I have been seeing temperatures about 20* higher than expected from my GM IAT. The sensor is wired directly, grounded at ecu ground by the TB and the other going to megasquirt. I am using GM settings and 2490 bias.

Tonight I ohm tested the sensor in boiling water and found that it read around 240 ohm. In Ice water, it reads 7640. According to the chart below, the sensor should be reading closer to 180 and 10500 respectively. The ohmmeter was tested against a 270 ohm resistor to assure that it was functioning correctly. It read 266 for all 3 resistors tested, which is within the 5% tolerance spec.

Aside from being almost exactly 20* off, the sensor seems to act normally, so can I simply use these numbers ive gathered, and adjust adjust the temperature I input to easytherm by about 20* for the middle one?

example:
212 -> 240ohm

32 -> 7640ohm

... wait as i am doing this, it is seeming more complicated. the curve has been thrown off and it is not 20* off everywhere, its about 8* off at 32 and 20* off at boiling... help me get that middle value right someone, my math sucks

 

A new sensor is cheap. Is it open or closed element? Closed element ones are everywhere.

 

open element

 

Just leave it....it doesnt have to be exact, just consistent.

 

it matters to me, I want accurate temps

 

Ok well according to some calculations I made, the curves would probably cross at around 37*C

Basically I discovered that the stock resistance at 0*C is 11550, and that my sensor is reading 66% below this. Then I figured that at boiling, my sensor is reading 77% higher OHMs.

66/0C x 177/100C = 37.3*C ... i know my math is probably screwed up, but the number seems to be about where it should be. So my easytherm settings will be as follows:

0C - 7640
38C - 1800
100C - 240
bias - 2940

This should get me a lot closer than it is reading now at least.

 

and it is working, looks to be reading within a degree or two from what I can tell.

 

Inspect your Easytherm output files after you are done. Easytherm really screws up if your vales are off by a degree, and it doesn't even tell you that something is wrong.

Jim

 

I dont understand what you mean by this, I thought easytherm simply took the temperature vs resistance curve and scaled it out?

 

Easytherm CREATES the T vs R curve from the values you give it. If the values you give it are off by a degree or two, the output will be TOTALLY wrong, like random garbage, not just scaled out with less accuracy.

Jim

 

ok ill take a look at it.