switching PV's from series to parallel

[doug7x]

Ok, folks...hope you can help. I've built 2 PV panels from ebay cells. 1 is mono, the other is poly. Both panels appx 196W. (49 6x6 8A 1/2V cells in series...7 strings of 7 cells) At max. output, I'll get 24.5VDC @ 8A from each. My GTI turns on @ 14 VDC, & it's max input is 28VDC. (300w constant, 600w peak rating).In low light, I'm not making enough voltage to turn on my GTI (both panels are hooked to the GTI in parallel) What I'd LIKE to do is find a switch or relay that will automatically switch my panels to series when light is low (dawn, dusk, cloudy day, etc.), & then back to parallel when the voltage rises enough to keep the GTI on line. Somebody told me that he thought a solid-state charge controller could do this, but he wasn't 100% sure. Thanks in advance....Doug

[Sr WiNdTeCh]

I'm not an electronics expert but from an electrician's point of view, I would get a photo cell (same thing that turns on and off security lights) and place that in the middle of your panels and then run that to a small relay that you have wired up for NC your parallel connection and then the other on your NO connection, then when it gets dark enough your photo cell will turn on your relay witch will switch to series connection... problem is... it may be to dark when this happens... not sure on that.. I'm building my own panels now... could you tell me how you built yours?

[Rover]

If you can perform a test that measures the voltage and current (BOTH) comming from the panels in series,in a low light scenario like you describe (below your controllers regulation when in parallel) (example connect them in series at this time with the ability to disconnect quickly before you fry your controller) , you should get an idea of what kind of gain you will get.

I'm thinking the power that you are trying to recover is so minimal, it may not be enough for the controller, and even if it is , may not be worth the trouble of pursuing this further.

If it is a cloud passing by scenario, I'd be afraid to even perform the test

[Ungrounded Lightning Rod]

Voltage is controlled by bandgap, and resistive drop (both of which are affected by temperature), NOT by light level (though insufficient current to supply the load will make the voltage drop off a cliff).

Light level controls available current.

If your voltage in parallel is too low to trigger your GTI under low light but is fine in high light, it's probably because there's insufficient current to handle just the parasitic load of the device under low light conditions.

So putting them in series, which will be TRYING to buy voltage by trading away current, seems likely to make the problem worse, not better.

[doug7x]

SrWin, I saw a youtube vid on how to build the panels, then I re-designed from there...I'm not using EVA, Sylgard, or glass...just 1/8" Optix acrylic. (this is the stuff they make skylights with) Sa far as the photocell...when it gets dark enough for the cell, it'll be too dark to get any juice out of my panel...

[doug7x]

Rover & Ungrounded...although I never took voltage & current readings once I hooked up my cells, I AM wired through a kill-a-watt meter, & I have connected my cells seperatly to it, just to see how they compared. It seems that my GTI turns on at appx. 50-55 watts, with a min. of 14 VDC. If, for example, both of my panels are making 10V & appx.20W each, my GTI won't turn on. BUT , if I hook them in series, I'd get 20VDC & 40W, & my GTI WILL start. I HAVE tried this & know it works...just can't remember the exact values. I'd really like to find something to do this automatically...(with a time delay... for the "passing cloud" scenario).Think about this...I live near chicago, & there are times I don't see good sunshine for days at a time...that extra 40-50W that I'm not getting now would add up. Thanks for hearing me out & for all of your input...Doug

[Ungrounded Lightning Rod]

Real data trumps theory.  B-)