NefMoto

The goal of this thread is to use a collaborative effort to provide guidance about some of the basic concepts and procedures involved with creating a stage 1 tune for the 2.7T S4 starting from a stock file.

Here's a quick link to the post where work done on the stage 2 phase is contained.
Stage 2 Change Summary (http://nefariousmotorsports.com/forum/index.php?topic=6695.msg79082#msg79082)

This page is the summary of the Stage 1 phase.
Stage 1 Changes (http://nefariousmotorsports.com/forum/index.php?topic=6695.msg72750#msg72750)

http://www.myaudis4.com/audi-b5-s4-ecu-tunes/
http://www.myaudis4.com/stage-1/
http://www.myaudis4.com/stage-2/


Files
http://nefariousmotorsports.com/forum/index.php?topic=6695.msg92145#msg92145
http://www.myaudis4.com/tuning_images/xdf_bin_files/


[editors note]
The stage1/stage2 nomenclature used here is a bit confusing and a bit non-standard. Typically, stage2 refers to full exhaust (dp+catback). I don't really know what stage2 refers to here - nyet

Ok. So what we want here is a clear, concise list of what a stage 1 file should consist of.

Guidelines that must be adhered to:

1. Safe for STOCK HARDWARE.
2. Safe for either 91 or 93 octane
3. Simplicity

I think we start with a list of maps. If you feel something should be included, post your reasoning. I will be cleaning this thread up periodically to make sure it stays easy to follow :)

I think we should show detailed information of each map what the axis's mean and when you change a value what it does.

(http://i60.tinypic.com/10cj9s8.jpg)

My list:

LDRXN - requested load cap (increase boost)
NMAX - rev limiter (easy, because it is a single value!)
VMAX - speed limiter (easy, because it is a single value!)
LAMFA - one of three ways to do fueling.. I'm not sure it is the easiest, because it requires two somewhat confusing related changes:
1) req torque axis - changing axis data is tricky, and to make matters worse, stock axis values are kinda screwed up
2) low pass filter/time constant changes - the time constant value is somewhat odd, in that its units are strange.

We could alternately do fueling via BTS or KR, but both have their own oddities. Opinions?

KFZW/KFZW2 - timing. The easiest thing to do for the guide is to tell the noob to just keep the two maps the same in high load/rpm areas... would talking about cam position make things overly complicated?

NWPMBBR/VMPMBBR - allow left foot braking (easy, single values!)

In TunerPro, CTRL-F for NMAX

Change this numeric value to 7000 or 7200 (I did 7000), save the bin, then apply ME7Sum and ME7Check, then it is ready to be flashed to the ECU using the Nefmoto flasher.

*Edit* I noticed in the S4 Tuning Wiki it mentions bumping up NMAXF (RPM DTC limit) to 300 rpm over NMAX, so I have revised my tune accordingly.

(http://www.myaudis4.com/tuning_images/nmax.png)

Step 1. Learn about internal combustion engines and understand how and why they work.

Then everything else can fall into place.

Far too many people don't seem to understand even basic concepts absolutely vital to engine tuning of any kind.

Most people have no idea why timing should be removed somewhere and added else where except for maybe they've heard/read that peak TQ will have the lowest value (not always true) or something to that respect. Or why targeting a certain lambda value on one engine/setup won't work for another. Most don't understand the other many influences that happen in a dynamic running engine.

If just creating a flow chart is the purpose of this thread then ignore my comments I suppose.  But changing a bunch of numbers around that someone can only relate to as more is better here and less is better there because someone said so isn't "tuning".

I understand, and agree. One of the prerequisites here is that people have a basic understanding of how an engine operates.

My vision was to treat the process like an apprenticeship, seeking out a few mentors to take people willing to learn through the process.

I fall into that category. Hardware on the car is almost straight and ready for tuning, will be ready to install downpipes soon and open possibility for some higher power levels. Getting software necessary for tuning downloaded and figured out and will be comparing some files to understand the changes being made before I try to flash anything. A how-to would be incredibly useful.

After looking at the VAVMX/VMAX (Speed Limiter) that Nye referred to (value displayed in km/hr) I decided to leave mine at the stock 143 mph (230 kph) level because that is faster than I anticipate driving.

Is Grad C a unit?

(http://www.myaudis4.com/tuning_images/vmax.png)

Grad C is degrees in celsius.

Where van i get such Nice defined M xdf? Would you share?

if you have the mbox xdf from here you just need to change how you view the file.
to paramater category and itll show it in the forum of the clean folders

I would love to learn the process! I am a mechanic and fully understand the basics of engine operation, but I have never played with software and just need guidance on how a certain setting not only changes parameters of the aspect it pertains to, but how it references other associated paramters to come up with a final value.
Say ignition timing for instance. In a distributor style ignition system you really can only adjust base ignition timing and the advance curve. In my 2001 GTI with a 1.8t I believe that timing not only changes due to engine speed, but boost level, driver requested load value, coolant temperature, ambient temperature, knock detection, etc., and is capable of varying individual cylinder timing if I'm not mistaken.

Basically I would love for someone to guide me through what changes affect what, and how it affects everything around it so I can grasp the whole big picture of how the Motronic system manages an engine.

