NefMoto

Hello.Can someone check my file?I have made calibration for bam 380cc injectors maf vr6 k04 turbo AGU engine 1.8t.The file is not tuned.Only calibration maps.Thanks in advance

 :P
Sure let me look in my magic globe here and see how you did.... Come on for real?? These cars need to be tuned on a wideband lambda for good results, you cant just take the maf curve from the vr6 file and expect it to work proper on the first shot when calibrating injectors too. Last one I DID was 440cc injectors and a vr6 maf with a k04-001 and it took a good 5 edits for fueling and maf, and another 5 or so for boost control and timing. IN most cases youll need to tune startup fueling too.

I had a similar projekt...

- there is no need to change FGAT0, leave it at 1.0
- check your KHFM -> 277/384*43860 -> 31639
- don't forget to reduce TLST.0/1 like KHFM
- check MLHFM, linearization too low for VR6 MAF -> 989@4.98V (859@4.98V)

good starting config to work on

Thanks a lot Marty and Leonhard for the response. You gave me useful information. Thanks again

Another question.This is an Excel spreadsheets to calculate afr from 002 block from Ross-Tech site but it doesnt work for agu log.I think must change the constant value.What is for m3.8.3 ecu?

you can use Block Data 002 as well, the only difference is load unit:
M3.8.X load comes with tl [ms/round], not rl [%]
check Block data 008 and 009, it's lambda-controller and fuel-adaption for lambda=1
and don't hasitate to use a wideband lambda for matching all over

Thanks.i will check these blocks.How i can make the excel file to work for my logs?What need to set on load cells?

yes, fuel trims only working for part load (lambda=1) narrowband,
but gives you an idea where to go.
This excel calculation needs to be feed with a precise air mass signal,
that you don't have right now.
Go with wideband and compare/tune calculation results at full load

Ok i will try.Thank you very much Leonhard

Just a comment I will not mess with MLHFM that way because you will modify load(ms) values in all related maps in that ECU.

yes, load-range (0 - 12.8ms) will be scaled according new Injector-size.
This has to be considered in ignition-maps etc.
I prefer this way because load(ms) keep it's ori weight and Block Data 002 will show the new ratios.

if you are using a VR6 complete MAF why no just use the MLHFM tested OEM flow values from the donor car. After that you can tweak final KHFM value to follow OEM ECU low load(ms) data as a reference. In my case and some others at 1000mts altitude above sea level, with K04, GT28xx and G25-xxx turbos you can reach around 18-24psi boost up to last diagnostic LOAD value of 12.75ms with injectors from 350cc to 630cc. In that cases the suggested KHFM value = 31639 was low. Last word in this ECU is trial and error. Just a comment.

yes, that's what I recommended for his config.
@Gpepes2012: I put the ori VR6 MAF linearization curve in this file unless you don't have it

Thank you very much for the help. I appreciate it. What exactly does kftlws represent? I see in the axis that it has a percentage of throttle rpm and load.I know that all these are noob questions but I try to understand how it works from the logs

you are asking the right questions  ;)
It's the load-calculation (precontrol)
very simple explained:
driver's demand (throttle position) of load (injection in first place) over rpm.
Corrections are following, like warmup, afterstart, lambda-control, WOT enrichment ect.
and yes...it should be corrected/lowered as well, after increasing injector size.
Bigger injectors needs less opening time ti[ms] for same inj.quantity -> ti is part of this load unit

Anyone know the correct factor for M3.8.1 (VR6) MAF Linearization? Don’t look the same as M3.8.3 (AGU), and the damos say they use same factor, but that don’t really make sense tbh.
-Didn’t look at ir awhile, so can’t remember why it dosn’t look right (:

So kftlws needs adjustment or not since the injectors end at 16.32 ms? After proper injector calibration kflf needs adjustment?In general, the strategy from now on is based on the actual load? Can I increase the request load and the valve duty kfldtv by% to achieve the desired?Thanks in advance

adjusting KHFM to bigger injectors will reduce load scale, means all load related maps/values needs to be shifted.
For example: 14psi are about 10ms load with original setup, changed KHFM in your case will lower this to around 7ms.
K04-20/23 won't need more than 12ms with new scale.
match fueling first at part load, measure boost and get (a feeling for) the new boost/load/airmass ratio (Data Block 002)
check/adjust base boost (N75 -> 0%), than build your new KFLDTV for desired boost target.

Thank you Leonhard for helping me a lot to understand how it works. One last question about kflf.Assuming I make changes to% in kflf at idle and in normal driving according to block 099 lambda reg and everything is ok does this automatically mean that everything will be fine in wot?I ask why a wideband gauge is not available at the moment. Or do I definitely have to have a wideband?

KFLF is actually not the map for basic fuel tune, it's for enrichen fuel at higher loads, outside lambda=1.
you see an factor and not lambda within KFLF. Factor 1.0 should stay for desired lambda=1, you can/should shift these factors according to new load scale. Check also TLRAN(load threshold lambda control) and TLWARN(load threshold fuel adaption) for shifting to new scale. Tweak KHFM to get close within lambda=1 area, KFLDSAK can help and TVUB might be a choice too.
And yes...Wideband is a must to verify at higher loads.

