1999 V70 Temp Needle overheat, fans on, car doesn't ACT hot.

[charleskennedy23]

I had a similar thing going on with my 98 V70. Temp gauge going hot and the fan running even though the engine and cooling system appeared to operating as it should. No engine codes to tell me what was broken. I Changed out the temp sensor and problem solved. Got an OE sensor from ipd for $55 plus shipping. You may want to try this before you start messing with the water pump. It's a real easy repair and if yours is original it may be time to do it anyway.

[voom8]

Okay, thank you very much for the suggestion.

I really had to wait until the weather permitted me to do the work.... I am in a parking lot... and it is MA... and it has been absolutely terrible these last two weeks... with rain, wind and crappy temperatures.

What I found, is that the new sensor Did fix the problem!

To me this was a surprise,
because the old sensor measured (between the two wires and disconnected from the computer) ohm readings that were typical to the chart that shows that the car was at operating temperature and not over heating. I was wondering how that could be... So I disconnected the upper radiator hose without the thermostat being in... and watched the coolant flow towards the radiator, it flowed, but not very much (not as much as I would suspect). So I was confused if this was typical or not (rate of flow)

But what I did find, is that the old sensor measures perfectly in a pot of boiling water. I measured a few points on the way down (cooling) and everything lines up with the chart.

The difference between the two sensors:

The old sensor measures ~ 40K ohms to the sensor body (which is common ground to the electrical system of the car) on either wire. When the temperature in the boiling pot of water got hotter... the resistance reading from either wire to the sensor body dropped to ~ 5K.

The new sensor measures totally open to the sensor body. (at least when coldish)

So I think the computer was being fooled either by a leaky path of current out of the two wire measurement circuit back to itself through the common ground (the computer) from the sensor body... or the computer was seeing voltage spikes the exists in the common electrical ground of the car.

It may have been said on other posts by other people, but a sensor could be bad even if it reads as expected with an ohm meter... the charts indicating ohm/temperature should also say that either wire has to be an open impedance to sensor body.

[abscate]

What's the impedance of you VM...can we report the sensor to ground impedance as >2 Mohm??