How fair would you suggest to run the timing on stage 1? An commercial tune i had on my S4 when i 1st had it logged 9deg at 5860rpm with .8bar i dont have the file or ever did but have the VCDS logs.

The map was ok produced 305bhp

Went really well but havning not drivern a standard S4 and them jumping to Stage 2 i cant compare.

Im happy with a stage 1 map for S4 but i need AFR min so i can flash it and log it would prefer knock detection but dont have the funds for that yet.

Thanks, odd units for that axis.
 

That is Nye's XDF
http://nyet.org/cars/files/defs/ (http://nyet.org/cars/files/defs/).

Awesome project :)


Is there any recommended reading on this? I understand the principals of IC engines, what all the bits do and how they work, but wouldnt have a clue to know where timing should be advanced or retarded, or how to work out what lambda values I should be aiming for.

Presently I am logging the following variables with ME7Logger:

nmot (RPM)
gangi (GEAR)
wdkba (ThrottlePlateAngle)     

pus_w (BaroPressure)               
plsol_w (BoostPressureDesired)
pvdkds_w (BoostPressureActual)
ldtvm    (BoostFrequencyValveDutyCycle)
mshfm_w (MassAirFlow)
tans   (IntakeAirTemperature)
zwout (IgnitionAngle)
 
dwkrz_0  {IgnitionRetardCyl1}
dwkrz_3  {IgnitionRetardCyl2}   
dwkrz_2  {IgnitionRetardCyl3}           
dwkrz_5  {IgnitionRetardCyl4} 
dwkrz_1  {IgnitionRetardCyl5}                     
dwkrz_4  {IgnitionRetardCyl6}                     

rkrn_w_0  {KnockVoltageCyl1}       
rkrn_w_3  {KnockVoltageCyl2}                   
rkrn_w_2  {KnockVoltageCyl3}                                         
rkrn_w_5  {KnockVoltageCyl4}   
rkrn_w_1  {KnockVoltageCyl5} 
rkrn_w_4  {KnockVoltageCyl6}   

vfil_w  (SPEED)

fr_w {LambdaControl}   
lamsbg_w (AirFuelRatioDesired)
ti_b1  (InjectionTime)
tats_w {EGTSensor} 
tats2_w {EGTSensorBank2}       
tabgm {EGTModelBeforeCat}           
tabgm2  {EGTModelBeforeCatBank2}

usvk  {O2SVoltageSensor1}         
usvk2 {O2SVoltageSensor1Bank2}
rl_w  (EngineLoad)
fzabgs_w  (CountMisfireTotal)

here is my list (semicolons are comments)

https://raw.githubusercontent.com/nyetwurk/ME7L/master/logs/typical.lst

Lots of information out there, but this document does a great job at getting some of the very important fundamentals across (attached).

My advice is to read as much as you can get your hands on... If you're the type that learns better hands on, get friendly with some local shops/engine builders or consider a course for ICE or even tuning (beginner courses usually touch on ICE basics).

Loving this thread already... since this is all info I'll be looking for after I sort out the problems with my allroad.

One thing - the XDF on here for mbox is missing a lot of the variables that are listed on the wiki.  It would help to have an updated XDF with all of the parameters from the wiki defined.

Another thing - should include some clear directions for disabling emissions equipment... there is a lot of information on the forum and wiki, but it's tough for a newb to follow

A lot of basic info but was a good read figured id share https://drive.google.com/file/d/0B9AlTN-WfgmnN2huYjhaRnJFcGs/edit?usp=sharing

Here is my contribution. These files are stock from a calibration standpoint. The Bosch files have obviously been converted to Bosch spec, and have had cw_can_r set so as to allow coding for non-esp cars without DTC's. Please look at the xdf, and realize that I understand some of the terminology is wrong. I went for universally understandable as opposed to black letter. Additionally, emissions and codewords are switches now, just click one button and they are set appropriately. No more explaining 1's and 0's. All units have been converted to mph, LDRXN has been converted to psi.

Thoughts?

Very cool ^ So when you hit the switch for cat efficiency it sets the value to whatever the mask is in the XDF perameter to bit 0 or 0x01?

I'm not convinced making ldrxn psi is a good idea...

I'm tempted to say you can have both a PSI version and a load version of the map in the xdf, but I guess that would confusing (even though you could use it to illustrate there is a relationship).

In my map packs, I doubled up all the lambda tables into lambda and AFR. Also, i have a krkte double that is cc/min... and KUMSRL (uh i think thats the map?) that is displacement.

Or maybe this is the wrong time to bring it up :/ I'm a bad teacher so ignore me at will.

I know, technically everyone should read the FR and understand how ME7 converts load request to boost, but that's beyond the scope of what we're trying (at least me) to accomplish here. Once people see appreciable results from their OWN work, I think that'll create the drive to dig deeper. I feel as though if we keep the redundancy to a minimum, it'll prevent a lot of questions later.

With that being said, I'm just approaching this based on the questions I get most. I would love to hear from some of the new members, fresh off the boat. I figure I'll add map names in parenthesis in the xdf descriptions for the stage2 xdf, that way the natural progression steers people in the right direction.