Thanks again. I made a log to see if everything is ok. I have it all but I have not installed anything(vr6 maf,bam injectors) yet except the k04 turbo.Everything stock except turbo. Does the wastegate need some adjustment? Why do I have a higher actual than the request?

KFLDTV (N75) precontrol starts with 62% at this point(map), too high for requested load with K04.
Boost (air mass) control isn't fast enough to hit Target. K04 just delivers more air mass.
If you like to match K04 with ori setup, reduce values in KFLDTV.
Try not to touch wastegate-actuator hardware, you might need this setting for later.

If that log is with K03 than I would say that is a mehanical problem

k04-023

Thank you very much Leonhard. I appreciate your help. I have to reduce the tlran so that I am in a close loop so that the ecu is in control? Can you tell me in simple words what exactly does kfvakl and kffa do?My car is decat.Need to zero kffa?Can i adjust kfvakl by log ms on deceleration?

yes, TLRAN is the threshold for close loop lambda=1 control and should be corrected for bigger MAF and injectors too.
KFFA is for cat protection against overheating, it will be added to factor of KFLF.
For example 1,21+0,08=1,29 at 7ms-5500rpm, get's very rich for cooling exhaust gas. And yes, no need without cat.
You might think about measuring exhaust gas temperature, if I remember correctly, K04-023 has a hole for sensor.
You can adapt a common NiCr-Ni thermocouple to tweak full load lambda.
You need at least the voltage signal of ori NB lambda sensor for checking KFVAKL, being still too rich or lean enough.
Respect dead times of these values. WB-sensor would be more precise.

I have never seen so many details and help anywhere. The best site with the best tuners ever.One last question. How can I tell if the injectors have reached the limit?Stock injectors max 16.32 ms and bam injectors 16.32 ms too.If I put 630cc later I will do the same procedure in the calibration?Khfm,TVUB,TEMIN,TLST.....?

basically you can calculate:
injector opening time(te) - that being injection quantity - is limited by duration of 2 revs (4-stroke)
for example: 1000rpm -> 1 rev = 60ms -> 2 revs at 1000rpm are taking 120ms; 6000rpm -> 20ms; 8000rpm -> 15ms
If you reach these values for te, get bigger injectors.
And yes, but with 630cc Injectors you might be able to max out the range of VR6-MAF.
Good luck, looking forward to read from your project

16.32ms injection time is diagnostic limit, it will not log higher, no matter what (:

Anyone can confirm factor for MLHFM?

M3.8.3 AGU - Factor 0,125000, offset -200 - Max  673kg/h

M3.8.1 ABV - Factor 0,125000 - Max 989kg/h

Yes 4.98---> 988.63kg/h

It's not diagnostic limit, it's scaling limit IIRC in ECU. Or the scaling limit is close to that.

Scaling limit?

i know load max at 12,75ms (8bit, 255*0,05 = 12,75), but not sure what you mean with scaling?

Just took it for diagnostic as it will max at 16.32 but it will inject more (:


Throttle angle?
I noticed that my AGU only goes to 82,9% WOT (Using pedal and manuel on the throttle)
-Original throttle axis goes from 77% to 94% so it will interpolate between those columns! (Example: going from 8ms in 77% to 10ms in 94% will land you just under 9ms at WOT - 82,9%)

anyone else notice that? or is it just my throttle that's worn? (Changed axis to 82% so I could use the whole map)

Another question. According to the wideband in which areas kflf map do I make changes? From where to where is the area for idle, part load and wot? The changes must be made in each area separately by% to the desired?

For example is this right?

Your X-axis is wrong. (Factor 0,050000)

You log RPM, load and AFR (This is tricky as you probably log RPM and Load on PC and AFR by the gauge)

then you look at RPM and Load at the AFR you want to correct

Looking at my map:
Let say 6000RPM and 10ms you want to lower AFR, then increase 1,359 (35,9%) so it correct for more fuel.

log Load at Idle to find idle area (Pretty sure it's up to 1,5ms, maybe even up to 2ms)
and so on.

Thank you.For example how much need to increase 1,359 value if afr is lean?

Right now it inject 10ms*1,359 = 13,59ms (so in 13,59ms it inject fuel)

If you want to go leaner (Rise AFR) you need to decrese the value.

10ms * 1,300 = 13ms. (you inject less time = less fuel)

Ofcause you have other correction maps, so the real output will be different.

Hello, I open the topic again. Suppose I put 630cc injectors.Khfm must be 277/630*43680=19205?

Hey, im having the very same project (440cc inj, k04001). A friend of mine doing the software but runs in some difficulties with linearising injectors. We didnt touch MAF map so far, running stock maf right now.

Would you be willing to share your file with us? You would really help us out!!
Best regards, Moritz