Bypass diode can cause panel to blow???

[TucsonJim]

Ok guys I got some answers. I had a long discussion with Jesse McNeeley today, he started the company with his dad Fred. They do recommend running their panels in parallel. Because they have no control over the number of panels someone can run in series. The specs on the diode in my panel is "THY2550 2cz20r45 25A 50V 20A 45V MBR2550 Schottky diode”. Apparently, the increase in voltage can overheat the diode and can cause a fire because the IP68 box isn’t rated for the higher temps. The diodes originally in my box would handle my 2 panels in series, 3 panels would exceed the 50volt of the diode’s specs. Jesse is willing to change out my boxes in April when I stop at his shop.

Ed - Thanks so much for the follow up. That makes sense.

Jim

[lobsters]

Ok guys I got some answers. I had a long discussion with Jesse McNeeley today, he started the company with his dad Fred. They do recommend running their panels in parallel. Because they have no control over the number of panels someone can run in series. The specs on the diode in my panel is "THY2550 2cz20r45 25A 50V 20A 45V MBR2550 Schottky diode”. Apparently, the increase in voltage can overheat the diode and can cause a fire because the IP68 box isn’t rated for the higher temps. The diodes originally in my box would handle my 2 panels in series, 3 panels would exceed the 50volt of the diode’s specs. Jesse is willing to change out my boxes in April when I stop at his shop.

Ed, thanks very much for getting this clarification! Makes sense, and good to know the limits.
-Al

[howson]

Al (and everyone else),
I received Renogy 100W Eclipse panels yesterday and spent some time examining them along with how they are put together. I don't know if any of what is written below will make any sense to anyone but me, but it helped me to understand why the Eclipse has bypass diodes and why they are soldered where they are.

If anyone identifies a factual error please let me know and I'll fix it ASAP.

Let me emphasize the drawings are over-simplified to convey concept and impact, not to teach electronics or account for every source of power loss in a circuit.

First, the panel itself. The purpose for the first drawing in the graphic below was to understand how the system was wired together. As noted, the Eclipse is a 32 cell panel constructed with 4 columns of 8 cells. The diagram of the cell interconnection (I used a battery symbol for each cell) is an educated guess. The shot of the diodes was shot by me of the actual panel.

The rest of the slides were assembled without a lot of techno mumbo-jumbo with the hope of conveying what was happening with limited techno-speak. Those slides will follow next.



The first thing I wanted to understand is how I'd expect the panel to perform without diodes. My goal was to make the diagram easy to grasp--I'd appreciate some feedback (from anyone).



Now what would I expect if one of the 16-cell strings encountered shade? Note (in my over-simplified circuit) that power drops to 0 from the entire panel.



Now with the diodes--starting with ideal conditions.



Now what happens when there is a shade issue?



...to be continued...see next post for the final picture and conclusion...

[howson]

...continued from previous post...



Conclusion: Without the diodes, shade will impact the output of the entire panel. With the diodes there is a chance of 1/2 of the panel still producing power.

I did not expand these drawings to a series of panels--if anyone requests it I'll be glad to show what I think would happen. I also did not show the effect of full shade. The output is expected to be 0W. (Again--this is an overly simplified concept post! I suspect when I test my panels in real sunlight and shade I'll still get measurable output. Those results will be posted in my solar install thread.)

-Howard

P.S. If you really want to know about schottky diodes, I found this Afrotechmods video very helpful: