NefMoto

Technical => Tuning => Topic started by: Rick on November 29, 2010, 04:53:30 PM



Title: Requested lambda <> Actual lambda
Post by: Rick on November 29, 2010, 04:53:30 PM
Currently I have a standard MAF with 60lb injectors. 

My KRKTE is scaled correctly, and this shows in my long and short term trims.

Cruise and idle i'm always at lambda 1.  LAMFA is standard, so set to 1 across the board. 

On WOT, when LAMFA is set to 1, will the ECU stay in open loop?  I suspect not, as on my car with maximum boost of 1 bar, the AFR as soon as i go WOT goes to lambda 0.9. 

So, it can't be closed loop anymore?.  I'm wondering if this is due to BTS?  All BTS maps are currently std.  Do they kick in instantly when you reach a certain load due to calculated EGT?  1 bar is obviously more than std boost, so it could be this.  I'd need to disable BTS to try.

The other option is that requested LAMFA (1) is not matching actual lambda despite MAF and injector scaling being correct.  I'm wondering if I can use KFLF to adjust this.  I realise i can adjust fueling using LAMFA and BTS maps but i'd first like the requested lambda to equal actual lamba.

Rick


Title: Re: Requested lambda <> Actual lambda
Post by: ArgDub on November 29, 2010, 09:46:43 PM
Lambda value is selected on the basis of a priority system. The priorities of lambda selection are catalyst protection, component protection(BTS), catalyst heating by secondary air, post-start-warm up... whenever lambts is less than lamfa this one will be overridden by lambts, so when your egts reach tabgbts you lambda starts to follow kflbts. what you describe seems normal to me

Guillermo


Title: Re: Requested lambda <> Actual lambda
Post by: Rick on November 30, 2010, 02:39:44 AM
Hi Guillermo,

I understand the priority system.   I was expecting that the lambda would follow LAMFA for longer before following KFLBTS.  It seems to happen as soon as I go WOT.  This suggests that I am instantly exceeding TABGBTS, which is not the case - a short WOT pull will not raise the EGT enough in reality.

Does the EGT model take this into account, i.e. is there a time delay before calculated EGT starts to rise?

The overall goal I have is to run 0.85:1 WOT for the majority of time, then when the EGT gets too hot have it richen up to 0.7x.  If BTS is triggered as soon as I go WOT then I can't see a way of doing this as BTS needs to be set to 0.7x.

Rick


Title: Re: Requested lambda <> Actual lambda
Post by: Rick on November 30, 2010, 07:18:16 AM
After studying the function in more detail, it appears that there is no time delay for KFLBTS, only for the delta function due to retarded ignition angle.

Calculated EGT for compnent protection is drawn from a map.  There doesn't seem to be a way of increasing enrichment as EGT increases.  Once you are above TABGBTS, you are stuck with the KFLBTS table.  This needs tobe rich enough to keep things in check in the worst case scenario

With ignition retard, the further you retard the richer you can run.  Not really much use for a modified tune as you don't ever want to be retarding from optimum.

The function I still haven't found is the one that looks at the actual EGT sensor inputs.  I've got a feeling that this could be used to trip a further enrichement, which would be ideal.

Rick


Title: Re: Requested lambda <> Actual lambda
Post by: elRey on November 30, 2010, 10:07:32 AM
There doesn't seem to be a way of increasing enrichment as EGT increases.  Once you are above TABGBTS, you are stuck with the KFLBTS table.

There is such a map. I believe it's FBSTABGM. But it probably doesn't work the way you want it to. It factors the KFLBTS value based on temp above TABGBTS. So, it's not a direct lambda value.


Title: Re: Requested lambda <> Actual lambda
Post by: turboskipper on November 30, 2010, 12:57:15 PM
Are you data logging this? Requested lambda and actual lambda should come from a data log, not what you think LAMFA, LAMBTS and so on will give you. If your charge detection and fuel mass parameters are correct you should always have desired lambda equal to actual lambda.


Title: Re: Requested lambda <> Actual lambda
Post by: ArgDub on November 30, 2010, 10:05:30 PM
Rick, elRey is right KFLBTS is scaled by FBSTABGM which is function of EGT, and FBSTABGM is cero when EGT is TABGBTS, so BTS correction is progressive and increases with EGT.

Seems that lambts doesn't cause the increase, maybe there is an accel enrichment function. When throttle is suddenly opened, the airflow will increase more rapidly than the fuel flow due greater inertia of liquid gasoline, which can cause a temporary lean condition, to correct this situation carburators use an accelerator pump, it's logical to think that there is an analogous in me7, but I couldn't find such function yet.

The other option is that your KRKTE not properly calibrated and you see this when you go wot where KRKTE is most significant.


Title: Re: Requested lambda <> Actual lambda
Post by: ArgDub on November 30, 2010, 10:06:08 PM
Are you data logging this? Requested lambda and actual lambda should come from a data log, not what you think LAMFA, LAMBTS and so on will give you. If your charge detection and fuel mass parameters are correct you should always have desired lambda equal to actual lambda.
But, what is the requested lambda?, I mean does anybody know what is the me7 name of the variable you are loggin as requested lambda?


Title: Re: Requested lambda <> Actual lambda
Post by: elRey on November 30, 2010, 10:10:19 PM
with vagcom, mblock 031 usually shows requested and actual lamdba


Title: Re: Requested lambda <> Actual lambda
Post by: ArgDub on November 30, 2010, 10:20:03 PM
with vagcom, mblock 031 usually shows requested and actual lamdba
I know, but (mblock 031- Lambda specified value) must have a correspondence with some variable in ecu, I would like to know which variable.


