I own a 2000 Escalade. When I bought this vehicle used the temperature gauge did not work very well, the needle would fall well below the dial's 100 degree mark, and only come up to about 150 or 160 when driving. I might add that this is the only gauge in the cluster that is not working right. I've had the water temperature checked for fear that it was getting hot and I could not tell, but the temperature was fine.
I have checked the thermostat housing ground on suggestion from a mechanic and I've changed the antifreeze. After changing the antifreeze I noticed that the gauge needle stopped at 100 and went up to about 190 or 200, but this only lasted for about one day and then it went back to below the 100 degree mark.
Could be the actual gauge inside the instrument cluster causing the problem, but in this situation the circuits have to be tested correctly to find out for sure.
There's a special tool designed to do just that. General Motors' service departments often use the Kent-Moore J33431 Signal Generator/Instrument Panel Tester. This tool can replicate signals such as vehicle speed, fuel level and coolant temperature to aid in instrument panel diagnosis.
First step on your Escalade's coolant temperature inaccuracy complaint would be disconnecting the coolant temperature sender connector, and hooking up special tool J33431 in its place. Then we can replace the signal from the sender with correct hot and cold signals from the tester. In this case we're dealing with resistance (ohms), and the specifications are 1305 ohms "cold" (100 degrees) and 49.5 ohms "hot" (260 degrees).
Once connected correctly, ignition on, we adjust the tester resistance to 1305 ohms. The temperature gauge needle should be pointing right at 100 degrees. And the same goes for the other end of the scale, tester at 49.5 ohms and we should be reading 260 degrees on the gauge.
If both readings are correct, that confirms the coolant temperature sender being defective; it's sending out an inaccurate signal and needs to be replaced. Either that or the thermostat is stuck open and not raising coolant temperature. If the needle is pointing at the wrong temperature with the correct resistance applied, the problem lies upstream.
This is when we work our way toward the gauge with the tester. Best bet would be to remove the instrument cluster and apply the correct resistance directly to the coolant temperature gauge. If the needle is still pointing to the wrong temperature, the gauge is bad and will require replacement of the entire cluster. If the needle is now dead-on accurate, there's a poor connection (high resistance) in the wiring between the sender and the gauge. The weak link can be in a number of areas, but the cluster connection itself is a common spot to cause problems.