HOWEVER:

These are just my thoughts. I don't want to monopolize this, or trivialize anyone's ideas. That's not what I'm about :P

This is correct. Same with all of the other switches. Try it out.

As a kickoff I think the XDF is excellent.

I could live with LDRXN not being in PSI, but that's only because I do a separate calculation on a spreadsheet to get an estimate for what PSI would be.  For now leaving it as PSI is probably the way to go, erring on the side of simplicity is my preference, it's easier to introduce advanced concepts after a basic understanding has been established rather than getting bogged down trying to explain things that could have been avoided in the first place.

For those following along at home, should we have set the Rev limiter and Speed Limiter at this point?  Looking ahead, next we will...

I put those in the xdf :)

I think having both versions in the xdf (load and psi values for LDRXN for example) and so on would be great to help people learn on their own the relationship between the two. Once they get comfortable they will most likely move over to using the tables/maps correctly (load values for LDRXN). Just took a look at the xdf...Might also be good to add the ME7 acronym along with it's English description for the maps, and adding axis descriptions for the ones that aren't obvious. Great job by the way!

That is all stage2 territory in my mind.

Alright, I guess the next step is, who is familiar with me7logger, and wants to guinea pig the stage1 and log it?

Question, why does a stock Mbox 2.7 map run all 1's for LAMFA. When compared to some "custom" tunes that have modified LAMFA table? Is either/or being run in closed or open loop?

I'll be glad to log it.  I posted the variables I normally log previously thinking I might get a suggestion about what else to log.  I don't want to log a ton of variables just because I can, I like to limit what I log to what I expect to use, if I need to create a new config file for troubleshooting I'll do that.

About the Stage 1 tune, at this point we have a stock file so logging in the current state would not be necessary, right?  I saw the comments in the flyover box about making changes but would be interested in getting some suggestions or discussing a little more.  Based upon the notes does the "Desired Air/Fuel Ratio" table remain uniform throughout?  Presently all table values are 14.7 - I'd have thought tailoring that might be part of a Stage 1 tune.  Is it not due to the range of the table, just 50-100 pedal position and rpm basically 3000-6000?  Is the table interpolated between the 1000 and 3000 rows?

I think it would be great to get some stock logs first..log the same variables in the same order for a good comparison between tunes.

Agreed. 100 percent stock logs would be wonderful. I can't recall ever seeing any of a healthy car.

Then, I will wrap up the .cfg everyone will be using for stage1, and we'll get a log.

Good work in here.  daz, I like the xdf.  Very good descriptions and advice in the details of the maps!

I'd like to participate, but have an ME 7.1.1 allroad and that might throw too many variables into the mix.  that and I have a damned apr chip soldered to my ecu. 

So for the stock logs..What gear(s)..what variables?

Heh, I spent quite a bit of time on the stock files and xdf today. Tomorrow is the day I create the logging configuration files :)

I would say start with the markp variables, then neuter the list a bit to remove the confusing stuff.

3rd gear from 2500 to readline, WOT

my list is here

https://raw.githubusercontent.com/nyetwurk/ME7L/master/logs/typical.lst

semicolons are comments.

Here's a set of stock tune results.

Stock Tune Baseline (http://www.myaudis4.com/2014/09/05/s4-stock-tune-baseline-logs/)

Just goes to show how dead on the stock tune is :)

Look at that compressor map overlay!

BTW for the purposes of the training guide, we should have a step by step how to for using ME7Logger with ECUxPlot ..

I love how you have compressor outlet pressure btw.. is that real or calc?

I plugged a pressure sensor into the compressor signal line so it is actual pressure.

I've applied the suggested changes for a 93 octane gas.

(http://www.myaudis4.com/tuning_images/lamfa.png)

Very nice.
I think it would be usefull to right down real name of the maps as well.
This is LAMFA and ??

My suggestion for the axis values is in the description, but I go with:

50 60 70 80 90 97.5

As for the actual AFR, I figured something a little richer to account for 91.

Results of the implementation of step 1.

(http://www.myaudis4.com/wp-content/uploads/nef1_afr.png)

Remainder of the charts are here:

Nefmoto Tune Rev1 (http://www.myaudis4.com/2014/09/06/nefmoto-tune-rev1/)

Hey Jeff, can you post full csv's for us?

Sure, here's the interesting part of the drive.

http://www.myaudis4.com/tuning_images/me7log_20140906.csv

Can someone explain how the pedal map for desired air fuel ratio works?

VERY roughly: KFPED generates mrfa, which is the torque axis of LAMFA.

The result goes through a low pass filter

In any case, hopefully you have read this first before reading this thread :)

http://s4wiki.com/wiki/Tuning#LAMFAW

Thank, Nyet!

So in Flyboy's example, it looks like the LAMFA axis hasn't been changed.  ??? Or is the pedal position map the LAMFA axis?

(http://www.myaudis4.com/tuning_images/lamfa.png)
[/quote]

Yep, it looks wrong still.

 :)

Is timing change 100% nessasery for stage 1?

Ive seen 2 "pro" maps one for S4 B5 were it had KFMIRL upped and 1.4bar in LDRXN but timing was 100% stock, the other was 1.8T octavia 180bhp same but 1.2 bar (0.7bar standard)

My opinion would be to remove and negitive timing as a starting point and test/log

this is awesome guys still need guinea pigs for logging?
id be happy to contribute.
and having an xdf file with load and one with psi seems like a great thing for comparison

Suggestion for making it correct in your view?


