solar troubleshooting
[originally posted to reddit]
before relo
This morning ~9am I picked up camp to go to the next 14d spot. Started the van and observed solar+alt were charging at ~40A combined (net into bank). I didn’t check the actual mix because it didn’t matter: both were charging at normal combined rates for that time in the morning. {I looked at the stats later and it was roughly 50/50}
in town for mail
I parked at the local library to hork (slow) wifi, updating devices, pulling down media, backing up online data. Dog and I left the downloads in progress walked down main street to the USPS to pick up my General Delivery mail.
Came back and the downloads continued. The library wasn’t actually open today (very small town) so I didn’t feel bad about slurping their wifi.
something is amiss
Before I left for the next camping spot I saw the battery monitor display; 80% SoC and 7A discharge. That’s odd. I doublechecked to make sure the van wasn’t shaded (it wasn’t). Hmmmm.
Pulled up the Victron app on the laptop (linux version) on the laptop and looked around.
The charge controller was OFF due to insufficient panel voltage. 0.3v, so probably just backfed voltage from the controller itself.
Maybe it’s a glitched. I disabled/enabled the controller; no joy. I’ll troubleshoot at the next camp. The alternator was delirious with emotion at getting to do all the charging. Normally it’s in [very distant] 2nd place.
I remembered I’d seen ~0v voltage once before when a tree limb swiped under the panels and physically disconnected the MC4. But there were no limbs anywhere near the vehicle this morning. I stood on a nearby low rock wall but couldn’t see the solar cables clearly on the roof.
Fine, I’ll figure it out at the next camp.
actual troubleshooting
Got to camp. I didn’t want to drag out the ladder and climb up there unless I had to, so I started checking things inside the van.
This particular meter is relatively new to me, so I started with known good first: checked the BATT terminals on the controller. 13.x, fine. Checked the PV terminals: 300V. I did not code brown but my eyes definitely opened wider.
After a few seconds and a few retests I realized the meter was autoranging and had a small mV icon in the corner. Oh. 300milliVolts (0.3V). That makes more sense. Sweet baby jebus where are my heart pills.
I looked at the fuse in place to allow easy maintenance1 and it was broken.2 It shouldn’t be broken based on how little current flows through it, but this is what we’re dealing with.
It was hidden behind my “why didn’t I do this sooner” wire racks so was a bit of a PITA to get out. Whatever.
Got it out and the fuse was intact. What? I put it back in and looked again; from that angle a shadow from my headlamp had fallen across the center span of the fuse where the break would have occurred. A continuity check across the fuse would have saved me wrestling with the hidden fuse, but lesson learned.
So now I checked the panel side of the fuse and NEG terminal for voltage; same, 300mV. So the problem is either on the roof or in the entry gland. I don’t like where this is headed.
I dug out the ladder and scampered up. Started checking connections. They’re all plugged in….. except…
One of them looks weird. Deformed. I tried to separate the male/female MC4 but could not. The reason for the failure became clear (see below) but for now I had to fix it.
repair
Luckily, a couple years ago I bought an MC4 crimper and MC4 assemblies so I could do field repairs. Time to figure out how this works….
I cut the wires off as close as I could to the old MC4 to preserve wire length. I looked inside a working pair of MC4 to figure out which barrels to crimp onto each wires.
{Note: at this point a nondescript pickup with an obvious VHF antenna and parked about 50ft away. The driver got out with a clipboard and headed over. I assumed he was a Ranger recording plates and was hoping I wouldn’t get run off some kind of forbidden site. Turns out he was surveying water levels in water tanks and just wanted to walk up a game trail to check the nearest water tank. }
Crimpy crimp. Assemble. Connect.
I climbed down from the ladder and enabled the Victron again; it saw Voc and quickly ramped up to its full 45A output. I climbed back up to collect tools and broken bits, then put everything away.
the answer to the question
I normally keep my array[s] wired in series. But last summer was brutal and I needed to hide in tree shade. I’d already given my 2x 100w portables to A Worthy Boondocker. So I rewired my panels in parallel.
I’d carried the 3-way Y’s around but hadn’t used them before. The fit was not great and required force to get them on. When I reconfigured back to {series} recently that blog entry mentioned:
Took a while to wrestle off the too-tight connectors, but I was glad they were off. Rewired the panels…
I think either the forcing together or forcing apart damaged one {of} the MC4 so it didn’t fit right internally anymore. My guess is it arced and overheated.
denoument
All is well. I checked put my paws on each set of MC4 to make sure they weren’t warmer than ambient at full throttle, and they were all fine.
I made some mistakes but was only down for about 2 hours, including time driving to the new location.
While writing this post I looked back at the data and saw that the failure occured just after 11am. I dug the paralleling Y’s out of storage and cut them in half with a razorknife so they can’t be reused. Will dispose of them next time I’m near a trash can.
{edited for missing words, typos}
comments
mastodon comment thread for this post
lemmy comment thread for this post