bypass diodes fried

[cardamon]

I have a bit of a conundrum that has me scratching my head.  Background: I live off grid and am an electrician.  The friend I am referring to knows nothing about electricity, solar, controls, diodes, etc.  She bought a some astropower 120 W 12V modules used.  I went over to temporarily hook up two of them in parallel to charge a battery.  Upon inspection, I noticed one of the bypass diodes was cooked.  I set that one aside, grabbed another one and hooked them up to a basic PWM charge controller that I was loaning her.  The LEDs on the charge controller didnt light up so I assumed it was bad since I hadnt used it in years so it seemed reasonable.  Came back later with another one, same thing.  Turns out I was only getting 9 volts from the panels. I determine that this other panel has a bad bypass diode as well.  I think it may have fried sometime between then and when I first looked at because I am pretty sure I inspected it after I found the existing bad one.  I was in a hurry on the way to a wedding so I told her she could swap it out, white to white, black to black.  She calls me later and said it still wasnt working and she was seeing smoke coming off one of the bypass diodes and that she was sure it appears fine before.  So what is going on here?  Is it possible this panels suffered some sort of event that fried the diodes or made them about to fry?  Is there anything stupid she (or I) could have done that would damage the bypass diodes?  I was trying to think of how one could kill a bypass diode by doing something wrong and I couldnt come up with anything feasible.  Any ideas?  Ive been away for three weeks but will be back this weekend so I can take a look at the rest of the panels and try to make sense of what is going on...   Thanks

[Flux]

I think the answer may be that they were used panels.

It is near impossible to kill bypass diodes on a simple small set up with a few panels.

The problem usually comes from large installations with multiple parallel and series paths. If a few panels of a parlallel string get shaded in a set up with a lot of series strings, a bypass diode can be forced to handle several times the pannel load circuit current, Unless the diodes are over rated to cope with this they can fry.

Short of a serious error in connection I can't see what you can have done so I supect you ought to check all diodes on those panels before continuing..

Flux

[madlabs]

I'd add that visual inspection won't always find a bad diode. So they may well already have been bad or damaged. As Flux noted, it's usually in series/parallel arrays that bypass diodes blow, so I'm guessing they were previously beat on.

Jonathan

[cardamon]

Ok thanks for the reality check.  figured it would be next to impossible to destroy one playing with just two panels.  I recommended new and the used ones were  only slightly cheaper......

[SparWeb]

If I was in that situation, I'd just assume the diodes had been stressed in their previous life, and replace them with an identical spec.  Or match up new ones as close as possible without under-rating them.  Any thoughts/disagreements?

[Flux]

Having agreed the panels came with faulty diodes and it's not something you have done, I can think of one more possibility other than failure from shading in a series parallel string.

I recently had to deal with a set of panels where the cable ran close to a tree that had a direct lightning strike. The cable blew to earth and I really had little hope of any panel surviving.

One panel shattered next to the connector box and was useless. All the others blew the bypass diodes short circuit but the panels survived with no obvious loss of performance. It seems as though the best use for bypass diodes may be to protect panels from lightning strikes.

I found some diodes from ON semicionductor that were aimed at this market with a rating of 9A. They were a similar size to the original.

I agree with Sparweb, replace all diodes whatever the condition, they are cheap enough.

One thing I found tricky was that on some types of panel the diodes were soldered to the connector wires before the terminal box and burried in silicone. Rather than risk damage to the fragile looking connectors I carefully clipped the dide leads and put the new ones directly on the connector blocks, just as some of the other type of panel had.

I suppose it is quite possible that those panels could have a milder lightning strike with some diodes lost and they could have been replaced for that reason.

As you say buy second hand and you take a chance, sometimes a bargain , sometimes not.

Flux

[OperaHouse]

I think they hooked the panel up backwards on a battery or had the panel reversed when in parallel with others..

[jack11]

I am not sure if a "fried" bypass diode could result in a short circuit or an open circuit at its PN junction, or if it could be either one, depending on how it fried???

But with a bypass diode shorted, I don't think you'd ever get any current out of this cell (or a string of cells protected by the same bypass diode), because all of it is going thru this short. Then I think the total panel current depends on the balance between which bypass diodes are shorted in that panel, and which are not.

With a bypass diode open circuit, there is no bypass protection, and you may get a hot spot at the shaded cell when you short the panel's leads (or some device/controller does it for you if it operates the panel at a low voltage).

Maybe that can explain why your panels did not put out what you expected.

[Mary B]

Common failure mode for diodes is fail shorted. Open fail only if the surge is enough to vaporize or crack the diode in half.

[Flux]

My experience is that diodes fail short circuit. Only in extreme cases would there be enough energy available to blow it open circuit.

Even with the panels I mentioned that had the lightning strike all failed short circuit.

It is certainly possible that the panels have been connected backwards across a battery, we shall never know if they were bought used.

Larger panels are often split and may have more than one bypass diode so you may stil get half volts. Until you check all the diodes and look for a pattern we shall probably never know what happened in their previous life but if it's only the diodes they may still be a good buy.

Flux

[cardamon]

Yes these are soldered on which is a little bit of a hassle.  Three of the six panels appear to be putting out half volts. Ill be charging her for the repair for sure.  Love it when people try to save 50 bucks and either end up paying that 50 in parts/repairs or  have someone else spend their $50 of time on it. 

[SparWeb]

Yeah...
That never happens to me.

Never...

[fabricator]

I just got done with a lot of welding on a "really great deal" on a trailer off craigs list for a friend of mine, my work cost almost twice what he paid for the thing, and it's still an old rusty trailer.