Absolutely, I don't have a monopoly on this, in fact the more people that will participate the better. 

http://s4wiki.com/wiki/Tuning#LAMFAW
http://s4wiki.com/wiki/Tuning#cite_note-13

Will the emissions related addresses be the same with the 551T? such as CDKAT?

Very unlikely. Also, the beginner won't be worrying about that. For the purposes of this project, assume everyone is using the M or L-box. If you want to use a different file, I suggest defining it manually and starting your own thread :P You will get help there.

Jeff:

My advice for LAMFA is in the comments.

Rescale the axis to 50 60 70 80 90 97.5

Ok, when I looked at the axis values I was thinking the 0.50 0.60 etc was the way the percent value was written in the table, which is why the last value changed from 1.0 to 0.97  I'll make the update.

Awesome. As you have it, any torque request at all (over 1 percent) will result in an AFR request of 12.2-lol

The S4 wiki tuning page explains that this is wrong and there are links to the thread here which breaks it down in detail.

PLEASE let me know if the tuning page is missing something; I really want to make sure it is complete with respect to this tutorial.

If you didn't see it, that is a bit distressing...

I'm not following the S4 Wiki tuning page, at least not unless I am directed to look at something specific.  I have seen the relevant section but was not certain of the applicability to the thread's XDF. 

Here's the corrected settings:

(http://www.myaudis4.com/tuning_images/corrected_afr.PNG)

I've successfully used it! After a little time with winOLS with comparing a defined 551M ecu it actually was pretty easy. Added KFZW's axis, IOP, IRL corrected addresses for 551T.  I cannot post any logs at this time because A.) I can't get the visual logger to connect and B.) the throttle hose keeps blowing off.   ::) When I am finished debugging the logger/hose issue I'll post them up!

Looks good. I'd leave the last entry 100.0 to prevent confusion.

Also, the XDF should probably explicitly call that table the axis... maybe something like

"Pedal position axis for the Desired Air/Fuel Ratio table" so it is clear it is related to the "Desired Air/Fuel Ratio" table.. I wish TP let you change axis data directly :/

Download the xdf nye, look at the comments. There is a paragraph for each map explaining everything :)

No, I saw it, just saying maybe it would be better if it specifically said "axis" in the table name itself..

Also, as to your question (from that other thread) if we should use BTS instead... I think that is actually more complicated. Too much stuff has to change.

Also, this way introduces the idea of changing "axis" data in gentle way (I hope).

My 2 cents of course..

This was another reason why I suggested adding the true map name along with the English translation. There is no way to directly relate to the wiki.

+1

I agree. I was hoping to use this thread to put a little more focus on specific maps and what they do. I am on Me7.1.1 so my xdf is still in German which makes it that much harder. I find I am still having to jump from 10 different threads to feel like I am getting somewhere.

Honestly LAMFA is probably the most straight forward fueling map to use for beginners. You will also have BTS if the user is on stock ish exhaust system as a safety. In order to get BTS to work efficiently it would take the user some time with playing with the EGT temp trigger and dialing it in.

Alright. Once Jeff rescales the LAMFA axis, and posts a log (preferably with light throttle/closed loop leading to WOT) We'll call fueling good. Then neuter timing a bit for the next revision? What say you? I'm tempted to go right to boost, but on 91 octane, even stock timing at peak load can get messy.

I think we can go straight to boost.

Are the changes I posted most recently correct?  Saw the other discussion where possibly making the change with a hex editor would be needed.

Making the changes in the axis map in the xdf is sufficient. After you do that, if you open LAMFA, the axis should reflect your changes.

Changes incorporated and logged the car again.  To keep the clutter down in this thread I'll provide a link to the charts I made.

Nefmoto Revised AFR (http://www.myaudis4.com/2014/09/10/nefmoto-revised-afr/)

Excellent. Can you also post a csv so I can filter it for the official version?

Here's the ME7Logger csv, pared down slightly.  AFR is contained in a separate file that I have not uploaded.

http://www.myaudis4.com/tuning_images/me7log_20140910.csv (http://www.myaudis4.com/tuning_images/me7log_20140910.csv)

Here is a proposed boost profile, peaking just under 14 psi at 2500 rpm, and compared to the stock profile.

(http://www.myaudis4.com/tuning_images/ldrxn.PNG)

Jeff, don't try to increase the initial rampup. Stick with the stock datapoints until 1850.

My opinion of course.

Elevated but more closely mirroring the stock profile:

(http://www.myaudis4.com/tuning_images/ldrxn2.png)

What Daz has done to the xdf file I'm happy to see. This is more or less what I was looking for when I 1st came onto the forums. I do agree some things should be left for a stage 2 file builder. But from what I see so far this makes me feel more comfortable coming from other tuning software like OpenECU. Thanks Daz and others, wish I could find a way to help other then a tester and see how a start from scratch file comes out.

I will be following. I of course will try this out and still use the other 2 xdf's needed for my files. But glad to see this progress. A clean-up later maybe should be taken away from discussion and a new thread started, I think this one is needed as it helps the knowledgeable communicate with the less informed.

