Panel issue, or??

[Texas-Mark]

So I have some Renolgy 100W panels that I have had in service for about 10 years. They are on the roof of my shed. These are set up as two panels in parallel (joined at the panels with Y connectors) in multiple "banks." The two lines from each bank come into the shed to a combiner, then to my charge controller. Everything was working fine for years. Then a few years ago we had a hail storm and bank #2 started acting up. I would have no current flow from that bank (measured directly across the positive and negative lines coming into the shed). Then suddenly they would start working again for a few days or even a few weeks. In other words they are very intermittent. I checked the Y connectors (no easy feat) and they seemed fine. Note: Even when I measure no current, I will still read 18V across the lines (while disconnected from the combiner).

So here is my question. Since I have two panels in parallel, other than the Y connectors, is there any individual panel malfunction that would cause this behavior? I had assumed one or both panels had suffered a hail strike that was making one of the internal panel connections intermittent. But I am not sure how one panel would affect the other in this way (no current at all). My next step I guess will be to separate the panels at the Y, and run them individually into the shed to see if it is just one panel acting up, or both. This will be a PITA to do.

 

[Alphacarina]

Then a few years ago we had a hail storm and bank #2 started acting up. I would have no current flow from that bank (measured directly across the positive and negative lines coming into the shed).

"Acting up" - Exactly what does that mean? How are you measuring current flow?

When you disconnect a bank of two paralleled panels from the load, you should measure voltage (18 in your case) but there would be no current, because you have them disconnected from the load

Don

 

[chrisski]

How do the panels look? Visible Hail damage?

Where did you get the 18 volts 0 amps? Did you trace that back to the panels on the connector?

Another reason to get 0 amps on the ampmeter is the internal fuse being blown. You mentioned 2 panels in parallel, so this could easily blow a fuse internally. Can you remove the fuse and check it? My first DMM is rated for 10 amps, but would blow the fuse any time I measured amps and read 0 amps.

If the fuse is fine, can it read amps on anything, even a resistor on a AAA battery? If the DMM works, recommend undo connections and measure amps on the way back.

The last thing that could go bad on a panel is the diodes on the back. They should be really low ohms one way, and open the other way.

 

[Texas-Mark]

"Acting up" - Exactly what does that mean? How are you measuring current flow?

When you disconnect a bank of two paralleled panels from the load, you should measure voltage (18 in your case) but there would be no current, because you have them disconnected from the load

By acting up, I mean they work fine for a while and then it doesn't Since it somewhat seemed to be temperature related, I always thought it was something in the actual panel was bad (like a cracked trace), especially since it started right after the hail storm.

I am measuring with a Fluke on Amps directly across the leads. All other banks will give me around 11 amps (I don't do it in full sun as to not blow the fuse in the meter). This bank will read zero. I read 19.7v across the good banks and also the bad bank.

The more I think about this, the more I think I have a bad cable (connector) coming off one of the Y connectors that has a lot of resistance which will still give me an open circuit voltage, but no current. If one panel was dead (open), I should still be reading the other since they are in parallel. If a panel was shorted, I would have bigger problems. The odds of both panels being intermittent is slim.

 

[Texas-Mark]

How do the panels look? Visible Hail damage?

Where did you get the 18 volts 0 amps? Did you trace that back to the panels on the connector?

Another reason to get 0 amps on the ampmeter is the internal fuse being blown.

No visible damage.

I measure that at the end of the line in the shed. Getting to the Y connectors is not easy (but I know I am going to have to get up there).

Not a blown fuse as I get good readings on the other banks (when disconnected from the combiner).

 

[PVorBust]

If you've got 2 (or more) panels connected with a Y and you have no output, either:

• both groups of panels have failed
• the failure is below the Y

I agree with the suggestions above that Volts, but no current is highly suggestive of a blown fuse.

 

[RCinFLA]

A common stress failure is cell interconnect solder joints due to repeated temp cycling which is normal to panel environment. This can show up as intermittant connection.

Sometimes conductive dendrites grow between cells and panel frame. Its root cause is perimeter sealing failure allowing moisture intrusion. The leakage due to the dendrite can vary based on humidity and recent rain fall.

Check for corrosion in panel junction box where you will find bypass diodes.

Also check any MC4 connectors for corrosion. DC voltages and moisture create a lot of issues. Just ask Ma Bell telephone personnel when you see gas cylinders of dry nitrogen near their junction boxes.

Since you have both panels's wiring brought down you can separate them and check Isc with ampmeter shorting across each panel. It will do no damage to panels and will show up a drop out if there is an intermittant connection within a panel or cabling.

 

[sunshine]

How many panels do you have. 2 in parallel or 4 in 2p2s.

I read 19.7v across the good banks and also the bad bank.

Suggests just 2 panels in parallel

All other banks will give me around 11 amps

I cant work this out when you have 100w panels and no output from one (or one bank)
The current should be 5'5 amps though each side or 11 combining the sides.

Even so I go with the others on bad mc4s !.
My own experience as well as a lot of posts that start with- ' I've had my system a year (low voltage) was working well but now it aint!!

 

[Texas-Mark]

Problem solved (well, found). The output of the negative Y connector and negative MC4 connector is toast. I had to cut the tabs off to get it apart. I cleaned it up as best I could and now I have current. Luckily this one relatively easy to get to. And the cable is only about a 4 foot run.

I may have looked at the wrong one when I first checked the Y connectors. Or maybe I could not see inside the female port enough to see the beginnings of the corrosion.

 

[Texas-Mark]

I cant work this out when you have 100w panels and no output from one (or one bank)
The current should be 5'5 amps though each side or 11 combining the sides.

Just to clarify, I have 8 pairs of panels (each pair of panels is in parallel). All of the Y connectors are up by the panels. So each pair should be giving me about 11 amps at the combiner box in the shed. That is what I was reading on the good pairs.

As for the blown fuse theory, as I noted, I was comparing to good banks (pairs) so I knew the meter fuse was good.

 

[sunshine]

Just to clarify, I have 8 pairs of panels (each pair of panels is in parallel)

That clears everything up.

I am doing away with MC4 connectors. They are convenient but seem to present problems after awhile with low voltage systems and where the connectors are from different companies due to dissimilar metals used.....and my crimping ability...

 

[PVorBust]

I am doing away with MC4 connectors.

What will you use instead?
I bought a batch of 20 MC4 pairs only to discover the inserts wouldn't seat in the shells
The next batch were fine, but I wasn't happy about it (naturally I was on the roof when I found I had a problem)

 

[12VoltInstalls]

What will you use instead?

I get that!

Sure some connectors fail in time but I’ve gotten to the point that crimp/solder joints can be more of a pain than just keeping extras of ‘x’ connector on hand.
The vaseline-like greases don’t seem to last, and while I’ve been annoyed with the gunking up of stuff, fluidfilm seems to still be nasty in three years and doing it’s job. Ymmv

btw a bad connection will often read full voltage. Merely a difference in potential- not an indication of good continuity. So while voltage is information- ohms is evidence especially if you do a voltage-drop test.

 

[sunshine]

What will you use instead?

6mm crimp lugs to bolts through a pvc board pop riveted to the panels frame. 12v system 60amps in places.