I noticed a couple of posts from others showing some IR images, and thought it would be comforting to know how hot everything was running, without having to stick my finger on it.  
So I purchased a HK micro camera which plugs into the phone and has a manual focus ring.

After recieving it, I ran the inverter at about 3500w for 10ish minutes and then opened it up to see what would be revealed.  (I doubt the unit is fully heat soaked, but this is enough time to get things warmed up)

Here is the transformer.  Although I had siliconed the transducer deep as I could into the windings,
the inner windings are about 40F hotter than what the sensor is reading.  I knew there was going to be a difference,
so it's good to know what it is.



No issues or hot spots with the wiseguy inverter board itself.



Further scanning of the rear internal panel wiring revealed the following alarming picture.



I know it's hard to see exactly what is going on from the picture, but I am pretty sure
that 1/2 of the 120v split phase circuit breaker is bad and is running quite hot.
The heat then propagates into the surrounding wiring attached to that terminal.  I still need to pull it apart to confirm as the wiring is difficult to get to without the rear panel.
Nothing else in my setup is a concern at this point, but I am quite glad I picked one of these cameras.

Thermal cameras are good at giving you an idea of relative temperature within the field of view but not so good at accurate measurements of actual temperature.

The lens material emits some heat so its temperature affects the result. Apparently black germanium is best, though expensive. Same applies to the sensor array itself, though that can partly be compensated for. In up market ones the sensor array is refrigerated to reduce the problem.

The other issue is the thermal emissivity of the objects in the field of view.
Eg when doing a training course at the local fire station with our new $9000 thermal camera and pointing it at the kitchen boiling water unit the white insulated body of it correctly showed up warm, but the chrome plated tap that had just been used and was close to 90°C showed up cold. Chrome has a very low thermal emissivity.

Yes quite a few years ago back when I was repairing gearbox's we had a IR temp gauge and I did find out on shiny machined steel it wouldn't read the correct temperature.

There was another employee who thought his **** didn't stink and I had an onsite to go and do, so I told him put a fence so this tapered coupling that weighed around 300kg can't go past the seating position. I got told where to go as he knows what he's doing so left it there.

A couple of hours got a call on my mobile you better get back here as the tapered coupling has gone down too far. Not only didn't that guy put that safety fence in he used the IR heat gauge on shiny steel and instead of 180C the discoloration of the steel did show it was alot hotter.

Now the coupling was supposed to sit flush with the output tapered shaft and it ended up 65mm down on the taper. Now trying to get the coupling off broke a 60 ton enerpac puller and when it went the 5 ton crane got hit taking it out.

To cut a long story short the repair bill was well over 30K so when using a IR heat gauge don't use it on shiny steel as it will be much hotter than whats on the screen.

Regards Bryan

Both of these are good points, thanks for posting them.

For the temperature ranges I am using, I believe the accuracy is "good enough".  If I need an accurate temperature measurement, I really should be using the RTD transducer.
They do use IR for detecting if you have a fever, so the accuracy can't be that horrible.


Here are some comparisons between an RTD/meter and the IR camera.

I learned about thermal emissivity after getting my first IR therometer, a BK Precision 635 many years ago. Was kind of dissappointed as the most of the things I wanted to check were bright aluminum which has very low emissivity.

Did some experimenting with a metal soup can that had a bright shiny finish, a Comark C28 K type thermocouple meter with a K type probe, the BK 635, some Krylon semi flat black spray paint, using boiling water as the reference.

The semi flat spray paint seems to have emissivity quite close to the .95 default of the BK 635 IR therometer. With only part of the soup can painted you could see the (quite large) difference between the shiny and the satin black surfaces.

The black spray paint made a quick way to get resonably accurate measurements on shiny surfaces. The paint wipes off the shiny surfaces pretty easily with lacquer thinner if needed.


Gary