IAT varies too much to be used in SD tune?

[forcefed86]

Working with a turbo LS engine. (done several now with the MS3 now). I usually end up flat lining the IAT/MAT table which causes some issues Id like to avoid. Main issue is at start up the sensor is around ambient, and as the engine warms up the IAT rises. If I’m sitting in traffic on an 80* day my IAT temps can jump up to 125+ Then when I accelerate they drop rapidly closer to ambient again. THis causes the tune to bounce all over.

Temp sensor is mounted to the OEM plastic intake manifold, so it isn’t getting “heat soaked”. Are the generic supplied MS3 settings for the MAT table geared towards reading ambient temps only? Seems like I’d have to place the IAT sensor in an isolated “cold air” intake tube to work properly.

That said since this is a turbo build I use the IAT charge temps to pull timing so installing the sensor externally isn’t really an option either. In this case would it be better to place another IAT outside the charge pipe for cruising SD calculations. And use a second IAT sensor in the charge pipe to pull timing?

The EGO and auto tune seems to take up the slack if I just zero the MAT table, but that seems like a half assed way of doing it.

What’s the solution here?

[krisr]

I find that you need to set the MAT Air Density Table to 100% across the board until you ascertain the air temp operating range of your engine when you do your tune. Say for example you tune the engine between 80-90*F, then i'd set those cells to 100% for a start. Then you can adjust either side for hotter or colder days to maintain AFR's back to target without touching VE. The stock MAT curve is uses "ideal gas law" from memory, which IMO doesn't work unless you tune at around 20*C, therefore I find that MAT correction MUST to be tuned once your VE table is dialled in.

[Reverant]

After fiddling with stock ECUs, the answer is that you can't just take the IAT reading as-is and use it in the calculations; you have to use a model to calculate the actual air temp derived from the IAT reading, based on a number of other variables (coolant; speed, RPM x time, airflow on the snesor, time since last shutdown to name a few).

[whittlebeast]

See under <Basic/Load Settings> <MAT/CLT Correction>

And also <Basic/Load Settings> <MAT Air Density Table>

Between those two corrections, you should be able to dial in almost anything you find as a repeatable response.

Andy

[vw_chuck]

IAT varies significantly based on massflow through the engine. Your assumption that it should be a constant temp no matter what is flawed.

[whittlebeast]

<Basic/Load Settings> <MAT/CLT Correction> Is a mass air flow correction. MAPxRPM is essentually Mass Air Flow

[elutionsdesign]

MAPxRPM is essentually Mass Air Flow

It is N/A, once someone uses a weenie turbine and the TIP shoots to 3-4X MAP it's a poor indicator of actual flow.

Years in the chemical process control industry has taught me not to mess with the ideal gas law, instead figure out where your sensor needs to be for the best information. You're just lying to yourself flattening out the physics and it will bite back at a really bad time. The OEM sensor locations were for OEM process control and they spent a lot of $ on development, once you change the controller you start over.

[forcefed86]

Should be 1:1 on the drive pressures on this twin setup FWIW.

[whittlebeast]

MAPxRPM is essentually Mass Air Flow

It is N/A, once someone uses a weenie turbine and the TIP shoots to 3-4X MAP it's a poor indicator of actual flow.

Years in the chemical process control industry has taught me not to mess with the ideal gas law, instead figure out where your sensor needs to be for the best information. You're just lying to yourself flattening out the physics and it will bite back at a really bad time. The OEM sensor locations were for OEM process control and they spent a lot of $ on development, once you change the controller you start over.

Please post up a big data log of a well tuned turbo motor that does not fit this assumption. I have seen way more that a few that are real close.

Andy