RomRaider

Two things, I was doing a bunch of work on the wife's 06 LGT and forgot to plug the MAF back in. I was suprised at how well the car ran. So obviously there are MAP based (Speed density) maps on this damn ecu.

COBB is claiming a Speed Density flash with AccessPort "coming soon"

I would like to see it available for the open source community too. I have no experience writing code, but I can provide monetary support to the cause. I donate to RomRaider every two weeks or so and would gladly increase my donation size if that helped in any way... even if it was just "motivation"

up

I think I've found the engine load calculation based on MAP. For an 07 sti:

engine load (g/rev) = (Manifold Pressure(psia) * .0954) - .17

I don't think there's anything more extensive than this for calculating load during MAF failure, although I'm not 100% positive.

Hmm, doesn't take into account the charge temperature. I would think for normal / spirited driving we would want to fully apply the ideal gas law. I would relocate my MAF for something like this.

J

Not saying we should use this. Merely a response to what happens during MAF failure and the fact that it is not very useful for what we want to do.

This feature also apply for 16 bits ECU. It's a poor estimate which has been implemented to prevent engine stall (eventually to be able to drive the car to the closest service point), but not good for daily driving. Throttle body air temp and VE are required for a good calculation (as well as differential throttle input).

well, VE should be easy to calculate by datalogging a naturally aspirated pull and using the maf sensor and IAT and RPMs to derive VE.

for WOT/near WOT MAP-based fuelling,
this would require:

new logic, the actual SD equations

(I'm playing around with these a little bit, we are in luck with our fuel temperature sensor because we can get a more accurate specific gravity number!)

one "VE" map, 2d map with 12 values, VE vs. RPM.

one tps crossover map, 1d variable.

that would be a lot of work, for sure, but thats near what it would take for a good SD system.

we might also consider cannabalizing the EGT sensor input for a post-turbo IAT input, basing our densities on the relatively cool IAT on the MAF housing would make the cars run ludicrously rich.

So the current IAT sensor is pre-turbo?
I always assumed it would be post... that complicates things.

its integrated into the MAF housing, subaru basically figures out the delta T of the MAF vs. manifold temp of the stock system in a lab, then applies that to all of its calculations in the ecu as a fixed offset.

obviously with non-stock parameters or hard parts their fixed offset becomes completely inaccurate.

you'd be surprised at how well it DOES work.

basically you're banking on the fact that the ic core sits at the same temperature as the intake air flowing past the iat sensor.

to the degree that this is true, it works perfectly. as a result, it works VERY well with fmics and CAIs, and pretty well with tmics and SRIs. the closer you keep the temps of the iat and the ic core, the better.

not saying it's the best way, but phil grabow tunes hydras (exclusively SD load based) with the IAT in the stock position every day of the week, and they seem to run ok. my own experiences with the utec in SD mode back that up too.

that said, i AM looking at ways of moving the IAT to become a MAT.

Well, it works like crap if you have a shortram intake and a FMIC.

I'm sure it works well enough, but putting the air temp sensor in the manifold is better.

Ideally you would not use to the temp sensor at the MAF to tweak your timing. Let the MAF temp sensor only be used for MAF readings, nothing else.

Ideally you would have a temp sensor as close to the cylinders as possible and use that feedback to adjust timing and boost as well.

This would let you tune for actual (or at least a lot closer to actual) engine conditions so you would not need to worry about heat-soak or other conditions which might raise/lower in-cylinder air temps when the MAF air temp stays the same.

i'd imagine it would!



chris vincent and i have done SOME work on this... have you seen this thread?

http://forums.nasioc.com/forums/showthr ... ?t=1088149

i'd really like to relocate my IAT too, but i'd rather not destroy my MAFs just to get a little, hard-to-mount thermistor out of the deal.

ken

I agree about keeping the MAF IAT sensor for MAF temp corrections.

really, the best place for an SD IAT sensor is right where the pressure reading is being taken.

ditch your EGT thermistor, recal the curve, and presto! spare temperature input.

it would require some rom work though, but so would SD.