Backfeeding panels.

706jim

My original system consisted of 7 ancient Arco panels in series with VOC of about 42 and 4 Astropower 12 volt panels in series parallel with VOC of 20.7 for each of these panels. Well, 20 years later, I've expanded my system with 3 Canadian solar panels 320 watts with VOC somewhere in the low 40's. (Label has faded).
The original system was located on my cabin roof and I've removed these and relocated them to the ground. So the "old" panels (approx 430 watts) are now pointing south and the "new" panels are pointing west. With this arrangement, I can avoid overloading my Trace C40 pwm charge controller as the "new" panels kick in when the "old" panels come into shade.
For a lark, I installed a breaker and ammeter in series with the feed from the "old" panels and realize a current of 4 to 10 amps as the sun swings by.

However......Today I noticed that when the old panels were shaded, the new panels were backfeeding the old panels to the tune of 3 or 4 amps.

Not what I intended!

FWIW, the open circuit voltage of the Astropower panels is only about 36 whereas the Arcos and the the Canadian solar panels are well into the 40's. (My system is 24 volts nominal) 

So, what is the solutions here: Some sort of diode in the feed from the 36 volt array? Scrap the old panels and replace them with more Canadian solar panels of similar capacity?

I'm open to suggestions.

My system was expanded to be able to power and electric fridge which has been working just great so far despite my ancient Trace inverter.

One other solution would be to scrap the old panels, buy more Canadian solar panels and an mppt controller. That said, would it work effectively with some panels in the shade while others are in the sun?

mike95490

Diodes are like 1 way valves  - but with a penalty - Voltage loss and heat

Diodes have a forward voltage rating  Vf when they are biased and conducting,   you multiply the Vf x Amps going through the diode and you get the watts (heat) that MUST be dissipated.

A common silicon diode has a Vf of about 1.2V, a Schottky diode Vf 0.6V      Lower voltage = lower heat

At 10A, you will need a heat sink, as this is continuous power.

Silicon   1.2V x 10A = 12Watts

Schottky  .6V x 10A = 6W    half the heat of a silicon diode

https://en.wikipedia.org/wiki/Diode#Forward_bias

A quick look at some Stud Mount  100V  20A Schottky diodes should get you in a decent ballpark

here's one for $0.60, but you need insulator & heat sink too

https://www.mouser.com/ProductDetail/Diodes-Incorporated/MBR20100CTF-E1?qs=sGAEpiMZZMtQ8nqTKtFS/AdJbs6Ht01BlvFofXwj8LQuWQQmjZq%2bFQ==

  it won't work for more than a minute or two without a heatsink.

BB.

With an MPPT controller--Your best bet its to disconnect those old panels. There is no good way of placing the old+new in series/parallel for an efficient system (it sounds like--As always, details matter with solar power). If the panels are working OK, they really are probably too far apart in Vmp/Imp ratings. Need the Vmp/Imp for both the Canadian solar and Arco panels (or at least, if you can count and see how many cells are in each panel--I.e., groups of 36, 60, or 72 cells are common (Vmp~18/30/36 volts respectively).

If using PWM controllers, really Vmp-array needs to be >~35 volts to reliably charge a 24 volt lead acid bank. Even if you have higher Vmp panels, it should not hurt anything... The panels will simply output (a maximum of) Imp to your PWM controller (MPPT controllers try to match Pmp/Vmp/Imp and having "different Vmp panels" can cause confusion, or if too low of Vmp, simply not supply useful current for the lower Vmp panels).

If you can get a good deal on "matching" panels and retire the old panels (or sell them to somebody that can use a small system), would not be the worst thing in the world.

Adding series blocking resistors--As Mike says, just more things to worry about (electrical insulation, heat sinks, weather proofing) with "Extra Diodes". If you have that much reverse current flow--I start to wonder how good the old panel cells are (they are just diodes and should block a reasonable amount of current/voltage).

Newer panels should have bypass diodes (they go around ~12-24 cells at a time to bypass "dark/shaded" cells to prevent damage.

If you have older panels designed for 12 volt systems, they may not have bypass diodes--And without bypass diodes, "dark cells" can be damaged when several panels are placed in series and back-fed from other panels and a few cells are in shade (basically over voltage the diodes and cause junction breakdown).

-Bill

706jim

Thanks for the comments. I think the best thing to do would be to replace the old panels with similar 72 cell panels and retire the older ones. The cost of the old panels was very high in comparison to what I can get nowadays. On seeing the results of my present setup, I am very happy that I didn't put the old panels up onto my newly reshingled roof.