I've had a couple temperature gauges in OM617.952 W123 diesels read 10 or 15 degrees high. Cleaning the grounding points didn't seem to help -- the gauges were just consistently off.

I had to repair the instrument circuit board on my '82 wagon last month and temporarily swapped clusters from my parts car. The temp gauge from the parts car indicated 95-100C while the original gage read 85-90.

The parts car was my daily driver before it got rear-ended a few years back, and it almost always registered an engine temperature of 100-105. As a test I once removed the EGR thermo switch from the front of the cylinder head after warming the car up (the pressure cap was loosened but even then the coolant surged out a little -- be careful) and put a candy thermometer into the water jacket. The thermometer read between 170 and 180 degrees, as I recall -- not the 212f (100c) indicated! So that's a poor man's infrared gauge for those who don't mind spilling a little coolant. JUST MAKE SURE THE COOLING SYSTEM ISN'T PRESSURIZED (pardon the caps). And wear rubber-coated work gloves to prevent parboiled fingers. I'm probaly stating the obvious, but playing around with hot coolant can be very dangerous

One more cooling system anecdote: I remember reading a theory on Rich Sexton's old mailing list by a guy who was convinced that the thermostat housings on 617 diesels could erode slightly at the thermostat bypass valve seat. The erosion, he theorized, allowed some coolant to bypass the radiator at all operating temperatures and thereby raised the engine temp to unhealthy levels.

His fix was to drill two or three holes into the outer ring of the thermostat. He claimed that it forced the thermostat to open wider and seal off the bypass circuit. I forget why he said it worked that way -- probably increased the flow of heated coolant through the stat -- but it's an interesting theory.

So if it is ruled out, are there any theoretical explanations based on the fact that it is acting this kind of this particular behavior? We all know for certain that, even though they are all the same makes, year, model, and even serial numbers and production date, each individual car/engine has its own personality, right? Hence the reason a said that my car, specifically, is acting wierd ever since I acquired it. So what made you conclude that everyone with a 617 would experience the same behavior?

You're making an argument that the temperature fluctuation is due to instability of the system as the engine reaches operating temperature. If this was inherent in the design of the 617, all engines would exhibit the same behavior. Variation in the tolerances of the build won't provide a suitiable margin for acceptance of this behavior as normal.

I previously recommended that you check the temperature of the fluid with an IR gun. You have failed to make this attempt and further discussions are moot until you accomplish that task and obtain some data.

Found problem
 

A replacement fan clutch dropped the temp down to normal range. When I shut off engine, old fan clutch spun a bit to long before stopping. Put one in from my donor car and that seems to have fixed the problem. Thanks for all the input & help.

If it ran cool without a thermostat, but hot with a thermostat, why would a new fan make a difference? Thermostat sets the minimum temp, right? Without tstat no minimum temp, so runs cool ... add a tstat and runs hot ... but with new fan runs at tight temp?

After having a spate of bad thermostats in the early 1980s on a 2002 BMW, I learned from several of the BMW club that they consistently drilled small holes around the perimeter of the tstats disk (internally, of course :-)). I think the logic was that you allow some flow of warming water which helps the tstat open ... have done this consistently since then on most tstats - I suppose it will, as a side effect, slow the warm up cycle, initially?

This fundamental question baffles me as well, however, I'm quite convinced that components such as the radiator and the clutch fan do affect the operating temperature.

The SD runs at 95°-100°C. now..........no matter if the ambient temp is 85F. or 60F. The brand new thermostat runs exactly the same as the 20 year old thermostat. The chances of the two of them being NG are zero.

I have the suspicion that the difference in location of the sending unit for the gauge and the thermostat is the issue here. If you notice the gauge on downhills, it drops to 81°C. or so. Why? The thermostat would close all the way down and the temp would be maintained close to 90°C. But, this does not happen on the gauge. The only possibility is a difference in coolant temperature between the gauge and the stat.

With a marginal radiator and/or a marginal clutch fan, the temperature swing on the gauge seems to be far greater than one would observe with brand new hardware. Some members have stated that the gauge sits at a rock steady 85°C. despite the load on the engine. These folks have installed a brand new radiator.

I don't fully have the answer to this conundrum, however, I'm positive that the thermostat is not the culprit.

I'll be searching for a solution for the SD as the higher temp. is starting to annoy me.

Tstat should be closed until some temp slope prior to tstat's rating (is it 80?)and then start opening, then, as temp drops it should start closing to allow some coolant to circ to maintain rated temp until load on engine relative to outside air temp begins to increase, causing it to open.

So, if temp goes down to tstat rating and holds going down long hill, then tstat is working near closed position (allowing some coolant flow likely), but if immediately goes above rating when headed back up by a significant margin and stays, then something isn't working correctly, right? Should increase some, then start regulating, IF all other things are equal - like radiator capacity, no air in coolant, no blown head gasket, pressurized system, viscous clutch fan, etc - 100C electric fan probably not in play, nor would the electric fan linked into air conditioning pressure switch, right?

Agreed.

However, we're talking about small discrepancies between the gauge and what the thermostat would likely be seeing with regard to temp.

In the case of the SD, the gauge will, occasionally head up to about 99°C and then back off to 93°C. or so. But, I do note that it will also, on occasion climb to 90°C. and simply hold there. The ambient temperature is not a variable.

I still wonder if a slight bit of air remains trapped in the head???

Perhaps - think I would "burp" it and see ... if you do take the tstat out again, try the hole drilling exercise - I think what it is doinig in a very rough way is smoothing out the tstat open/close movements ...

I need a positive method of removing any air. My methods are questionable.

Roy has a positive method by pulling a vacuum through a specially drilled and tapped expansion tank cap. I'd rather not go to that extreme.

The hole drilling exercise is capitulation. I'm not ready for that.

I recently replaced the original 21 yr old tstat with an after market german part. The new part lowered the temp about 10 deg for me. Forgot the brand but it was stamped 80 deg. 100 deg used to be normal. Now it stays between 85-90. I also replaced the coolant from green to yellow, so i dont know if it was the coolant or the tstat.

Coolant.
 

Nobody's mentioned coolant. Is everybody running 50:50? For those folks in warmer climes, try thinning it out a bit.

Jay.