1990 XJS is my temp gauge accurate?

Hey guys! Thanks in advance for any help you might lend. I have been curious about my temp gauge as it sits halfway between cold and normal for the most part. 190 stats installed. I ordered an IR and temp readings from the top A bank radiator hose, near the stat are 195 ish. Is the gauge under reporting or is the car running cool? 1990 XJS convertible v-12

I would always believe the IR meter.

I would go with the IR meter.
You could test your temp. sensor and see if you are getting the correct readings and replace if needed.
This could be the gauge. I would be concerned about a false sense of security with a low reading.

This a new sensor maybe it’s not reading correctly. Also the reading on the left bank in 10-12 degrees higher than the right. Is this a normal discrepancy?

Check it while it heats up. If you don’t see movement on the gauge that corresponds to actual temp changes per the IR thermo then you may have a faulty gauge, or a dummy gauge that reads the same between two, widely different extremes. These were introduced at some point to calm nervous cat owners. I do not know the date or part number.

I see slightly above N at 195F. Maybe an eighth of an inch.

Thanks for the reply. Do you take readings at the stat housing to get 195? I wondering if I have a tired or bad stat on the left bank as my IR readings are 12 or more degrees higher than the right.

I recently tested my Jag gauges (the standard dash plus 2 digital) for accuracy. I was advised by a company who specialises in calibrating devices who said simply put a seperate thermometer in boiling water (100c/112f) and see what it reads.Then use this thermometer to compare (in near boiling water) to the Jag gauges.
I found that my digital meat tester was the only thermometer I had (5 in total) that came close to accurate. I then used this to compare to the ones in the Jag. Many surprises…

That 195F reference was a recent reading when she got a little warm. I read the temperature of the temp sending unit with a Fluke IR thermo to determine what “slightly over the top of the N” equals.

Before I retired I worked as a repair tech in a Military calibration laboratory. I have repaired and tested 100’s of Fluke 65 IR thermometers and I can tell you with absolute certainty that the uncertainty of measurement is ±5°C in a laboratory setting at 100°C and that is with a reference blackbody and the 65 mounted on a tripod at 100mm from the source. Transfer that to an engine bay and the uncertainty could be as much as ±10-30°C

My point is there are a number of factors that will effect the reading of an IR pyrometer.

Emissivity of the item being read. Aluminium is about 0.50 and your Fluke will be set to 0.95 from the factory. If your unit has adjustable E then set it at 0.50 to read off the aluminium housing.

Distance to spot ratio the 65 has a DTS of 8 to 1 so at 8" from the source the spot is 1" diameter, and the IR will read EVERYTHING in the spot and average it. So if you are reading the thermostat housing you might also be reading part of the exhaust heat shield.

Lastly depending on the model the laser could be offset from the measurement beam. The Laser in the 65 is offset 18mm, so you might not be measuring the thermostat housing.

Hopefully armed with this info you can revisit the measurement and see if it changes.


Thanks for that info Warjon! I was aware of the limitations of the IR thermometers. I also have a Fieldpiece with +/- 1 degree F K-type thermocouples if I want more accuracy. But I really appreciate the E setting advice! Mine is adjustable. I’ll test to see how that affects the readings.

Wow! Great info! Thanks!

Thanks, Warren. I’m always surprised at how much faith the automotive community puts in the inexpensive IR “gauges.” The reason, of course, is little or no understanding of the heat transfer or the challenges of accurate temperature measurement.

1 Like

Correct: that’s why, to at least minimize some of that inherent field error, I would always use, where practicable, a piece of black gaffer’s tape, as a target.

The ir device i have is really small, so i just would hold it very close to what i’m trying to measure. Carefully aimed, multiple measurements, taken less than an inch from the surface. I got very consistent readings from the relatively cheap device. Those led to me changing the 22 year old (at the time) radiator and wow what a difference. Making correct observations and measurements, avoiding human error, is the most important step before moving on to conclusions.

Where would you think is the best target for the most accurate reading.?

On each side I did it directly on the shiny metal that covers the thermostat. basically right next to where the hose is clamped on.

I would use the radiator hose, it’s already black and away from exhaust.

With the IR you can use a bit of tin foil shiny side up on a backing under the hose, the tin foil will read way less than the hose and this can reduce errors.

You’d better. The temperature at the thermostat is NOT the temperature throughout the circuit. When things are operating properly, the coolant will be much cooler at the water pump inlet.

Ed_Sowell…Thanks, Warren. I’m always surprised at how much faith the automotive community puts in the inexpensive IR “gauges.” The reason, of course, is little or no understanding of the heat transfer or the challenges of accurate temperature measurement.""

My IR is a cheapie but I had the luxury of being able to test it and found it to be pretty good.

In the lab we had a Raytek MX4 we used as the reference, even this was out of tolerance IIRC 3°C at 100°C but as it was used as a reference it was never adjusted. We just corrected for the error.

While we’re discussing this: It has come to my attention that modern cars run the coolant backwards through the engine. It comes out of the water pump and goes in the top of the head and comes out of the water jacket to the thermostat housing. My 2002 Mazda works that way, and I believe the Chevy LS engine does as well. Supposedly you can get more power from the engine because the heads run cooler this way. Of course, it seems to me that if the objective was to run the heads cooler, why don’t you just run cooler thermostats? They supposedly run hot thermostats these days for fuel efficiency and emissions reasons, and if so how does running the coolant backwards not cause the same issues as running cooler thermostats?