NefMoto

Hi,

Great forum this with lots of info, and I've read a lot of topics which helped me get where I currently am with my setup, but now I need some assistance please with getting my setup working properly. I have a 2.7T ARE engine from an Allroad, 4bar FPR, RS4 MAF, RS4 injectors, RS4 intake plumbing, RS4 FP and RS4 K04 turbos with RS6 compressor wheels, 76mm DPs and 76mm FF exhaust, no cats. I managed to get a T-Box ECU from the US and an S4 cluster meant for the Canadian market, so no IMMO. Long story short, I managed to get the motor running with a stock M-box flash obtained here, and through this forum managed to code out the rear O2 sensors, LDP etc., and sorted the maps for the RS4 MAF and injectors.

I toyed around for a while with copying maps from the K-Box over to the M-Box flash but was never able to get rid of the boost overshooting when going WOT from low rpm. I then tried flashing a proper K-Box but got blocked by the IMMO - for some reason I was never able to get bootmode going to sort that out and only after giving the ECU to someone with an eeprom programmer who de-soldered the 95040 and wrote an F-Box 95040 file with IMMO-off onto a 2nd ECU (from an A6 2.7T - I didn't want to mess around with the T-Box / M-Box flash) could I get the K-box flash running. But the boost is still overshooting at WOT and low rpm, and I'm having CEL because of rear O2 and also SAI DTC and RS4 cam timing DTC. So this is where I'm currently at - the K-Box XDF I found here doesn't seem to have all the parameters to code out the rear O2s and SAI - unless I'm missing something - can someone please help with this?

Also the boost overshooting is spoiling a lot of potential fun I was expecting from all my effort. I must say I have read a number of threads about PID re-calibration a number of times but I just can't quite get my head around the KFLDIMX and KFLDRL maps, and how to set CDMDAPP = 8 - I'm assuming this is a parameter one should find in the XDF but I do not see it in the K-Box XDF I have...? Firstly, do I really need to re-calibrate the boost PIDs?

I have logged a couple of parameters using Zetsi's logger - here is the csv (I hope I'm logging the correct variables!) as well as an Excel file with some graphs (sorry about this - I'm having mixed success with ECUxPlot!):

Thanks...

Browse to 10B70 in the RS4 file. The next 13 bytes should be

00 C0 EF FC 0C 00 F3 FF AA FA 55 55 55

Replace them with

00 FF EF FF 3C C0 FF FF AA FA 55 55 55

It'll eliminate the errors associated with the missing hardware specific to the RS4.

As for the rest, you have a lot more reading to do. It sounds as though you have no grasp on PID control.

DD can you elaborate?

Thanks, as usual.

10B70 is the start of Endstufen Konfig (Power stage Configuration), or ESKONF. The next 13 bytes are basically a map of the installed components. (although I'm sure you knew this).

Copying the M-box ESKONF into the RS4 file is sufficient to prevent the NWS/N331/SAI circuit codes, but he also wanted the rear o2's deleted as well as the LDP. I just went ahead and combined them. The end result is an ESKONF that's configured for only the components he has installed.

For those of you that are unfamiliar, read this.

http://s4wiki.com/wiki/Tuning#ESKONF

RE: PIDs - Recalibrating is much easier that taking a blind stab trial and error.   You are overboosting because you are trying to use a calibration for a turbo that you are no longer using.

PID takes a while to understand how everything interacts, but you will be very happy when you get it.   And you get it by continuing to follow the PID from square one thread.  

CWMDAPP is 1 byte, no scaler, x181BD in the M-box and G-box, likely the same in K.

Make sure you know check the FM and learn what the axes for your PID variables mean.  Some are requested boost, and same are boost error, etc.   That was probably my biggest hindrance to learning....

and have fun.   Tuning boost is fun, because you get to constantly rod the hell out of your car all in the name of knowledge.

you will have to calibrate your KFLDRL first and foremost. There is a thread on this forum on how to do it specifically, but basically you need to lock WGDC and do pulls for full RPM range increasing it by 10% each time until you reach reasonable max, which would be 80-90% on small turbos and as much as you dare to throw on large turbos (I stopped at 60% on my tials when I hit 38psi or so lol). Once you get the data, you can work supplied spreadsheets to calculate new linearization table for turbo which will make everything MUCH smoother as you will no longer request unsuitable WGDC and over/under boost.

Awesome gents, thanks for the input so far.

DD, re ESKONF, I guess your suggestion re copying from M-box would sort the RS4 airbox flap also, although I do not see anything in the Wiki at ESKONF about this? Also, would I still need to edit CDKAT / CWDLSAHK / CDHSH etc. or would ESKONF sort everything out on its own? I'm asking because with my M-Box flash I edited the individual parameters and didn't touch ESKONF?


I had a look at the 'PID calibration from square one'-thread already before but I obviously need to look at it again, and probably again  :-[

Also, and not meaning any distrust, but can someone please confirm if x181BD is the right address for CWMDAPP in K-Box? Not sure what will happen if I change x181BD and CWMDAPP sits somewhere else!?

I'm guessing the 'PID calibration from square one'-thread, right?

Am I correct in saying that, since with small turbos the boost tapers down at high rpm, if a bigger turbo can flow enough at high rpm and assuming fuelling is adequate, it would not be uncommon to see boost in excess of 1 bar at max rpm - i.e. the 2.7T stock internals will survive that?

I will spend some more time reading and hopefully have some progress to report on after the weekend...

You will still need the codewords to disable function, ESKONF just prevents diagnosis of missing hardware.

Great, thanks!

Do please confirm.  You don't want to be wrong and I can only confirm in the M-box.   

0x181BB in k-box

I'm thinking to be sure about the location of CWMDAPP one may also use an M-Box flash (CWMDAPP at x181BD for sure) in the linearization of KFLDRL, but with WGDC values for KFLDRAPP from the KFLDRL axis in K-Box (i.e. 0, 3, 10 etc.)?

What I haven't been able to find in the FR is where the description of CWMDAPP is and why it should be set to 8 (only bit 3 set)?