Can someone explain this?

**justchat_1** · 03-11-2010, 03:59 PM

Ok test vehicle is my 2005 mustang gt....

Outside temp is around 50 degrees.... so garage temp is maybe 60.

Query is 010F (Intake air temp)

response is 41 0F 41

obdII spec says that the temp is the value - 40 in Celsius. Which would make this 77degrees fahrenheit?

Considering the car has been sitting overnight is this normal to be off by that much? Am I screwing up a conversion somewhere? Or is this value special because its an american car? (I know all the manufacturer specific queries return fahrenheit temps.)

**p2psmurf** · 03-11-2010, 04:42 PM

Reply is 41. Or 65 in decimal. Minus 40 is 25 degrees Celsius or 77 degrees Fahrenheit.
You are converting it correctly. What's the coolant temperature at that moment? They should be (almost) equal. If you disconnect the air temp. sensor what reading do you get then? I have seen examples where a replacement (calculated) value was given in stead of the actual sensor temperature.
Possible sensor fault or software bug in calculating OBD2 value.

**jfitzpat** · 03-12-2010, 12:28 AM

FWIW, the IAT sensor is often in the intake manifold, very near the cyl head. So if the engine is running, 20 degrees F, even right away, is not all that unusual.

Depending on the setup, a 15 degF offset isn't that unusual with the engine off either. It depends on the type of sensor used. Certain thermocouple setups cannot be read below ambient temp plus a small DC offset.

-jjf

**SNOtwistR** · 03-12-2010, 01:25 AM

that sounds about right for the choke sequence of ecu, cold start the ecu tells fuel system its -40 to add more fuel since fuel injection has no choke butterfly valve just my 2 cents hope it helps SNO

**justchat_1** · 03-12-2010, 06:36 PM

well the sensor is in a cold air intake pretty much even with the fender.....IAT was about 8 degrees above coolant temp so i can't really explain it-but with the engine running temp is dead on. As long as I wasn't screwing up a conversion somewhere thats all that matters.

thanks for the feedback guys...

**jfitzpat** · 03-13-2010, 07:48 PM

Originally Posted by justchat_1

but with the engine running temp is dead on.

This is what I was referencing above. Think how a simple thermocouple works. Basically two different types of metal. When there is a temperature *differential*, there is a tiny voltage (from something I believe is called the 'Seebeck effect').

So, when a thermocouple is at the same temperature as the reference junction, no voltage is generated. So 0V does not equal temperature X, it equals 'ambient' or 'reference' temp, which can change.

This leads to two common types of problems at ambient air temp. First, we are talking a very tiny voltage, so 0V is down in the ground noise range of the system - so you can get some false offsets until there is a small temp increase.

Another common problem is that the 'reference' is wrong. Remember, it is a _relative_ temp measurement, so the overall system has to have something to reference it against. There are a bunch of schemes, impedance of another sensor, precision cold reference joints, etc., but many of them have the problem that they are in error until typical engine compartment temps are reached.

-jjf