Air Temp Sensor

[sammy]

Hi Everyone

I've been looking everywhere for the GM Air Temp Sensor that’s referred to in the MS Manual. I figured this would save a lot of headaches given that the resistance values are commonly known.

From what I can gather, the sensors are not available down here in Australia, so I got hold of a sensor from another manufacturer. I can’t remember what car it comes from, but the sensor is an open type.

To determine the resistance values I used a multimeter with the following tests
o Cold water (result was ~10 C)
o Room Temp (result was ~21C)
o Heated cup of water (result was ~30C)

I wired the sensor into the relay board, hooked up the MS calibrated the Therm values and started the engine.
The ATS was not installed in the air box; it was just sitting loose on the fender next to my Thermometer.
I was disappointed to see that MS was reporting ~ 6C less than the Thermometer.

I assumed I didn’t gather the values correctly so I started over again , but each time it was different (inconstant results)

I’m guessing the problem is with me .... and someone has probably done all this before (only properly!), so I’m wondering if there is a list of Sensors and resistance values (apart from the GM one) that people have used over the years?

Or even is people can help by providing their experiance iI would be appreciative

Thanks

Sam

[Bernard Fife]

Sam,

You need to use temperatures that are a lot further apart than that. Use ice water (0ºC), room temp, and boiling water (100ºC) as recommended in the manual.

For ice water, start with all ice cubes, and let it melt to a 50/50 mix of ice and water and the water will be at 0ºC. Boiling water will be at 100ºC. These will eliminate some of the potential inaccuracies in the temperature measurements. They are also far enough apart to give a true sense of the response curve.

Lance.

[sammy]

hi Lance thanks for your help with this, i didn’t notice this mentioned in the manual to this extent, I'll go back over it again and see if I missed anything else.
Meanwhile I’ll try what you said and see how I go

Sam

[Spittybug]

I found that when I did the calibration that Lance suggested, I still had some discrepancies at high temps. I plotted curves for my result, documented ones from the manual and a few other sources. They all seem to be exponential, not linear. Essentially that means at the very high temperature, a small change in temp results in a fairly meaningful change in the resistance. So, since these things are calibrated by typing in known resistances for determining unknown temperatures, you can be off by several degrees pretty easily at the top of the range. Once mine was installed I noticed that TS was giving me a temp reading about 10deg higher than my car's gauge and what I was able to get using an infrared gun. I basically just fudged the high temp resistance reading in the calibration table until it was close.

Remember, the absolute temperature over most of the range doesn't matter, it is the rate of change that you are looking to understand. At the high end though, the absolute is important so that you don't get set points for automated control of cooling fans or or overheat warning devices.