Really odd problem...

Flint

I have a string of nine 280W panels, but the voltage adds up like seven panels. If I remove the last panel from the string (so there are only eight), the voltage goes up to what it should be.

Any ideas?

westbranch

Pls add more PV details, make, model, specs, wiring,  etc

Flint

Winaico WSP-280M6.  280W - Voc: 38.45V - Isc: 9.44A - Vmp: 31.37V - Imp: 8.93A

Wired in a string back to the combiner.  There are four strings, and three behave normally, while this one exhibits the behavior I described.  Swapped the wires between strings, and the problem moved (ie, it stayed with that string of panels, not the wires), so it's definitely a panel issue, not a wiring issue.  Tried bypassing each individual panel, and when I got to that one, the voltage actually went up when it was removed.

westbranch

Is it possible that the polarity is reversed? Or that there are blocking diodes mis-installed?  Just thinking outside the box....

BB.

Probably a failed panel... On occasion, it can be a shorted bypass diode (inside the J-Box)--But many newer panel J-Boxes cannot be taken apart to change the diodes.

Otherwise, look closely at the panel and see if you find any differences (delaminating, discoloration under glass, corrosion, etc.).

Taking the panel and shorting just its output to measure Isc will frequently show problems (full noon time sun, current >~0.5x Isc, panel may be OK--Compare Isc of other panels under the same sun--The Isc of good panels is usually pretty close).

Similar measure Voc (voltage open circuit) of the panels--Can find shorted diodes in panels.

A relatively inexpensive DC current clamp DMM (like this one from Sears) can make current measuring much easier (and safer too).

http://www.sears.com/craftsman-digital-clamp-on-ammeter/p-03482369000P

-Bill

Photowhit

So I can't figure out how to just start over, gosh I liked when I could just write my own HTML!

Flint

Voc is the same as the others.  I'll have to see about getting a DC ammeter - mine's AC-only.  Until then, is there anything else to check?

I made some adapters to reverse the polarity, and that lowered the voltage of the string even more, so I'd say that's not it.

I opened the junction box and visually checked the diodes - there does not appear to be any damage, and they look like they're installed facing the correct direction.

I forgot to mention that this panel is newer than the others.  Same model, but it comes from after they switched to MC4 instead of Tyco, so it's definitely a different production run.  Not sure if there could be any incompatibility, somehow?  I just tried swapping in another panel (also from the newer run), and the problem persists for that one, as well.  Is it possible that the new run is somehow different from the others?  Or is there something else I should be checking with the rest of the string?

zoneblue

PW, you can edit the html by clicking the right most button on the tool bar.

So I can't figure out how to just start over, gosh I liked when I could just write my own HTML!

Flint

Any thoughts?  This one is just not making any sense...

Saipro

Frankly, I can only think of reversed polarity issues - and the diodes should fix that/prevent flow of current altogether. If you've got the time and other resources, try it in parallel with another PV panel from the old batch and see if the current output is as expected.