Guru
Joined: 03/06/2019
Location: AustraliaPosts: 1584
| Posted: 12:38am 17 Jan 2025 |
I always felt bad that I could never get this going being that Tony designed it mainly to help me out. Don't like to waste peoples time when they are good enough to do something for me which seems like it wasn't appreciated which was the complete opposite. Simple as it is, was too much for me to figure out and felt I let Tony down.
I can say that I learned a lot from this though.
As perhaps inappropriate as it may sound, I'm glad others that are well ahead of me also had trouble with it. Obviously more to it than meets the eye.
I did a couple of variations on the principle which Tony and I discussed.
The first one was using an arduino with a modified "Blink" sketch to just pulse the MOSFET. I realised that the timing was not really important nor was the voltage levels. As long as the panels weren't dragged off their VMP or at least more than 10 % off it, the rest was pretty small potatoes.
Also Realised as long as the caps weren't allowed to get Full and therefor waste solar power, voltage and timing wasn't real critical either.
I played around with the timings for the cap bank by putting a 500W halogen light as the load. I could see the circuit simple as it was did work because I got next to nothing with a direct connection and an actual amount of useful light when I connected the pulse Circuit.
I played with the timings on and off till I got the lamp to glow the brightest in good sun and then went a bit higher which had no effect which was allowing margin but more importantly, it worked great in the rising and falling light.
And that was it. The thing runs all the time, initially from a small USB power bank thing which lasted days anyway. For simplicity I ended up solar powering it with just a big fat cap same as I used for the main storage as bit of a battery with a voltage controller. .
Always felt that using an arduino was a waste so I bought a pre-built timer that went down to Hz for $12 I think it was and used that as the driver instead. Had the advantage of being able to ajust timings easier although I used the ones from the arduino and that was it.
All it does is turn the fet on and off and that's all that's needed. The caps pose virtually zero impedence on the panels so they work efficently.
The power in the caps does not really matter as long as they don't fully charge and leave no more room for the power the panels are generating. Even if they are 1/20th Full, all the power gets dumped into the load so there is no more loss than if they were 98% full.
The aim of the game is not to drag the panels to 7V or whatever where they become highly inefficent and keep them in their happy place.
This very crude and basic setup seems to do that. Also don't matter what size the cap bank is. You just run a higher frequency to dump the load faster.
I have a pretty decent cap bank to allow for plenty of panels at slower timing ( around 50Hz I use ) and I don't even feel the FET getting hot.
I would have liked to get this circuit working but it was beyond me but it did give me a more through understanding of what needed to be done to make the idea work and it seems my even more simplified soloution works just fine.
I am also now looking to use the very same Circuit as a Battery Desulphator.
They just pulse the power into a battery prefrably at high amps and voltage and this self same circuit is perfect for that.
Even if I slow it down enough to get a full hit every time and loose some power, so what? It's the electric Vibration I basicaly want and this fits the bill perfectly.