**After getting some closer looks, I think some smaller/simple wording should be included in the descriptions just tho's who are 1st starting off, I'm stunned how much easier this xdf makes the approach to tuning able to be done.... I know a lot of my confusion was based on axis and means... to me afr of 1 didn't mean much, coming from other tuning software that dealt in idk what it would be called but just xx.xx so that already helps me never mind someone new to tuning period. I wish I knew more how to use TunerPro more to help with the XDF building but I'm lost... anyone got a PDF of the instruction manual or something??

Another vote for maintaining the original FR names in some form.  This will help make the information useful for someone using an XDF based on FR names instead of one that has been pre-defined with user-friendly names.  Selfish request, really (I'm tuning an allroad 'S' box), but it's been wonderful being able to take the B5 S4 2.7T info on this site and apply it to my own ECU.

Updates?  :) Also how does using LAMFA/KFLBTS differ in terms of fuel economy? I'm currently running a version very similar to Flyboy's fueling map, and proposed changed to LDRXN, and I seem to be getting worse mileage that other tunes I've run with a lot more aggressive maps..This or I'm just beating on the car more..hehe

Did you fix the axis as discussed?

Long time ago, but you might have me mixed up with Flyboi. hehe

As discussed in this thread I mean. You have rescaled it? If so, fuel economy during part throttle (closed loop) operation will remain unchanged on a stage1/2 file.

When you look at Lamfa table Changes, Do you understand Changes that has been made? That should answer your question.
GL!!

I do understand, but because of the lack of resolution using LAMFA, and simple approach for this project, I assume there was a trade off of some sort.

Back to logging the car.  This is with the first modified boost profile.

http://www.myaudis4.com/2014/09/22/nef-stage1-modified-boost/

How does the car feel? Thanks for the logs  :)

Please post raw csvs! At some point we need to add graphing instructions as well.

I'll work on posting the csv after I pare it down a bit.  Updated the post with a pic of timing, looks like that'll need some attention.


Not bad, but the timing drop is noticeable.

Here's the csv file data.

http://www.myaudis4.com/tuning_images/me7log_20140922_pared.csv

Is that because of timing being pulled? How does one figure to lower boost, or lower timing in this situation?

Or add fuel. Tough call. The only real way to know is a dyno.

You can try to measure acceleration with road logs, but you have to be scrupulous about keeping variables to a minimum.

Timing drop?

If you are talking about the curve from 3600ish it looks like the switch from KFZW2 back to KFZW.

Some of what you may be feeling is also cam changeover.

You always log the CFs as well

Of course. But this still does not point to whether the timing pull should be resolved with less boost, less timing, or more fuel.

No it doesn't, but obviously it tells us if timing has been pulled.
One step forward :)

I added a couple more charts.  Suggestions on the way forward or is this good enough for Stage 1?

http://www.myaudis4.com/2014/09/22/nef-stage1-modified-boost/ (http://www.myaudis4.com/2014/09/22/nef-stage1-modified-boost/)

I say we pull some timing and smooth the curve :)

Then I'd say we call it good!

I will post some screenshots tomorrow.

Here's camshaft angle overlayed with the timing curve.  Rest of the charts and csv file are in this folder:

http://www.myaudis4.com/tuning_images/k03_me7x_nef3_files/ (http://www.myaudis4.com/tuning_images/k03_me7x_nef3_files/)

(http://www.myaudis4.com/tuning_images/k03_me7x_nef3_files/nef3_camshaft_angle.png)

Here's what I'd like to see Jeff.

Camshaft angle, ignition retard, and load on one graph (put camshaft angle and load on the same axis, CF on the other)

Or, just post the raw csv and I will plot it myself :P

...

Here is how I view logs for timing work:

(http://i.imgur.com/AIPbwwg.png)

If the camshaft angle is 0, that means KFZW is being used. If it is 22, that means we're in KFZW2.

In this log, we can see that from 3800 onward the values in KFZW are being used. Prior to that, it is KFZW2.

We can also see two knock events. One at 2200rpm, 150ish load. One at 2500, 160ish load. Take a look at the values for in KFZW where 2500rpm and 150 load intersect. Reduce timing in that area. NOTE! Just because CF is 3, does not mean you have to pull 3 degrees! One knock event at that RPM means CF's will be 3. This is actually not bad, and can be left as is, but I wanted to post for illustration purposes.

I would then suggest making sure KFZW/KFZW2 have the same values around 140-160 load at 3500-4000rpm to reduce the notch in timing. After that, relog. If you are not showing any CF's, I would then suggest adding more timing up top, from 4500 onward (to KFZW).

Great explanation. I would like to see overall ign timing on that graph as well if possible. Would you suggest removing timing by say half of what you see in correction? Also is it good to remove timing a little earlier than when you actually see the correction? Might be nice to have LDRXN expressed in both load and psi in the boost folder of the xdf so we can relate to KFZW/2 better.

Like I said, remove as little as possible. As KFZW is 8 bit, the increments are .75 (I said this in the post). As for ignition angle, that is a separate graph. The point of this is to keep it simple and do one thing at a time.

If you feel you have a better way, by all means, post an illustration! This is supposed to be a collaboration after all.

