GM temp sensors

[Richard Wakeling]

Hi Frank and group,

> Does anyone have the resistance values of the GM air/coolant sensor
> at temps over 212 deg F. Have values at lower temps from GM Fuel Inj
> Diag manual and other sources but those start at -40C/F and go to 212F
> and go no farther. Am trying to develop a lookup table for my datalogger.
> Any help greatly appreciated.

The resistance values versus the temperature on both sensor are not
linear to my knowledge. The table does not represent a straight line
with regards to voltage. As for lookup tables, you could of course do
the obvious and actually heat up the sensors in boiling water or
something (of course only get it to about 95 degrees celsius), and
measure the voltage. However, if you are using datalogging devices such
as Diacom, you can always manually change the voltage, emulating a
sensor, with a variable resistor, then read back the temperature on
Diacom. 
 The GM data stream sometimes puts out the sensor as a reading of
temperature (i.e. a single byte is dedicated to the temperature, and
another to the voltage). I know that this however does not appear on the
Sunbird data stream and I had to create my own lookup table. Once I
obtained the voltage versus temperature, I did plot the result in excel
but had to use a best line of fit, with a polynomial of degree 6! This
of course was the best resolution possible. The result however still was
a bit out, with some readings giving about 1.5 degrees celsius out
(Australia here), however the overall map was plotted correctly. I
believe using a lookup table would have to be more accurate however this
uses up a reasonable amount of CPU time, depending on your aim.

I don't really have the above problem in OZ though, coz the datastream
in Oz puts out both the bytes dedicated to voltage and temperature. Hope
this helps you out.

Andrew Wakeling
kojab at ar.com.au