Pb when rescaling MAF on a STI8

[Kci-sti]

Hello,

I'm currently remapping my EUDM STI 8 (AH992 ECU type) which has been slightly modified: Garrett GT30 and some other mods. The car should be pushing approx 400 bhp.

Of course I'm maxing out the std maf reading which stops at 4.69V for 292g/s. But I'm not maxing out the MAF capability as the max voltage I get is approx 4.94V.
So,based on the 2.5L STIs (they got a 350g/s MAF scale) I've rescaled my MAF.
Everything would be perfect if only it was possible to save values above 300g/s in the map!!
For example, if I say I'll have 342g/s for 4.92, when I save the file, the value saved will be automaticaly modified to 42g!! It removes 300g/s for all values set above 300g/s!!

Could you explain me how I could get rid of this automatic correction which is really getting on my nerves...

Thx a lot for help...

[Jon [in CT]]

The 300g/s limit is a result of the number of bits allocated to hold the values in ROM.

[Kci-sti]

The 300g/s limit is a result of the number of bits allocated to hold the values in ROM.

Yeah,it's sad...I've seen that after having post. 300g/s is multiplied and it gives 65535...which gives FFFF in hex. So the max number in hex with four digits.
So I'm fucked up... I gonna have to fit a bigger MAF,rescale it and change the conversion in the XML file so that it will display the proper value in ECU Edit.
Such a waste of time knowing the MAF is not maxed out at all at 4.69V!! I'd like to know how different is the code of the 2.5L STIs to allow to enter data over 300g/s... Anyway,I don't think an upgrade of 2L STIs would be possible as it would equals in far too many changes without even being sure it will work.

Time to serial mount my Perrin MAF and the std one to do a proper rescale!

[Freon]

Don't mix up your limits. There is a physical limit. The fix there is to increase the MAF housing pipe diameter, i.e. BigMAF. This has really nothing to do with the limitation of the 16bit format.

The 16bit format limitation is a real bitch. All you can really do is hack all the references to MAF g/s scaling in the XML. Or you can just remember that MAF g/s in your logger, and the load columns in your maps, are actually much higher, then just use injector scaling to roughly compensate for a larger MAF housing pipe. Using injector scale to tune a large MAF housing is opposed to doing the "correct" fix, which would be adjusting the MAF scaling table up to 350-400g/s or whatever. Your logger is never going to read right, so you'll just have to know.

There is no elegant fix, for sure.

The DBW, 2.5L turbo cars use a completely different ECU and processor. It is a 32-bit CPU w/FPU and uses 32-bit floating point for the MAF table and all the airflow calculations. 32-bit float allows an enormous range of values, including negative. The ECU overall is much more powerful.