Only thing I have to contribute right now is a slightly modified xdf description-wise.  :)

Thanx gays this is by far most giving explanations i have seen.
Question.
Can we say that 2.7T Engines Always use in BOTH cams KFZW2 at low rpms, and KFZW both cams in high rpms ( meaning after swichover)???
If we lower or bump up timing for ie .75 deg, is the effect lowering or bumping up LOAD in those areas? Meaning getting more TQ in this areas??

wow, I was under the impression that the kfzw tables worked opposite of that.    ::)

I thought the switch over caused the tables to go from kfzw to kfzw2.  Good to know. 

Do that by adjusting one downward or the other upward?  Down seems like the safe route but almost certain to need additional modification back upward.

Depends. If you have no timing pull after changeover, then yes, increase KFZW. Lower KFZW2 if it is knocking prior to the switch.

Me too  :P

I have a question as to why the adjustment would be made in KFZW and not KFZW2 where the knock event was logged based on this statement[Take a look at the values for in KFZW where 2500rpm and 150 load intersect]. Reduce timing in that area

KFNW uses rlsol_w, so when you go WOT the cam advances right away and begins retarding just after 3600 RPM.

In the graph that DD posted, KFZW2 is used where you see cam state at 22*, KFZW is used where it is at 0*, and they are interpolated while the state is changing. fnwue is the percentage of KFZW2.


That was an accident I'm sure. It would be KFZW2, but I wouldn't touch it as it's two cylinders with one knock event each.

I would also advance KFZW during/after changeover.

Made me confused as well.
Would you correct this guys??

What are you confused about?

What is the knock event indicator?

Edit: And to stay with the naming convention: KFZW = Ignition Timing, KFZW2 = Ignition Timing (VVT Active)

B_kl, but you won't catch it 99% of the time because you aren't logging crank sychroneously.

Or were you asking how I decided it was only one knock event on the two cylinders?

I was wondering how it was determined from the chart on the previous page, containing this explanation.

Ignition is retarded by krfkw from KRFKN for each knock occurance. KRFKN ranges from 1.5-3* based on RPM.

In this example ignition was retarded by 2.25* then began advancing back to 0* retard. If there were more knock events, then ignition angle would have been retarded again/further.

I agree. If I were tuning this I would not alter timing for knock on 2 cylinders. That, to me, is ideal. It shows you're at the very limits of what is possible on the setup without surpassing them. I just explained how to do it because not everyone will be in as good of shape after increasing boost.

I still can't fathom how everyone was confusing KFZW and KFZW2.

I am very glad to see you here phila. Drop some nuggets of wisdom. You've see where we started, what, in your opinion, should we be doing differently (if anything) and next? I don't want to monopolize this thread.

This is the stock tune with 93 octane gas.

(http://www.myaudis4.com/wp-content/uploads/image004.png)

Not really comparable.

Generally lambda 1 and much lower load.

I thought it was interesting to see comparable ignition retard.

After making the changes the charts are shown here:

http://www.myaudis4.com/2014/09/24/nef4-timing-adjustment/

and image files with the csv file in this folder:

http://www.myaudis4.com/tuning_images/nef4/

About writing mistake DD had, that was also pointed out above.
This is very clear to me, I only asked for misspelling being corrected BC of the nature of this thread

I also had a question before that, just after DD posted this graph. Would you be kind and straight it out for me?

Interesting what the desired boost profile results in:

(http://www.myaudis4.com/tuning_images/ldrxn2.png)

(http://www.myaudis4.com/tuning_images/nef4/nef4_boost.png)

HBN limit maybe?

http://www.myaudis4.com/tuning_images/nef4/me7log_20140924.csv

No idea what HBN is.

KFLDHBN - max pressure ratio

Gays?

What do you mean by both cams?

B1 and B2? Or intake and exhaust?

Both banks are adjusted together. Only the intake cams are adjustable.

Advancing timing towards MBT increases torque.

I think the stock values are around 2.3 tapering to 1.8 (17 to 11 psi relative), so that makes no sense anyway.

Something else is amiss.

Oj, sorry misspelling.

Yes, I mean B1 and B2.
I am just trying to understand this 100%, that's why stupid questions :)

You are assuming ambient pressure.

Easy solution is to log the variables.

Has KFTARX been modified?

ambient is around 998mbar


Looks stock, spec load peak is 164. corr spec is 159-161 at 2500 rpm

Still should be requesting more than 11psi


http://www.myaudis4.com/tuning_images/nef4/me7log_20140924.csv

I'll hand the development baton off to someone better suited to the job.

Oops, guess I was the one assuming.

Just add fldrrx_w, rlmxko_w, and ldrlts_w to the log and the intervention should be revealed.

I am digging this thread. I know it seems elementary, but I think it's going to reduce the number of "zomg! what do I do!" threads significantly.

No problem, we can do that, but we still need your logs! Can you add the variables that phila asked for?

man im so happy this thread is moving like this ive been so damn busy at work. ive edited a stock file from the first page trying to follow up with what were doing and i really help i can post logs this weekend to try and help out and add more data

Guys, it'd be best to have a larger sample group. It's not fair to Jeff to be the only one logging! If you have a relatively stock car, feel free to post!

