solar budget of 2700.00 Canadian

Salvo,

Diagram is nice but reality is the diodes are across 20 to 24 cells that produce ~1/2 volt each. Not sure that mattters.

I understand the on or off nature of the diode but I am still not convinced about the results of partial shade.

A compelling case for how bypass diodes work or don’t work is indicated by the typical panel AV curves. They show full amps for 1000W/m sq and about 20% amps with 200W/m sq irradiance with no voltage difference. This means that as the light levels drop the amps drop and the bypass diodes are inactive.

It seems reasonable to me with certain types of filter light like some trees parallel panels which add amps may supply more power than serial panels which will be limited by the smallest amp panel unless very low light levels cause some bypass diodes to activate. Clearly there are different types of shadows like “soft light” vs. partial full shade like from an AC shadow.

The basic shadow differences come down to where the panels are mounted and where you park relative to shadows. If you have little shadow difference for parallel vs. serial then serial would involve less wire length, smaller gauge wire and no need for a combiner box.

For RVers who like trees (and who doesn’t) parallel panels may have an advantage but the best advantage would be portable panels mounted in full sun.

Take a close look at the "Typical Application" diagram in your link. One bypass diode is across a series of 8 PV cells. Any small amount of voltage produced by those cells will back bias the bypass diode. The diode will not conduct. The larger current from the other cells are now forced into the shaded cells. Forcing the current through the shaded cells will generate a voltage drop. If the forced voltage drop is large enough and shaded cell voltage generation small, the bypass diode can start conduction.

In your testing, if you cover up cells (total darkness), the bypass diode will conduct. On the other hand, if you tilt one panel 45 degrees away from the sun, the diodes will not conduct (as shown by Gale's testing).

Sal

smkettner wrote:
I don't see how partial shade could be worse as any excess should pass forward and any minimal production will allow some power to go through the cells.

http://www.ti.com/lit/ds/symlink/sm74611.pdf

Still more reading to do.

smkettner wrote:
Forward drop of a Schottky is about 0.4v and you push through three in full shade for a total of less than 2 volts lost. So my 2 panels still in the sun producing 7.5a and 60v will only put 7.5a and 58v to the controller. This is a loss of 4% in addition to the third panel not producing. That is the theory to test anyway. I don't see how partial shade could be worse as any excess should pass forward and any minimal production will allow some power to go through the cells.

http://www.ti.com/lit/ds/symlink/sm74611.pdf

Still more reading to do.

Clever. Their use of a charge pump makes sense.

Jim

Forward drop of a Schottky is about 0.4v and you push through three in full shade for a total of less than 2 volts lost. So my 2 panels still in the sun producing 7.5a and 60v will only put 7.5a and 58v to the controller. This is a loss of 4% in addition to the third panel not producing. That is the theory to test anyway. I don't see how partial shade could be worse as any excess should pass forward and any minimal production will allow some power to go through the cells.

http://www.ti.com/lit/ds/symlink/sm74611.pdf

Still more reading to do.

Sal wrote:
I'm no solar expert

Sure had me fooled:B.

Yes, that's how I see it, but I'm no solar expert. My knowledge is from testing solar myself and analyzing other people's results.

Salvo wrote:
If one panel is shaded it will still output current. But the shaded panel current is significantly less than the others. The diodes in the shaded panel will not conduct (the panel is producing power). Due to the series string, the shaded panel will significantly reduce total power production.

The parallel configuration will be much better.

Interesting. I think I understand...paraphrase.

If the cells are Partially shaded, they will be producing voltage/current, but less than all other series cells. However, their voltage/current Is sufficient to keep them connected in series, the diodes Don't bypass the weak cells. Therefor, the weak shaded cells act as a "hotspot" and series voltage/current flows Into the shaded cell in an attempt to equalize the voltage/current across the entire series of panels.

In series, the max current is the max conducted through the shaded cell; which is actually consuming power.

In parallel, all other cells/panels voltage/current flow through large diameter low resistance parallel wires, so the parasitic drag of a shaded cell is far less consequential.

So far I have stuck to home run installs rather than junction box. It would be fun to have a serial parallel switch to be able to flip flop at will and comapre under various conditions. If I found serial was always better with my panels I'd direct connect one wire from each panel to save the extra run, but for now I like the flexibility and ability to try every permutation.

Jim

The test you linked was never completed. Also, covering up one or two cells is not how shading works. In the real world, shaded cells still get light. If the shaded cells get enough light then the bypass diodes won't conduct. Now the series system is in deep dudu.

I believe you did some testing by covering up cells with a blanket. That test will make series look good.

Sal

2oldman wrote:

Salvo wrote:
The series vs parallel debate is not as clear cut as some think.

No, it isn't.

Series solar testing

In my case of 6 panels in series I see no drastic power loss with one whole shaded panel. But, again, that's MY case. YMMV.