my itch about the "wasted" power. Another idea from the forum that has got legs! Thinking about it, I already monitor both voltages so I could probably do this in software - skip the LDR and just have a FET that gets switched in. hmmmmm. This is pretty much what the ideal diode does except the voltages are subject to a hardware comparator formed of the matched pair of PNP transistors. I could replace that side of things with a few lines of code.
I have the ideal diode built and ready to go and at 10mV it is almost the short-circuit that I seek. 10mV out of the panel's 6V (although unlikely) is 0.16% loss - way better than the 5% (probably much worse) I am getting now. I have been watching the traces this morning and it grinds my gears to see dV (VSol-VBat) at just +0.2v and know that nothing is going into the battery, whereas with a "short", the battery would be sucking every last joule from the panel - I aspire to that.
I'm gonna fit the Ideal diode this weekend (or maybe today). I anticipate the voltage traces will become a single line when dV>0 as it will be around a tenth of the resolution of reporting... will tweak that in the next software.
cheers
h
Edited 2025-06-13 19:55 by CaptainBoing