NefMoto

Hi.

I guess this is a pretty broad question with a hell of a lot of variables.   

Let me run you down on my spec first

1.8T k03s running Audi TT BAM management
82mm 9.25:1 JE forged pistons with IE rods
AGU head with large port inlet
k03s turbo with ported manifold - stock actuator
Forge supersize dump valve - recirc
Custom free flow 2.5" pipework into a large core intercooler
AGU coilpacks
All the N valves removed for dump valve, evap, sai ETC.
3" TIP with large foam filter

At the minute, im requesting fuel as below through lamda, and using BTS as my EGT enrichement protection map.
(http://i93.photobucket.com/albums/l42/carsey2006/lamfa-2.jpg)

At the moment, it seems alright on a k04 standard map,  10psi was pulling 160g/s with stock fuel and timing, but some CF's

Now ive tried to raise the load through LDRXN to mid 160s and that gives me about 15psi according to my boost gauge which appears bang on.
Only issue Im having is when in high gears, low RPM, high load situations where the ECU drops into 'soft limp' mode and limits me to 5psi until I release throttle and press it again.

Ive tried copying the KFMLDMN maf lower diagnosis table over from a 032 HN tune and that sorted all gear up until 6th where it would occasionally drop into limp.

Has anyone got any ideas what else could be causing it to drop into soft limp?

Attached is my 'latest' file which im using today.  No CF's it appears, just the occasional soft limp I describe above.   Holding 15psi pretty much throughout the revs.

Any advice would be appreciated.

Maybe I don't understand what you're trying to accomplish, but you should be richest at peak load when the turbo is spooled (and leaner on spoolup). Why are you JUST getting to an ideal AFR at redline?

Why are you upgrading your MAF on a K03?
Why do you have a K03 on a built engine?
Why have you not rescaled your LAMFA axis? The last column should be sub 100 (I'd say 95) so you're actually using it......

You really need to read more. Time to get back to the basics. Without meth injection (or race gas) that fueling is leaving a lot of timing on the table.

Why are you upgrading your MAF on a K03?
Why do you have a K03 on a built engine?
Why have you not rescaled your LAMFA axis? The last column should be sub 100 (I'd say 95) so you're actually using it......

MAF diameter is suitable for the ecu.  im running a S3 housing on a TT map....like standard.
K03s is because the engine required running in, so opted for a standard turbo to run it in, make sure everything was OK, before I went bigger (TSFI)
LAMFA axis is standard except the 89.9974 row.  I used the k03s RPM values to give me a better fuel control on spool (like an original k03 framed turbo map would)

Didn't even notice the lamfa axis, well spotted!

He's got a K04 064 waiting to go on, he's just running it in on the K03s just now and hence the MAF change.  

I've got fuel coming as early as I can, because if it's lean, it'll break something.

So yeah, I'd have fuelling coming in earlier, and use BTS too - once you've got an EGT sensor.  Lamfa is only preemptive.

The purpose of LAMFA is extra torque when requested by the driver.  It allows the enrichment of the mixture depending on pedal position/req load.  

Quote:
From the factory it is disabled by setting it to lambda 1 everywhere, and having an unuseable requested torque axis of 0-1%. Reqested will always be above 1% so only this load column gets used.  The idea is the factory wanted to run lambda 1 everywhere for emissions and economy, and only enriched when egts' rise.  What this means is that if you try and use LAMFA to tune fuel factory axis you will always be rich.  Change the axis to 0-100% and now you can set the fuelling according to load, which is how it should be done.  For example at 80% req load and above i usually request lambda 0.82, then enrich a little more with KFLBTS.

Yeah, but it'll interpolate between 89.99 and 100 to get a mean.  Set it to 99.99999999 and it should be reeeet. 

Why have it interpolate at all? I'm sorry, but what you have there is total shit.

Unless you have a 50 trim that hits peak boost at 5000rpm, and are running meth, in which case it's still shit, but not as bad.

I just want you to know I am very pissy today, so don't take anything I say as an attack or insult. I'm just reveling in my inner shithead right now.

Still shit.

Set it to 95 or even 97.5.

So setting the last row of lamfa to something slightly lower than 100% would be good.   I get that.

However, would it be causing my drop into limp mode issues im seeing where I get 5psi?

At the moment, the car feels pretty damn quick, just can occasionally smell a bit rich yet fuel trims are 0.8% idle and ~1.8% partial.

I know theres a absolute load of changes between the k03s and a k04 file, which is why I want to try get my map to a point where its producing good boost....21psi spike and tailoring down to whatever it will manage without it dropping into limp mode on high load situations.  Thats my issue right now.

I was getting a misfire/boost cut on spoolup with a higher load map, but have since raised KFLDIMX to accomodate that like a normal tune would.

Im new to all this stuff, and only changing what i have read about on here and the wiki.

Here's the thing. Running a K03 on K04 software or vice versa is a bad idea. There are maps that translate requested load into requested boost. (KFPRG, KFURL). You'll need to spend some time getting these right so you are getting the correct boost to hit requested load.

I'm a firm believer (now anyways) of using the software that was designed for the configuration. I think too many people are crossflashing these days.

Are you logging AFR? If not, that's another can of worms entirely. Running software not designed for your engine, along with a different MAF housing, there's no telling how close actual AFR is to requested. The entire calibration will be wrong. What you need to do is put the stock MAF back in, and dial in your fueling that way. Log wideband, and adjust the correction maps so that actual lambda follows requested. Then go about swapping the MAF.

