Hi all,
I got the car for myself with an unknown tune in it. All hardware looks stock 1.8T AUM 06A906032HJ with (k03 or k03s did not check yet).
On WOT run I noticed weird behavior: the power of the car is oscillating back an force. First thing I did logs and air leak check. Found some leak in the pre intercooler hose.
Tightened that one and it all holds now 1.2bar of pressure (used air compressor to test) with no hissing all the way from the intake hose (after MAF) and through the intake manifold.

I did the WOT log runs to understand what is happening, I can see that PID is constantly trying to adjust boost to fit desired. After the first overshoot it oscillates up and down all the way to the red line causing sometimes 0.3-0.5bar difference in actual pressure depending on the gear and how quick the boost gets to the first peak.
Any hints what to try to adjust? Start with PID-I max and lower it?
I remember I tried 4-5th gear WOT pool on a highway and actual boost followed the requested boost. So there is something with PID how quick boost reaches the setpoint.
If it is slow rising boost and RPMs, current PID can manage and it keeps boost very close to the desired. (I need to find this 5th gear log, can post later)

I also wanted to use LDRPID tool for linearization of WG and boost. I set CDWMAPP = 8 and started with 0% to 80% runs.
However, on 80% WGDC run I am experiencing throttle cut in application mode. As I can understand from the logs actual load/boost is way over the desired.
I can see that with CDWMAPP = 8 my desired boost stays around 1600mbar(0.6bar stock?). It does not follow what is in LDRXN map anymore. See the log and picture attached.
What map can I change to raise limit for desired load/boost for CDWMAPP = 8 runs ?

I am attaching a screenshot of KFMIRL, KFPED,LDRXN, KFLDHBN, KFLDIMX, KFLDRL and 0-70% WGDC runs just in case it can be useful to see something wrong with the turbo or WG.

Any help is appreciated, really want to fix this behavior.
thanks!

Its better to start from scratch for sure

Why are you digging so deep?
fix plsol curve first (flat kftarx, change ldrxn a bit) its not pid related when you request mad curve pid just tryin to follow that

do not change dimx & drl for stock turbo, you may want to add PR to dimx axis with corresponding values for PR > 1100, but do not touch stock ones if its stock turbo

I just wanted to understand why when CDWMAPP = 8 my loadCorrected was folowing LDORXN and not LDRXN. When in normal mode CDWMAPP = 0 there was never throttle cut even with this crappy tune and waste gate oscialtions.
So I wanted to do linearisation runs to understand what current stock turbo is capable of. I had no intention to keep LDORXN  = LDRXN in normal mode.

What I noticed as well with fixed WGDC 80% run, does MAF value of maximum ~157 g/s makes any sense at 6200 rpm with 0.8bar boost and only 100-130 g/s in the 3000-4000 rpm range?
Or there is also something with MAF?
I think I saw the real stage 1 VCDS logs with 150 g/s already in 2500-4000rpm area and like 175 g/s at the tail with k03s on AGU engine.




I agree that I have to start from scratch.
I will take stock file, and run it stock to do some logs to make sure everything is happy stock first and start from there.
I saw there are two versions of 06A906032HJ ver 0002 and 0003 on the forum.
My car has unknown 0002 right now. Is there any major differences between v0002 and v0003? 

I did flash AUM stock file,
The log was done in 2nd gear. There is an overshoot in 2500-3000 rpm area.
First overshoot is going to be there with PID anyway, as that is in its nature as long as it stabilizes after reaching the set point.
Maybe it is a little big bigger because of the 2nd gear pull.

I see it is very comparable to the other stock logs of healthy 1.8T I found in this post.


this is stock Audi A3 8L 2001 with AUM 1.8T should have stock K03s.

before I add something to dimx I can simply make it 1bar stable all the way to 0.8-0.9 at the tail.
This should be possible to achieve by raising the ldrxn a bit in corresponding area where turbo can achieve desired boost.
Plus it has to use 100 WGDC in spool area so that it gets to the desired as quick as possible.

You said "mad curve". Do you mean by mad curve: asking too much from the turbo where it can never reach desired load/boost?

For nice request curve you should tweak ldrxn / hbn depending on a strategy.
Did you know what is lambda regulation and fuel trims? Read fr before starting to tweak random things