Bad diode in panel?

[desmo907]

Adding to my other thread regarding 2 of my 21 panels reporting abut 65% of what other panels report (there is no shade, etc.).

Seems the replacement enPhase micros didn't help so it may be bad panels.
Finally talked to ConEd and he mentioned it may be a bad diode and some panels are divided into 3 parts (each with a diode). This seems to make sense as my panel production (on the 2 bad) are about 65%.

They will quote me a service call to maybe check/replace them.
(I had a call into Phono about warranty but wonder if they can find 8 year old replacements to match my others.).

So I was just wondering if this made senses (about bad diodes)?
Thoughts?

[Mike90250]

Panels do not need diodes in them to work, so I doubt a bad diode is causing the trouble.
Diodes exist to allow some power to continue in the system if shade covers a portion of a panel, it's a bypass diode.

[bob-n]

I think Mike is saying that a common panel failure cause is an open diode. If you're getting less than normal power from the panel in bright sun, it could be a shorted diode. That's less common. It could be other things, too, but it sure sounds like an in-warranty failure.

Diodes can be replaced. But to replace a diode means taking down the panel, taking it apart, breaking weather seals, and then measuring. The labor cost of doing this would probably equal the cost of a new panel, and there is a good chance that they will find the panel too difficult to repair. so I'd be surprised if any service outfit would bother testing diodes in the hopes of saving a few bucks on the replacement panel.

Yes, over the years, panel dimensions and appearances have changed, but it may still be possible to get a very similar replacement. If you can get warranty support, that's their problem, not yours.

If you can't get warranty service and you're doing the work yourself, measure L, W, and H for the panel, count the number of cells, and start calling the local dealers. Or you may decide to live with it. After all, 65% power for no extra money is not too bad.

Gee, that's another data point for microinverters being more reliable than panels.

[desmo907]

Thanks. Will see what quote ConEd says (and if they can guarantee a fix). ConEd guy said it was just a connection box on the back ot the panel to open to access the diodes. Hmm... Will also wait to see if I hear back from Phono contact. I guess I can live with it too..... haha.

[Mike90250]

In full sun, the only way a panel looses power from a diode, is if it's "shorted". That's the initial failure mode for most semiconductors, they short, briefly, then fuse open forever. Because a PV panel has limited power, it may take weeks or months to "fuse" a shorted diode, but usually it's never noticed.

[desmo907]

Thanks.
Is it true about some panels are divided into 3 parts (each with a diode)? It is a bit coincidental that my 2 "bad" panels are about 65%.productions of the others.

[solarix]

I bet your problem is a bad solder joint in the junction box where the three sections of the panel come together. One section is not making connection and its diode therefore is conducting - bypassing the bad section so that the other two can work. Dig out the RTV sealant in the junction box and examine the solder joints. I've seen this many times. What brand are these panels?

[desmo907]

I bet your problem is a bad solder joint in the junction box where the three sections of the panel come together. One section is not making connection and its diode therefore is conducting - bypassing the bad section so that the other two can work. Dig out the RTV sealant in the junction box and examine the solder joints. I've seen this many times. What brand are these panels?

Thanks.... I assume by "junction box"you mean that's on the panel itself (where the connector goes)?

They Phono Onyx 235-270 dark - installed in 2012.
It might be cheaper if Phono would replace them but still waiting for a return phone call..

[solarix]

Yes, the j-box on the panels. I've run into this situation three times - and good luck trying to get the manufacturer to do anything for you. I end up fixing the panels myself as that is easier than hassling with the manufacturer - but then I'm a hands on type of guy. The brands I've had trouble with btw are Suniva and Canadian Solar.

[desmo907]

Yes, the j-box on the panels. I've run into this situation three times - and good luck trying to get the manufacturer to do anything for you. I end up fixing the panels myself as that is easier than hassling with the manufacturer - but then I'm a hands on type of guy. The brands I've had trouble with btw are Suniva and Canadian Solar.

ConEd is supposed to send me a quote for a truck roll. I suspect the diodes are cheap and they should have a selection with them. Guessing most of the cost is removal of panel.

BTW - could using a snow rake cause any issue? I am rather gentle but wondered about vibration, etc. . I would expect a solder to be pretty strong.

[Mike90250]

Solder most often fails from a poor initial soldering. Then what happens is with daily thermal excursions, the joint is stressed and finally cracks . When you open it up, you can often see if a diode fried or a solder joint went bad, but testing each diode, ( once exposed) is pretty easy with a DVOM.

[desmo907]

ConEd called but thinks the diodes are in an epoxy and not accessible. I look at the Spec sheet for the Phone Onyx Series 235-270 and it mentioned Diodes: 6 pieces Shottky by-pass diodes. He will verify if they are accessible or not. (Spec sheet also mentioned a Junction box =IP 65 rated.)
Still no call back form Phono contact though ):

Wonder if I can find old stock online????

[bob-n]

The dimensions of the Phono Onyx (1640mm x 992mm x 45mm) are pretty-much standard, except for the thickness (modern panels are thinner). So you can replace that one panel with another brand panel easily, but you'll have to deal with the difference in height. We've talked about shims to raise a thin panel here in the past and agreed that it can be done, but not cavalierly.

[PNPmacnab]

I had several bad diodes after a lightning strike. One bad diode will reduce open circuit panel voltage by 7-10V. These were bypass diodes that shorted out several cells. Those cheap diodes saved my panels from being destroyed. Replacing microinverters makes me think you had an event..