NefMoto

Hi, a few months ago a tune was done to the car (Golf IV 1.8T AUM 06A906032DR) by a commercial chiptuner and the car did not work correctly, from certain revolutions began to go very rich lambdas and in driving the car was giving jerks.

Because of these problems and no one knew how to give me a solution I decided to learn a little about my ME7.

The first step was to get another ECU to do the tests and load the original software to see possible mechanical faults. My surprise has been when I have seen that even with the soft stock the car is still running very rich:

What can be caused by this behavior ??


Hello, a few months ago a reprogramming was done to the car (Golf IV 1.8T AUM 06A906032DR) by a commercial chiptuner and the car did not work correctly, from certain revolutions began to go very rich and in driving the car was giving jerks.

Because of these problems and no one knew how to give me a solution I decided to learn a little about my ME7.

The first step was to get a switchboard to do the tests and load the original software to see possible faults. My surprise has been when I have seen that even with the soft stock the car is still running very rich (attached image)


What can be caused by this behavior ??

I think LAMBTS is always giving me the richest values ​​I suppose for a high load, but ... Why? Maybe a high IAT? Could this behavior be associated with some mechanical failure?

Attached the complete logs, sorry if they are not very complete. This is the first time I use ME7Logger (awesome program) but I have the problem of not being able to scan more than 25 variables.

AFAIK, this software is made so it will fuel with KFLBTS as soon as you drive a little spirited.
And that map is pig rich from factory even with a stage 1/2 tune.
Change map to how you want to fuel, or choose a different fueling strategy :)

Yeah, I know it is very rich, but check this logs on a other 1.8T (attached)

It's a 06A906032QN stock, I dont know if car has better peripheral than my Golf (sure has better IC than my SMIC).

But check the desired Lambda, is the same as mine until 4400 rpms.

Maybe if someone has stock logs from Golf IV 1.8T AUM we can compare.

Thanks

Sorry, the attached

That's because the BTS trigger is set later than in yours.
All stuff that is pretty simple to change :)

all this is noted on s4wiki

I think that's not the problem, I'm triggered full time in the log, trigger is set in my stock at 400 ° C and up to 4400 rpms my desired Lamda is the same as the other 1.8T plot, from there Mine is much richer

The other car also has high exhaust temperatures, although I could not check it in the file I think it also has LAMBTS active.


What? LAMBTS, MAF or IAT effect?

I do not understand English and something may have escaped me.

A link to a section you mean would be very useful, thanks.

mine does that too, thats why i went with LAMFA and LAMKR for fueling and just disabled BTS fueling.
but i have a suspicion that TKATBTS is the trigger.
see if its set to 100C, if it is; set it to the point where you want to trigger.

Log tabgbts_w and see how high your modeled EGT values are during all kinds of driving.
You will be surprised how high EGTs are even at idle according to this model, something to keep in mind whentuning tabgbts.

Why?  The precision of the BTS maps is much better than LAMFA.  You can build a much nice curve with BTS.  The fueling table for knock is even worse.

I know, I've been meaning to revert to BTS again because of the better resolution.
But LAMFA and KR have actually been working pretty good for me :)

I'm agree with you seeing tabgm in logs.

This evening I have tried log again with MAF unplugged whit similar result (attached)

EGTs (tabgm) at start of the log 480ºc, 760ºc at wot.

The results are very similar but I think... how precise can be msdk_w (alpha-n estimated air flow) vs mshfm_w (measured airflow)? or my MAF is K.O.? ???

What diferences must be apreciate in a log with MAF vs a log whit MAF unplugged?

Don't bother to try to correlate this. If alpha-n was accurate you wouldn't need a MAF. It isn't. Maybe if it was speed density.

Ok, I only wanted to check the healthy of my MAF and if it was causing fake measures.

But comparing both logs alpha-n was very very accurate.

