Both reasons are a quite good example why it absolutely never hurts to watch for abnormal data and try to identify the cause.
A bad MC4 connection (also can happen if moisture gets in) can turn into an actual fire hazard over time if it remains undetected.
I’ve purchased a 150 bucks thermal usbc cam for the phone and twice a year look at everything under full load. It may not be +/- 2°C precicse but it will for sure show outliers before they reach critical temperatures.
Big Advantage on thermal imaging is, you don’t have to disconnect flawless connections for review and threfore not risk to unintended turn it into a bad connection on reconnection.
Topic drifting slightly, probably worth splitting to its own subject. I use a flir ONE pro, has high resolution, was $400-ish. It is also used to check for hotspots on panels, nothing like a bit of severe hail to damage a panel. We have found numerous faults developing, simply by comparing the readings between similar connections and checking wiring in places that aren’t always very accessible.
It really is a critical tool to own.
I purchased the MobIR Air infrared camera about 4 years ago when I did a major update on my solar system. The resolution is absolutely fine, but for me also a “better” frame rate was important, as some of this cheap cameras had sometimes framrates with only once a sencond or so… For less then 200€ an absolute Must-Have gadget.
Funnily, first I scanned with the camara was also my dog.
Exactly that’s why it doesn’t need to be super expensie and high precise as all.
When you look at 60 Connectors, all displaying for example 35° - and than there’s the one showing 45-50 - You know there’s something wrong, no matter the precision of the absolute values.
And it sometimes can unveil things you never would expect or be able to notice otherwise as well: For my AC-Distribution, at first I used some Snap-In-Terminals. Designed for 10mm² wiring, Rated 50A, 18mm long Ferule as the datasheet specifies - nothing to worry, right?
Yet, they went up to 50,55° under load. That’s not critical, I just disliked it
So, I replaced them with some good old screw terminal blocks - and they barely raise above ambient temperature with the AC currents I’m seeing.
Just one thing to mention: There are a lot of apps available, that claim to turn your smartphones builtin camera into a thermal camera. That’s not working, they just “guess” based on color and brightness of something how warm it could be.
Indeed, for heat sources like bad connections is quite enough something simpler.
But it is fun when you can see even metallic fixing structure behind a drywall with a high resolution camera with 0.01 degree sensitivity…
It opens a new level of diagnostic for everything…
That little voltage spike there, in December with loads of snow on the panels, certainly raised an eyebrow. The only freaking set of MultiContact PV cables that were supposed to be ‘factory’ crimped. The other seven were good though. This bugger alone justified the purchase.
If you are going to do thermal imaging. Two things to remember. (1) these cameras are generally set up for an impassivity of 0.95 which is great for cable jackets, paint on walls, and such. Metal has a much lower emissivity and will show up way cooler than they actually are. To get around this I always put strips of electrical tape on any metal surface that I care about as it has an emssivity of 0.95. Second when you want to see the heat source on heavy cables (e.g. 4/0) you need to take several images in succession right after you start a high current operation. Assuming that the attached image makes it. I shows the thermal transient on some 4/0 cable after turning on 175A of current flow. It shows that the heat seems to be supplied connection of the lug and not from the Class T fuse that is in the middle. Also there are little squares of electrical tape on the bolts and on the lug itself. In the lower image the light green in yellow is where the metal is peaking out between the heat shrink and the black electrical tape. The temperature on the metal is indicating about 10F lower than the lug actually is.
Good find, but please keep in mind, that the spread between min and max measurement is very low, so every single degree seem to be a hotspot.
These cameras often orientate on the min and max wirhin the measurement frame.
In your picture there is a difference of “only” 3°K, so this might be a point of interest for a later measurement under load to see if this really becomes an issue or not.
Thanks. We did indeed. And we found the problem. The wire had been stripped a whole lot longer than normal. Sticking out into the female connector. Blocking the male pin from entering. Flares on the side were gone. Only a short piece made contact.
Thanks Tom. I already found out, the hard way. Thought to have located a problem in my heat pump setup. Called the big guns. Guy asked me: where are we looking at? Showed him the ‘problem’. He explained it, but was glad that I called. Said. Listen. Do you hear that whirling sound? That’s a problem. Well. Every new day another lesson learned. A good thing in my view.
oh by the way. I now also got a roll of those little pad stickers. Thanks!