Title: Re: Requested lambda <> Actual lambda
Post by: elRey on November 30, 2010, 10:24:52 PM
lamsbg_w


Title: Re: Requested lambda <> Actual lambda
Post by: ArgDub on November 30, 2010, 11:05:06 PM
lamsbg_w
thanks


Title: Re: Requested lambda <> Actual lambda
Post by: Rick on December 01, 2010, 07:31:03 AM
Thanks for the replies, useful info here!

What are the axis for FBSTABGM?  One is EGT, but is it calculated or measured EGT?  Is there a measurment block which reads calculated EGT in VCDS?

Rick


Title: Re: Requested lambda <> Actual lambda
Post by: elRey on December 01, 2010, 08:44:01 AM
034 and 035


Title: Re: Requested lambda <> Actual lambda
Post by: setzi62 on December 03, 2010, 04:43:59 PM
What are the axis for FBSTABGM?  One is EGT, but is it calculated or measured EGT?  Is there a measurment block which reads calculated EGT in VCDS?

Rick
The FBSTABGM has only one axis and the input is from variable tabgbts_w. This should be
calculated EGT. The name of the map is "Faktor BauteileSchutz Temperatur AbGas Modell",
i.e. "factor for part protection exhaus gas temperature modeled".
Measured with VCDS in group 217 or 214 (if available) are lamsbg_w, lambts_w, lamfa_w, tabgbts_w.


Title: Re: Requested lambda <> Actual lambda
Post by: Rick on December 04, 2010, 11:47:37 AM
Ok, found  the map.

It is 4x1, and as stock is set to 1 for every value.

Rick.


Title: Re: Requested lambda <> Actual lambda
Post by: Rick on December 04, 2010, 12:14:10 PM
So i now only have one more puzzle to solve, and that is what the EGT sensors are used for.

I translated this little snippet, but it doesn't make a lot of sense to me at this stage.  It could be suggesting that if the actual measured temp is above calculated, then measured temp is used:

* If the actually measured temperature, the temperature has reached critical components must be read from the modeled temperature and registered in the corresponding threshold. Possibly. Error in the exhaust gas temperature model are thus in the emerging TABGBTS, TKATBTS, TIKATBTS and TWISTBTS again

Rick


Title: Re: Requested lambda <> Actual lambda
Post by: Rick on December 06, 2010, 04:06:28 PM
Bit of an update.

Yesterday i found my long term fuel trims were around -22%!

This would explain why when I went WOT and open loop the AFR was going rich.  I've since rescalled KRKTE and LTFT and STFT are less than 1%.  When I went WOT lambda stayed around 1, which is what i would expect.  The car felt noticeably slower now it's staying at lambda 1.

I then rescaled the axis in LAMFA so that i could request different lambda at different requested torque.  I requested lambda 0.85 above a requested torque of 70.  This has richened the mixture, but not to 0.85 - it's more like 0.91.  I could just tune LAMFA until I hit lambda 0.85 on WOT, but I'm going to edit KFLF to see if i can get requested lambda to equal actual lambda.

ElRay mentioned that calculated EGT can be read in block 34.  I don't think this is correct, as that is catalyst temp which will be lower than the EGT for component protection, can anyone confirm?

Rick


Title: Re: Requested lambda <> Actual lambda
Post by: fknbrkn on October 17, 2013, 02:35:44 PM
Hello
maybe im asking very stupid question

but i have a problem - requested lamda goes down from 1 to 0.8@5000 but actual goes the same only before rev hits 3000, after that it raised to 1..

im using 380cc (from BAM engine) with 3 bar fpr  and IDC goes over 100%
today im changing fpr to 4 bar without any changing in the file. i have no time to make normal logs. but saw that IPW 24ms@4500 so it is 90% IDC
but lamda still around 0.95 .. 1.0 at 3000+ WOT

any reasons why it can be happened?

new fuel pump
new filter
new injectors
old o2 sensor


Title: Re: Requested lambda <> Actual lambda
Post by: Rick on October 17, 2013, 03:01:35 PM
It really would help to have logs


Title: Re: Requested lambda <> Actual lambda
Post by: fknbrkn on October 17, 2013, 06:27:54 PM
only vag-com log for now..

well i see that after 4760 rpm IDC goes over 100% again
but why it so lean before..

maybe im not very accurately rescaled MAF but i think it shouldnt affect in the actual lambda measurement. is it?

btw
stft 0.8%
ltft (with 4bar fpr) -3.2%


Title: Re: Requested lambda <> Actual lambda
Post by: Rick on October 18, 2013, 03:41:34 AM
What is STFT when on full throttle and it is lean?  Have you tried with lower boost?


Title: Re: Requested lambda <> Actual lambda
Post by: fknbrkn on October 24, 2013, 11:52:15 AM
found a cracked breather hose .. (a ~year ago ive changed all pcv except this one  >:()
replaced it
pressure test - no leaks

change krkte to match 404cc`s (380@4bar)
and lowering load a bit

still have a problem
take a laptop today and make a short pulls at WOT second gear after flashing
can take normally WOT logs at 3 gear tonight if it needed


Title: Re: Requested lambda <> Actual lambda
Post by: fknbrkn on October 24, 2013, 12:03:51 PM
sorry
bad file



Title: Re: Requested lambda <> Actual lambda
Post by: fknbrkn on October 28, 2013, 04:03:55 PM
replacing o2 sensor make no sense

bying walbro 255lph inline (not fit yet)

maybe it is wrong way? stock fuel pump cant rich @18psi ?

any help needed..


Title: Re: Requested lambda <> Actual lambda
Post by: Rick on October 28, 2013, 04:24:51 PM
Measure fuel pressure under load.


Title: Re: Requested lambda <> Actual lambda
Post by: fknbrkn on November 01, 2013, 02:33:53 PM
thanks for your help Rick

just installed walbro inline today and car pulls amazingly now