As soon as my car has suspension again, I'll post up some logs for that.

Thank you! Can you make sure you log the vars phila requested as well?

And I'd also like to echo what daz has said, and thank Flyboy for getting us this far. He's done a great job logging and posting information thus far.

I've gotta get a new air suspension solenoid block for mine... then I'll be happy to join in the fun.  Other than EGT/SAI delete and numbing the pre-cat test... I'm stock

SAI? this thread will likely only contain Mbox (6sp) files, so you'll have to do the map transfer yourself if you are tip...

Is this a second gear run?

Not a big deal, using 551L file

Hey guys,

I would be interested in being a tester for this project, but with a couple of concerns....  My s4 is my daily driver so I can't risk being down and I can risk having something pop due to the tune.

I have a 2001 s4 with tirptronic, as far as I know it is bone stock. 
I skimmed through the entire post, and I have been a pc tech for about 20 years so I am very comfortable with cmdline, and I do have an older vcds cable (serial, I haven't upgraded to the USB one yet for my 2010 vw).

Would it be possible to have someone write up some direction for someone who has never used this particular software?

Bill

there already is alot of information on that bill
this is for logging data. http://www.audizine.com/forum/showthread.php/479416-****-Complete-Guide-on-How-to-Take-and-Graph-Logs-****

so id be a willing participant with logging abilities ....but i have a tip if some one wants to make an lbox file id be happy to do the the tip support

L-Box and M-Box addresses are the same, you can apply the same changes to a stock L-Box file

Ok, the first thing to do is figure out why the suggested LDRXN does not result in the proper boost request...

So if you could start with that, and add logging the variables phila suggested in an earlier post, we can get to the next step.

Thanks!

I've had the idea to contribute something like this for a while, and have never made time to do it.  Props to everyone contributing and making it happen..

I've also got a new allroad, stock other than catback, that I can possibly help with some logging graphs if I get a minute (and if I am able to do some work and determine that the K03's are still good, otherwise it will go big turbo right away).

BTW, I haven't checked the xdf's myself and am only going off of the screenshots, but if not already, I agree with the previous suggestions that having the real names for the maps in addition to the easier descriptors will help ease the transition for people to go from 'stage 1' into future self-directed research and tuning (and it would help people who already know the map names by knowing what they are actually changing).

I would be able to help for the Stage 2 file. I have a lot of open areas to log. In about 2-3 weeks Florida weather will break and the cool less humid air will come.

If you think my mods are fine for the stage 1 I'll help.
2000 A6 2.7T
- Custom Intake
- Push on Coils w/BRK7s
- XSpower Diverter Valves
- XSpower 3" Catless flex DP's to stock middle section and muffler, rear mufflers removed at axles
- XSpower SMIC's
- Upgraded TBB
- Daz Stage 1 TCU Flash
** Full TB job ** cars healthy and strong
~ Removed ~ SAI & REAR 02's -

I'll be watching the thread so lmk when you need my hardware setup for testing.

ok so i re read all the posted and caught the part about pedal positon for desired afr i changed things around to match that and ive displayed my other changes this will get loaded tonight any feed back before i load and logg?
also* i corrected my logger issue i moved somthing from a folder and it screwed the whole thing up

S4addict:

That looks reasonable. Log and report back.

so i adjusted afr and boost didnt get into timing cause didnt think it was discussed in this thread enough so thats stock .....ran it through me7check (easy check). and found 4 errors . what did i do wrong ? attached bin below......going to log if i can get this sorted....for the tip logs

edit....i guess i need to use me7 sum to fix the errors? uhhhhg this may take awhile i hate command line

Command line is like bacon. Everything and anything tastes better with it. ME7sum is so easy a fast to use that even if You've never used cmd, You would be able to run ME7sum in two minutes.

If You already know how to use me7check, me7sum is basically the same.

There's easy way to run a command line process with minimal typing, just drag and drop files

Open up a command prompt by pressing the windows key, type cmd, press enter.

If you want to use ME7Sum to correct checksums follow these steps:

1) drag the ME7Sum exe over the command prompt window.
2) select the command prompt window and press the spacebar
3) drag your bin file over the command prompt window.
4) select the command prompt window and press the spacebar
5) drag your bin file over the command prompt window again.  Rename that file path parameter in the command prompt if you wish to not overwrite your non-checked file
6) Press enter

If you want to use ME7Sum to validate your file, follow steps 1-3 and 6.
The same would apply to ME7Check, 1-3 and 6, but drag ME7Check in the first step instead of ME7Sum

Thanks all.. a big motivating factor for me dusting off ME7Sum development is that we need a free workflow from start to finish for these community projects... telling noobs to buy (or worse, pirate) software is a non-starter.

I appreciate all the time and effort others have put in!

if i could kiss you i would, hahahaha worked like a charm........


thanks nye for the dusting   :)  i'll get the tip logs soon

Very commendable

i guess i have a boost leak somewhere ill get that figured out...... :'(    but here are some tip logs 14 psi for boost and afr figured like talked about in this thread... timing stock  ......im getting massive timing pull
ill get this boost leak figured and post new one ......do you guys think the timing pull is from the boost leak?

