Well it seems the issue I was having with that one solar panel string has returned. So I am troubleshooting the issue and hope to resolve it soon. Piece by piece I am testing with the tools I have, and by a process of elimination I am confident I can resolve the issue. With some of the tools on hand I can test, but eventually I need some better tools to figure it all out.
First I checked the voltages at both ends of the cable. It is stable and all is fine, until you try to apply a load to it. Then it all goes wrong. The solark cuts off power to the string and throws an insulation error. Luckily I have two strings and soon a third string to utilize. So this leaves array-c in a degraded state but not offline.
We can see I get 286V DC readings into the building.
A lot of testing has occurred in the last few weeks. I tried plugging in the string that is acting up into a different channel on the inverter and it right away acted up. Making me think its not an inverter problem, but something out there by the panels.
Since I am running two paralleled groups of panels. I must separate them for testing. I also tried removing the splitter and running just a single circuit but the issue remained. Frustratingly for both sides. Making it seem the issue is in multiple places.
Removing power from the wire, and then connecting them together I can do a continuity test. This will tell me if there is a break in the wire going underground into the building. It is possible ground borrowing animals damaged the wire.
Though upon testing the wire it comes back fine. Producing 0.3 Ohms, right around where it should be.
I tested a second wire that is buried and it came back with the same results.
So what's next?
Since I have only checked some of the PV connectors. I really need to check them all in the degraded string. Last I checked the home-run cables and they seemed fine. But there are some others that need to be looked at and give them a good tug to see if they are lose at all. Maybe disconnect take off the cover and check for corrosion. It may be worth replacing the circuit breaker as well, never know if that is causing the issue but unlikely.
And if I cannot find any issues there, I need to buy some specialized equipment. There is a device called a Megaohmmeter or also known as a Megger. This device will run a high voltage current through the wire and grounded frame to check for leakage. I have never used such a device, but luckily an electrician I know does. So if I can buy the device for around $250 I should be able to do some further testing.
Another idea is to run current through the degraded circuit, and use a thermal camera to look for hot spots in the connectors, showing a possible issue. Possible I could do this in
But I do hope its just a lose wire I find upon physical inspection. As the two options above require running current through the wires in ways I have not done before. So hopefully the most simple of fixes is the one that works.
I had a thought, it is possible the ice that started all of this weighed down a lose PV connector eventually making it pull out enough to cause these problems. So next time I can get out there at night, I really need to check every connector and see if they are lose at all. During this during the day is dangerous, so it is something I must do at night.
Also I bought a heatgun and some heat shrink insulation, so I plan on adding that too if I find any corrosion.