Pages: [1] 2
Author Topic: Lambda request in BFB engine is driving me crazy...  (Read 8092 times)
technic
Full Member
***

Karma: +14/-4
Offline Offline

Posts: 227


« on: July 09, 2013, 10:32:49 AM »

Guys, I need some help. I have been working with this car for a long time; just to make a simple stage 1 tune  Roll Eyes
The problem is that I cannot get requested lambda quickly enough at low revs when the turbo starts to spool.
It gets (dangerously) lean... and the problem here is lambda requests. (actual lambda follows requested fine)

I have attached logs and original file (below is an extent)

Timestamp   RPM   Load   Boost (Req)   Boost (Act)   Lambda Act   Lambda Req

    /min    %    mbar    mbar      
657.85   2000   99.2   1820.0   1300.0   0.961   1.000
658.11   2080   110.5   1840.0   1420.0   0.969   1.000
658.35   2160   118.8   1860.0   1530.0   0.945   0.938
658.61   2240   127.1   1910.0   1660.0   0.922   0.922
658.85   2320   139.1   1970.0   1810.0   0.930   0.922
659.11   2440   154.9   2050.0   2000.0   0.922   0.914
659.35   2520   167.7   2110.0   2110.0   0.914   0.906
659.60   2680   172.2   2110.0   2120.0   0.899   0.906

659.86   2760   173.7   2110.0   2090.0   0.914   0.899
660.10   2880   172.2   2110.0   2070.0   0.906   0.899
660.36   3000   175.9   2110.0   2100.0   0.906   0.891
660.60   3160   175.9   2140.0   2110.0   0.899   0.891
660.86   3280   175.9   2140.0   2110.0   0.891   0.883
661.10   3360   174.4   2140.0   2120.0   0.875   0.883
661.36   3520   172.9   2150.0   2130.0   0.891   0.883
661.60   3640   172.2   2130.0   2140.0   0.891   0.875
661.85   3720   171.4   2130.0   2140.0   0.875   0.875
662.10   3880   171.4   2120.0   2130.0   0.891   0.875
662.35   3960   169.2   2120.0   2110.0   0.867   0.867
662.60   4080   169.9   2100.0   2100.0   0.867   0.867
662.85   4200   168.4   2090.0   2090.0   0.867   0.867
663.10   4320   168.4   2080.0   2070.0   0.867   0.860
663.35   4440   167.7   2070.0   2050.0   0.860   0.860
663.59   4520   169.2   2070.0   2050.0   0.860   0.860
663.85   4640   169.2   2060.0   2040.0   0.852   0.860
664.10   4760   169.9   2050.0   2040.0   0.844   0.860
You can see that @ 2100mBar boost request the lambda is 0.906 ... and I have seen higher in same situations.
(In the above example the time is short before enrichment starts, but please have a look at the attached logfiles)

Someone on this forum gave me a hint that it could be TLAFA. But that seems to be only 0.2s according to this ME7 logger -log. See below.
No problem with other 1.8T, LAMFA always works (if it is not a BEX engine - it has the same "problem" Wink




Logfiles attached. The file : a4_log_20130404_142045.xls is based on a stock run to try to figure out TLAFA delay value.

So, I'm pretty much out of ideas  Sad If someone could help in any way I would appreciate it.

Cheers
/Lars
« Last Edit: July 09, 2013, 10:36:47 AM by technic » Logged
phila_dot
Hero Member
*****

Karma: +161/-11
Offline Offline

Posts: 1705


« Reply #1 on: July 09, 2013, 08:39:02 PM »

Set TLAFA to zero and raise ZKLAMFAW considerably
Logged
technic
Full Member
***

Karma: +14/-4
Offline Offline

Posts: 227


« Reply #2 on: July 09, 2013, 11:13:14 PM »

Thanks. One problem is that I can't find neither TLAFA, nor ZKLAMFAW. There are no suitable damos available and crosscorrelation (based on logged value) with other me7 damos gives several options for TLAFA address. I have tried many of them...

Could anyone perhaps help me with locations of these?
Logged
nyet
Administrator
Hero Member
*****

Karma: +514/-109
Online Online

Posts: 11369


WWW
« Reply #3 on: July 09, 2013, 11:18:32 PM »

Set TLAFA to zero and raise ZKLAMFAW considerably

It looks like TLAFA is already zero in all ME7.1 2.7t files that I can find it in... does that sound right?
Logged

ME7.1 tuning guide (READ FIRST)
ECUx Plot
ME7Sum checksum checker/corrrector for ME7.x

Please do not ask me for tunes. I'm here to help people make their own.

Do not PM me technical questions! Please, ask all questions on the forums! Doing so will ensure the next person with the same issue gets the opportunity to learn from your experience.
technic
Full Member
***

Karma: +14/-4
Offline Offline

Posts: 227


« Reply #4 on: July 10, 2013, 01:12:51 AM »

The time it takes for the value in lamfaw_w to propagate to lamfaws_w is 0.2 seconds according to log. So TLAFA should be 0.2 in my opinion.
The question is where it is located.... Any clever souls with IDA setup for C167 and ME7 memory layout are more than welcome to give it a shot Smiley

ZKLAMFAW i think is at 26F10. But should this really be raised? According to FR it is the timeconstant (T) for a first order lowpass.

val(new) = val(old) + (in - val(old)) * dT / T.