I'm starting to run out of ideas :(

I've thought to log the lambda variables to see that it's causing such low demands. These are the variables I've seen related to lamda.

ladiff_w         ;{}                                 ; {Regelabweichung Lambda}
lambts_w         ;{LambdaPartProtection}             ; {Lambda f�r Bauteileschutz}
lamelsh_w        ;{}                                 ; {Lambdasoll f�r elektrische Sondendiagnose hinter KAT (Kurztrip)}
lamfa_w          ;{TargetAFRDriverRequest}           ; {Lambdasoll Fahrerwunsch (word)}
lamfaw_w         ;{}                                 ; {Lambdasoll Fahrerwunsch Anteil aus Kennfeld LAMFA}
lamfawkr_w       ;{}                                 ; {Lambdasoll Fahrerwunsch bei Z�ndwinkelsp�tverstellung KR, WL}
lamfaws_w        ;{}                                 ; {Lambdasoll Fahrerwunsch station�rer Anteil}
lamfwl_w         ;{}                                 ; {Offset Motor-Lambda-Soll im Warmlauf}
lamlash_w        ;{}                                 ; {Lambdasoll f�r Test Schwingungspr�fung hinter KAT}
lamrlmn_w        ;{}                                 ; {Lambdasollvorsteuerung zur Verbesserung der Brennbarkeit bei kleinem rl}
lamsbg_w         ;{AirFuelRatioDesired}              ; {Lambdasoll Begrenzung (word)}
lamsoni_w        ;{AirFuelRatioCurrent}              ; {Lambda-Istwert}
lamsons_w        ;{}                                 ; {Lambda-Sollwert bezogen auf Einbauort Lambda-Sensor}
lamzak_w         ;{}                                 ; {Lambdasondenistwert, korrigiert um Zusatzamplitude}

In the attached file there are few more, but not sure if all of they are useful. What variables should log?

Thanks

Four possibilities:

LAMFA (throttle position based)
LAMFAWKR (knock recognition enrich)
LAMBTS (modeled EGT enrich)
ATR (actual EGT enrich)

All 4 are described in the FR (and the s4wiki) at length.

And remember that the lowest map always dictates what lambda is.
LAMBTS and ATR has trigger point you can set.
LAMKR map gets activated when a knock event is present.

Log the "trigger bit" variables, that will tell you what maps are active.

0.75 lambda is normal operation in stock file (sometimes)
triggered by bts and you should never feel something wrong with it

Hi guys! It seems fukenbroken was right.

Logged AFR and it seems LAMBTS produces that lambdas (attached)

Checked KFLBTS map and is not as rich as the demands made, so the only thing that can be is the multiplicator factor (KFDLBTS x KFFDLBTS).

I could not prove it because the definition file I am using:

http://nefariousmotorsports.com/forum/index.php?topic=4378.0title=

It has no defined maps KFDLBTS and KFFDLBTS. Could someone help me find them?
I would like to prove my theory that the first Golf 1.8T AUM (06A906032DR) are richer than the last AUM (06A906032HJ) due to the multiplier factor of LAMBTS. KFLBTS is the same in both.

And my last idea (Potentially a bad idea and I'm sorry because I'm sure it's already been talked about in the forum) would it be a good idea to handle LAMBTS using only KFLBTS and overriding the multiplier factor by putting 0s KFDLBTS or KFFDLBTS?

Thanks!

You can absolutely do this.

That's how I do it.  I can't be bothered with trying to figure it all out and doing it that way makes it much easier. :)

Hi guys!

I have been investigating all this time and doing some tests and I think I have found the cause of the problem: My high EGTs.

In the log you can see how tabgbts_w reaches 900º in 3700rpms reaching 940º in 5700rpms ...

Remember that my car is 100% stock hardware with stock map. But look at the difference with other cars (attached) one of them is a Golf IV also stock exactly as mine (1.8T AUM & K03s & 06A906032DR) and the other an Audi A4 1.8t also stock.

Now ... Since my car has no EGT sensor (they are modeled) I do not know what can be the cause of these EGTs, I have come to think that some component in poor condition or some sensor does not measure correctly it is generating them. Or that the factors of the area where I live make it so, these last logs (attached) are made with an ambient temperature of 25° and gas of 98oct in an area at sea level.

So need a bit of help, Is there any component that can cause an erroneous reading that causes an incorrect calculation of the EGTs modeled? What factors have influence in model EGT calculations?

Thanks as allways  ;)

Timing, load, AFR

And there are something to solve my high EGT problem?

Cant remap the car in that condition and i would like it.

As the EGT is modelled have you tried another 02 Sensor as that's what the reading is modelled off..

Also have you checked the cam timing? as even being 1 tooth off can cause high egts

Well I think if desired Lambda and Obtained lambda They are similar, I can not say that the probes are in good condition, right? Maybe I'm wrong.

I had not thought that, the car is second hand and although it apparently works as it should the previous owner is who changed the belt of the distribution.

Could you check the timing with me7logger in some way? What variables should I log? What timing should have a 1.8t in good condition?

Thank you, I start to run out of ideas, I do not know what else to look at, apparently the car is perfect.