I'm currently working on converting my M3.8.2 car to ME7.5, and the last piece of the puzzle is to put together a basemap to get the engine running.

I've started with 018CB code from a TT, as i figure this most closely matches my turbo. However my engines in an A4, and uses a different MAF and injectors to the TT, so the map needs corrected for these, as well as the various emissions things my car doesnt have, lack of VVT, EGT etc.

The maf i'm using is the standard 2.5" A3/A4 part with the wideband 049 sensor (06A906461L), so i've extracted the MLHFM table from an A3 map (8N0906018BR) that uses this MAF, however i'm slightly stumped with the KFKHFM table. Theres some suggestion that it should "match" the MAF sensor, and other suggestion that it should "match" the intake, so i decided to try and find an A4 file which uses the same MAF. I eventually found an XDF for an A4 ECU (4B0906018CG) on here which also uses the same MAF, but has an intake arrangement much closer to that of my own car, and compared it with the table from the A3 ECU and found the two tables are WILDLY different.

The A3 table is mostly set to 1, with small corrections around there at various places and looks nice and smooth and flat in 3D view, whereas the A4 table is almost all above 1.0 by quite a few percent and looks a right mess in 3D view.

Is this right? I mean sure i expected some differences, but it just seems TOO different? Maybe the table is defined incorrectly in the A4 file on here?

Any one have any thoughts or pointers?

What engine in what car? With what turbo?

If you're starting from scratch and not using a stock intake, you're best bet is to set KFKHFM to all 1's. Use the A4/A3 MLHFM and dial in your injectors on part throttle and idle. After your fuel trims are within ~5%, start massaging KFKHFM for various load/RPM points while logging to straighten out MAF readings.

The engine is an AJL, which is essentially the same as an AEB. Turbo is a hybrid unit, utilising a 2078 compressor (RS4) and S3 turbine wheel. Its much bigger than the K03, and only a little smaller that the BAM's K04, hence starting with the BAM file.

The intake (Airbox, MAF, TIP etc) is essentially exactly the same as that used on the AWM engine, which is why i went and dug out the 4B0906018CG file, in the hope of being able to simply copy all those values over.

I'm now wondering if the big differences i'm seeing are due to FKKVS table? I dont have FKKVS defined on the A4 file unfortunately (perhaps someone could help with that?) but i'm wondering if the A4 file has FKKVS set to 1 everywhere, and thus relies only on KFKHFM, whereas the BAM file i'm using, and the A3 file that shares the MAF that i'm using, has values in both KFKHFM and FKKVS, making the KFKHFM file itself look very different to the A3 one.

OK i'm digging this back up after a long winter hiatus.

I'm having quite a bit of trouble with this mostly becuase i'm really not familiar with the tools and while i know what i WANT to do, i cant actually do it, or at least do it in a way which makes sense!

So, i've got two files.

One is the 4B0906018CG from an A4, which i have open in WinOLS Demo (i dont have an XDF for it). As mentioned above, this contains the MAF calibration which matches my car.

One is 8N0906018CB from a 225-TT, which i've got open in TunerPro, using the XDF in the definitions section. This is the file i want to "tune", and so i need to load the MAF calibration from the above file into this one.

I want to lift three maps from the OLS file (namely MLHFM, KFKHFM and FKKVS) and place them into the 018CB map in tunerpro.

So the problems, first up the XDF seems iffy. FKKVS looks more or less sane, however KFKHFM axis dont line up properly at all, and MLHFM crashes tunerpro when i try to open it..

I scraped around and fixed the axes on KFKHFM by adjusting "address step" from -4 to 0, and now KFKHFM looks sane. MLHFM issue i hunted around and fixed by turning the data set round so it was 1x512 rather than 512x1. Great, cept it was full of garbage... Referenced the OLS file of the TT map and found the map location was incorrect so fixed that and now all three maps look sane.

The next issue is the Map Axis are all different, so i can copy/paste the map between OLS and Tunerpro, however once i do that the data isnt being referenced against its proper value any more. I've hunted all round the options in tunerpro trying to find a way of changing the map Axis, but have drawn a blank.

Could someone perhaps advise on adjusting the Axes, or if theres some way round this problem?

I'm also a bit unsure about the data itself, theres quite a difference between these maps. On the A4 file, KFKHFM has values around 5-10% over most of the table, whereas the TT file has values that are all pretty close to 1. I've been going on the assumption that lifting the data from the A4 file will provide me the accurate factory calibration, but maybe i've missed something?

Cheers
Kev

ok, so a bit more digging and i think i'm getting somewhere.

Comparing the table definitions i noticed the KFKHFM map looked correct in the hex, and the table was set to "Major order:Column" in TunerPro and "2D Inverse" in OLS

However the FKKVS table (which looked wrong) was set to "Major Order: Column" in tunerpro and "twodimensional" in OLS.

I changed tunerpro XDF to Major Order: Row, and repasted in the map table, and now it looks right!

I've attached the file, perhaps someone could have a look over it and sanity-check my efforts?

Cheers
Kev

Hi,

Yes i know this, but then so is KFKHFM in effect? All three work together to translate a MAF value into an injected fuel quantity?

Like i said earlier in this post, the TT file has a KFKHFM with only minor corrections, but a FKKVS with fairly large corrections, whereas the A4 file is the opposite with a KFKHFM containing fairly large corrections and FKKVS having much smaller corrections.

I figured it would be best to keep them together. Perhaps i've missed something with this reasoning though?

Yeh i guess its nicer to have it in one place. My goal was to try to retain the factory calibration, and thus not need to have to go tweaking those tables for hours.

I guess i can try it like this and if its not right, have a rethink?

I've modified the file further with some essential bits like EGT,VVT and Rear O2 deletes. However doing that has brought forward some more questions.

I could make a new thread, but maybe its better just keeping it all in the one place?

In the Rear O2 section, the S4wiki makes reference to "fixing up the EGT model". Can anyone explain why the EGT model needs "fixed" after removing the rear O2? Usually theres a link back to here, but there doesnt seem to be for that section. The car does not have a decat. Also, do i want to disable "Cat Cleaning"? That sounds useful and i shouldnt be turning it off if i still have a cat?

I also wondered, given i've disabled the EGT sensor, do i need to make any other adjustments in the protection maps to compensate for the missing probe?

Cheers
Kev

Cheers for that!

I'm still wondering about this EGT model thing mentioned in the O2 delete info on the wiki (ATM?)

Is it only valid for people who have actually decatted their cars, or does it also apply to me, where i do have the stock cat, but do not have a rear O2 sensor?