I don't like your LAMFA at all though.

I havent cross flashed... Im using a 018H ecu with a 018H file which im modding.  Why would I put a smaller 70mm maf diameter in when the maf housing isnt scaled to that ECU?  To me, that rings bells that it would cause issues.

Fuelling is following requested through logs with ME7 logger to within 0.01 through logs

Cross flash meaning running a k04 file on a k03.

You change back to the stock MLHFM when you put the stock MAF back in

I'm still missing why I need to put the smaller maf for the k03s turbo back in?   Surely with a properly scaled maf to the ecu,  which this maf off a s3 is to my TT ecu being both from the same engine,  the readings will be what they are meant to be without being scaled.  Smaller housing would read the same airflow, just at a higher g/s?

Airflow is measured in g/sec.

I think what ddillinger meant is put the intake back to stock while you get fueling in line and get AFR actual to meet requested.

I havent got a stock intake.  99% of my engine is custom built to ensure good flow.

I think dd and others may have thought you were using the ori sensor in a larger housing instead of using a whole MAF assembly already in use from Bosch for the 1.8t motor in the S3/TT225. If you have copied the stock values from the file that had the larger OEM maf you will be just fine and should be able to count on your g/s values being accurate.
 

Just to clear it up.

Im using a FULL Audi TT me7.5 049 sensor from an wideband TT/S3.   Its fitted in the original 3" housing.  Im using a UNTOUCHED MAF tables from my 018H ecu, which I presume are perfectly calibrated to the 3" housing I am using. 

The only values I have copied across are the KFMLDMN values from a 032HN 180bhp k03s file.  I believe these are the lower values the maf needs to see or else it will trigger a soft limp mode?  The reason ive swapped these is because a k04 will obvilously see higher G/S at certain RPM/throttle sites than the smaller k03s.  Also done the same with the upper threshold for MAF diagnosis.

I was hoping that would stop me seeing limp mode on high load, high gear RPMs where the car would drop into 5psi 'limp' mode until throttle released and then depressed again.

I could be doing it wrong but one lad who uses these forums told me to look at those maps for the cure to soft limp mode.

try swapping the diagnosis maps over also, make them all match the bigger MAF settings. Really just switch every map over that involves the MAF. that way you can have OBD2 diagnosis on your new MAF that will be accurate also. Maybe that is why you get the soft limp because the ECU is confused due to settings used for the smaller housing which used the 049 sensor.

this is how I set mine up and have had 0 issues, you don't have to do anything but copy over the maps. I'm running a k04-023 on an AWP 1.8t. with the TT225 Samco TIP and 225 MAF. I was using a refurbished  225 MAF and kept getting the boost leak DTC. I bought a used real audi/vw 225 MAF and that went away which is interesting because I wasn't getting a bad MAF dtc either until I switched those maps. which is why i in turn bought the used  "real" MAF.

Im using the k04 maf settings on a k04 file.  Only areas I can see which would cause a issue is the ones where the k03s might under-read on some occasions due to the less air going into it.

Are there any of the k03 maps I should be switching into my k04 file to allow the ECU to see the 'smaller' turbo properly.

What did you do about LDRXN?

LDRXN is running my custom settings,  about 160s midrange and 168 load top value.

Soon as ive added extra load, its dropping into soft limp mode.  Fuelling is as above.  The latest map i have is attached in 1st post.

We need logs. 168 load on a K03 is pushing it. You need to log requested boost, actual boost, and wastegate duty cycle.

Just to note its a k03s.  Last one I had, albeit with stock actuator was spiking 21psi and holiding high teens with about 12psi at redline.

15psi across midrange is medi-ocre I think.

Only log I have which isnt the best as my me7logger isnt working as keeps disconnecting is this one attached.   Far from ideal to tell much tho.

IF you are running a k03s on a file calibrated for a k04 the first thing I would change turbowise is the boost PID settings to what a stock k03s file would have and start from there. KFLDRL(iirc, the map that adjusts WGDC is what I mean anyway) should be made much much closer to stock k03s file calibrations. This should take care of some of your woes.
I used to get 15 psi boost on my old k03s at like 1800 - 2000rpms so adjust ldrxn accordingly. The highest I've seen on my boost gauge at that rpm range with my k04-023 is about 10psi

lol If I had your setup I would see what I could get out of that k03s before it grenades... Hello 30+ psi..


before anyone else rips you a new one as its posted in lots of places here on the forum..  .xls files suck, save it as a .csv, no one is even going to DL that file most likely.

I'll have a look at kfldrl.  Thanks.  The k03s, on a completely stock file was holding 10psi lovely.

As I said my me logger is being a pain at the moment so will get some logs from that when I work out what's causing error 8 on it.

I'll give a go of copying all the boost pid maps from a k03s and report back

consider using ecuxplot to look at your .csv.

and as others have said, please don't upload .xls (or any other .csv).

The original is best, as it is guaranteed that plotting programs and other things can read it, and guarantees that all the headers are intact.

Only log i have from ME logger was when the car was on a totally standard k04 ecu file with the k03s and the rest of my mods.   Its attached if anyone wants a look.

Edit:

Swapped the KFLDRL maps over from the 032HN file, and rescaled the axis so all values are matched.  Also copied the KFLDIMX map over aswell to match the changes RPM axis.
Will have to try find some time through the week to flash it on and get it logged properly, see what its doing with the load map.