So if the expression dT/T goes towards zero, the new val = old val. So we would like T = 0 ?



Logged
IamwhoIam
Hero Member
*****

Karma: +31/-54
Offline Offline

Posts: 852


« Reply #5 on: July 10, 2013, 02:52:58 AM »

KFTLAFA_0_A at 195F9
KFTLAFA_1_A at 19629
KLTLAFAKR at 19659

I'll let you figure out map sizes by yourself, this should already help you I think.
Logged

I have no logs because I have a boost gauge (makes things easier)
technic
Full Member
***

Karma: +14/-4
Offline Offline

Posts: 227


« Reply #6 on: July 10, 2013, 04:35:16 AM »

@IamwhoIam : Thank you very much!

Is KLTLAFAKR same as KFLAMKR = f(nmot,rl) in my ME7.5 FR? (The only one me7 doc I have)

You suggest I use this to take care of the lean situation instead of ZKLAMFAW and TLAFA?
Logged
phila_dot
Hero Member
*****

Karma: +161/-11
Offline Offline

Posts: 1705


« Reply #7 on: July 10, 2013, 05:48:04 AM »

The time it takes for the value in lamfaw_w to propagate to lamfaws_w is 0.2 seconds according to log. So TLAFA should be 0.2 in my opinion.
The question is where it is located.... Any clever souls with IDA setup for C167 and ME7 memory layout are more than welcome to give it a shot Smiley

ZKLAMFAW i think is at 26F10. But should this really be raised? According to FR it is the timeconstant (T) for a first order lowpass.

val(new) = val(old) + (in - val(old)) * dT / T.

So if the expression dT/T goes towards zero, the new val = old val. So we would like T = 0 ?

I'm not familiar with your ECU, but that's not how it works in the S4 M box.

http://nefariousmotorsports.com/forum/index.php?topic=2637.0title=
Logged
IamwhoIam
Hero Member
*****

Karma: +31/-54
Offline Offline

Posts: 852


« Reply #8 on: July 10, 2013, 06:07:23 AM »

@IamwhoIam : Thank you very much!

Is KLTLAFAKR same as KFLAMKR = f(nmot,rl) in my ME7.5 FR? (The only one me7 doc I have)

You suggest I use this to take care of the lean situation instead of ZKLAMFAW and TLAFA?


KFTLAFA=TLAFA as a 3-dimensional map
Logged

I have no logs because I have a boost gauge (makes things easier)
technic
Full Member
***

Karma: +14/-4
Offline Offline

Posts: 227


« Reply #9 on: July 10, 2013, 06:31:31 AM »

Ok, so 8bit, 8x6, factor 0.1, axis are load vs rpm? Correct?

Quote
Is KLTLAFAKR same as KFLAMKR = f(nmot,rl) in my ME7.5 FR? (The only one me7 doc I have)

Was this correct also? And could it be used instead of messing with KFTLAFA and KLTLAFAKR to achieve same goal?

@phila_dot, thank you. Then there could be a difference i think....

EDIT: 8x6
« Last Edit: July 10, 2013, 06:36:51 AM by technic » Logged
technic
Full Member
***

Karma: +14/-4
Offline Offline

Posts: 227


« Reply #10 on: July 10, 2013, 06:37:30 AM »

Maybe one axis is gear... 
Logged
IamwhoIam
Hero Member
*****

Karma: +31/-54
Offline Offline

Posts: 852


« Reply #11 on: July 10, 2013, 08:15:08 AM »

axises are RPM vs gear (gangi)

KLTLAFAKR is Verzögerungszeit Lambda Fahrerwunsch bei Zündwinkelspätverstellung

KFLAMKR is Wichtungsfaktor für Anfettung bei Zündwinkelspätverstellung and that's located at 19128 6x6 map RPM vs load, factor 0.007813.
Logged

I have no logs because I have a boost gauge (makes things easier)
technic
Full Member
***

Karma: +14/-4
Offline Offline

Posts: 227


« Reply #12 on: July 10, 2013, 02:23:41 PM »

Ok, but I must still alter ZKLAMFAW because, even if the delay from lamfaw_w to  lamfaws_w is set to zero using KFTLAFA, there will still be a delay in the lowpass filter influenced by ZKLAMFAW ??

Logged
technic
Full Member
***

Karma: +14/-4
Offline Offline

Posts: 227


« Reply #13 on: July 11, 2013, 01:11:46 AM »

Thank you for all input, guys. Really appreciate it.
I will do a few logs with TLAFA set to zero, but I would like to try changing ZKLAMFAW also.
The only place I have found something when 2d-correating is at 26F10, but it doesn't match up as good as it should.

Anyone know where it is in this binary?

Logged
IamwhoIam
Hero Member
*****

Karma: +31/-54
Offline Offline

Posts: 852


« Reply #14 on: July 11, 2013, 07:00:45 AM »

KFZKLAMFAW_0_A at 26D88 8x6 16-bit
KFZKLAMFAW_1_A at 26DE8 8x6 16-bit
Logged

I have no logs because I have a boost gauge (makes things easier)
Pages: [1] 2
  Print  
 
Jump to:  

Powered by SMF 1.1.21 | SMF © 2015, Simple Machines Page created in 0.022 seconds with 16 queries. (Pretty URLs adds 0s, 0q)