temp gauge spikes to the top
 

I have a problem with my water temp gauge in my 1993 2.6 190E. It is usally pegged at the top of the gauge while the water temp is typically right where it should be. The gauge or the sensor is malfunctioning. I compared the resistance readings from the suspect gauge with another temp gauge from a different cluster I have and the readings are the same. I replace the temp sender and it did not help. Any ideas what to do next?

There are 3 temp sensors.........did you change the single wire one?

If you substitute a 38 ohm resistor for the sender , the gauge should read 100C..if not, the gauge is bad.

I'll grab a resistor and test it out. Any idea what the color bands are for a 38 ohm resister? I can figure it out after a while if I need to.

Just get a 100 ohm wire-wound adjustable and set it with your ohm meter.

Ah,
good idea, for a minute I was confused, but I get it. I variable 100 ohm pot and dial it in to 38 by using my meter. Got it. Great idea, I had not thought of it, but I wish I had. Common sense was out to lunch.
I'll pick one up at Radio shack today.

That way you can test the guage throughout it's entire Range.
I have a 38 ohm made up for my toolbox with a ground clip for fast testing.

Here are the specs for the sensor with the gauge reading at each value

OHM 110 = 60C
67 = 80C
38 = 100C

i have found problems with the ground on instrument panel causing such a problem and fixed it by making a ground wire one end under instrument panel screw and other at main ground behind instr panel.

p.s. very unusual to find bad gauge.

David,
Thanks, I was thinking I have not found many bad gauges either and I convert a lot of gauges to white faces so I see them alot. I will check in to the ground issue and get a test setup to check the gague, thanks for the advice.

I have the same problem with my 1994 E420. It spikes to it max range and gets stuck there. Are the resistor values the same for my car as well? Thanks.

Both chassis use the same Termistor/sensor..

The temp gauge [A1P1] has the same feed as the gas gauge, so it can not be feed or the gas would also not work. And the gauge recieves it's ground through the sensor [ B/13], so it is not a gauge ground problem. [The gauge uses a Switched Ground circuit via sensor]
If the resistor sub does not work, you have a bad gauge or open/shorted wiring from gauge to the sensor. If the sensor wire went to full ground , it should peg the gauge. B/13 Thermistor sensor is Negative Coefficient Temp sensor.[ NCT] , meaning as the temp rises , the resistance value lowers .
So, the lower the R value , the higher the gauge reads.

Arthur,
So it I pull the single wire off the sensor and use a test lead with the 38 ohm resistor in line from the senders wire to a ground, a good gauge should go to 100 deg celcius, correct? The sensor is providing a ground that varies according to the the temp right? If I substitute a resistance for the sender and go from the sensor wire to the resistance to a ground on the engine I can check the gauge at different resistances? If the resistances are not correct the gauge must be bad.

<<So it I pull the single wire off the sensor and use a test lead with the 38 ohm resistor in line from the senders wire to a ground, a good gauge should go to 100 deg celcius, correct? >>


Yes ..you are simply replacing the sensor with an R value of what a good sensor would have a value of if it was sitting in a coolant that had a temp of 100C..that's all..it is a simple susbstitution ...a Resistor of a known value in place of the sensor..simple.
You can use any value on that chart I posted, but the easiest one is 38ohm/100C
Then the gauge [ which you are fooling] thinks the engines coolant is at a temp of 100C
This test does not neccessarily mean the gauge is NG, it simply takes the sensor out of the diagnostic equation..and it can also be used to test a gauges accuracy.