In a world of perfectly-calibrated sensors, the temp:% ratio should be constant, right? The sensor should only affect the ADC:temp ratio. Of course, if you just manually derived it instead of, say, using easyterm, then I guess there wouldn't be any correlation. :/ But I'd still like to see it just from sheer curiosity.

-Mike

I have the same problem with my EMS and a GM open sensor. On a hot day after I shut the car off for 10min or so I'll be lean when I start it up.

No, each sensor has different characteristics. Take an open element and a closed element IAT sensor. You'll find that the open element sensor usually does mean ie 50C when it says 50C. Closed element sensors tend to heatsoak more, leading to situations where the IAT maybe 50C but the sensor reads 70C.

Jim

I did some more thinking / investigating on this issue this weekend, and realized that REQ_FUEL has nothing to do with it. Only two things can affect this issue (well, not including baro correction, intake/CLT correction, etc). That's the air density as Reverant recommended, and the injector open time.

PulseWidth = OpenTime + Accel + ReqFuel * VE * MAP * AirDensity * BaroCorrection * EGOCorrection * WarmupCorrection

If you ignore Baro, EGO, Warmup, and Accel, then you end up with:

PulseWidth = OpenTime + ReqFuel * VE * MAP * AirDensity

What I was thinking originally is that if I changed my ReqFuel, it would change the way AirDensity affects the final pulse width. But looking at that equation... If I hold AirDensity and MAP constant, because they're factors external to the normal tuning settings, if I change ReqFuel then I just make up for it in VE, yielding the same value before multiplying by AirDensity. So AirDensity will affect the outcome in exactly the same way.

However, if you increase OpenTime, leave ReqFuel alone, and hold MAP and AirDensity constant, then you'd have to reduce VE to get the same final pulse width. This means the entire portion that AirDensity is multiplied by would become a smaller portion of the final pulse width, making the final pulse width less sensitive to AirDensity changes.

All that said, I also realized that because the equation multipies by both MAP and VE, my pulse width at idle is only 1.0 + 7 * .38 * .30 = 1.798ms, which is exactly what I see at 80F idle. A 10% air density chagne (going from 70F to 130F) only makes a .0798ms difference in pulse width, which is below the resolution of the non-high-res code. This means idle is definitely not the condition for me to be looking at to worry about how my afr changes with IAT changes.

-Mike