What is with your throttle position?

i may have rolled on it at first ....that was my fault i was just checking that and kicking myself......
not ssure if this matters but this pull was right after i installed the ecu and did a tba

when logging it is super important to be consistent.

Also, this is why I really dislike tuning tip cars (kick down etc.) and i was hoping to keep this thread restricted to 6sp ....

I have a TCU program I use that restricts downshifts when in tip mode. Makes it more akin to a 6 speed.

Won't help here though.

my car wont down shift unless i slam it the extra inch...pretty sure theres a kick down switch ....its pretty easy to keep in third.... i wont clutter this thread just trying to help as it seems to have lost steam

No we appreciate your help, but if the car isn't 100% mechanically good (and that boost leak is really killing you), then yea, it is kinda cluttering :(

To get this back on track:

I don't think this is a HBN problem..

My thoughts:

* 160 req load with very low IATs will have a pretty low boost req - we should consider maybe requesting more load
* I don't think we should express LDRXN in boost, since the load->boost conversion is very IAT dependent (although we could counteract it in the spec load -> cor spec load path with KFTARX)
* req boost isn't tapering because of IAT rising from stock ICs... that is pretty much canceling out the spec load taper

I'm interested in this, is there a thread on it somewhere?

no her s4 is a tip tragic but its cool its lowered n shit
ok heres my file from my 2001 audi s4 six speed 4th gear. really shoulda done a diffrent gear but i can post more logs.
mods 3in to 2.5 ssac down pipes srm bipipes for meth with meth -meth turned off for logs-
stupid cold air intake and a true dual miltek exhaust.

Something is wrong. Likely a boost leak.

I'm rockin a TIP with a mod from Daz, wish I got the kick-down turned off lol... makes the car more like a manual plus logs are hard to get like at 3000k or below, you gotta roll on it low part throttle and then do WOT like my logs show.

Gonna suck for dyno guy. O well he don't get it correct they should have to do another run to make up for it. My last car then stopped dyno at 6500. My turbo was pte6152e didn't make 20psi to 4500, and the cams where ready to go already hks272's but I had it tuned to 7500rpms. I felt I got lower reading on dyno and seen and felt more like 370awhp not 345awhp on a 4g63

Here is the present tune relogged to include the three additional variables, though I do not show those new values in the charts:

http://www.myaudis4.com/tuning_images/nef4/relogged/charts.html (http://www.myaudis4.com/tuning_images/nef4/relogged/charts.html)

In this folder are all of the chart images as well as the CSV file of the data.  I have shown two 3rd gear pulls on the charts but within the CSV there is also a 2nd gear pull.

http://www.myaudis4.com/tuning_images/nef4/relogged/

so were supposed to requesting 14psi correct?

My understanding of the requested boost table (LDRXN) is that it approximates what the requested boost will be, but is not a direct determinant.  The boost curve that I log should roughly follow the shape of what is planned, but the magnitude of the boost value may not match what is in the table. 

What I'm finding odd about the resulting boost profile is how it climbs where the 'requested' values in the table are decreasing.  That's the part where the other factors come into play, and illustrates how the table is not the sole driver of the resultant boost profile.

Here's a chart with the new variables:

(http://www.myaudis4.com/wp-content/uploads/nef4_added_variables.png)

Yep. Nye and I spent quite some time talking about how to best approach this.

We decided to just try increasing LDRXN by an arbitrary amount and ditching the PSI conversion.

Care to give that a shot Jeff?

Sure, here's the revised profile.

(http://www.myaudis4.com/tuning_images/nef5/nef5_boost.png)

Do you think timing is ok or should I try refining it some more?

Care to explain what those extra variable I logged are telling us?

Yes, I've just increased the values (black line) slightly versus the old (red line) values and will see how the boost changes.

I'm also going to slightly lower timing around 2000-2500 rpm and increase a little around 4000-5000.  All timing changes will be on the order of 1-2 degrees.

Updated results:

http://www.myaudis4.com/tuning_images/nef5/charts.html

Nef4 (previous - yellow) vs Nef5 (current - blue) Boost

(http://www.myaudis4.com/tuning_images/nef5/nef4_vs_nef5_boost.png)

Here's the timing change progression plotted out.

(http://www.myaudis4.com/tuning_images/nef5/nef5_kfzw2.png)


(http://www.myaudis4.com/tuning_images/nef5/nef5_kfzw.png)

Please post logs, and ECUxPlot graphs if possible..

IDC is maxed, so I think we are good there. Don't think we can do much more boost on stock injector/fpr.

Timing looks good.

Don't know about the rest until i can look at the logs themselves, in particular, PID stuff.

Added the CSV to this folder:

http://www.myaudis4.com/tuning_images/nef5/

This has the charts:

http://www.myaudis4.com/tuning_images/nef5/charts.html

IDC is <85% @ 5500 rpm

Whatever you are using to process the csvs... please stop. You padded it with commas.

Even worse the first line is


which makes automated parsing of that csv impossible.

PLEASE at least try to open it with ECUxPlot (to verify your munging didn't mess anything up), since that is what others will be using to visualize the data, not hand cutting/pasting into excel, or similar silliness.

Also, at this point, we are going to need the rest of the PID variables, since we are underboosting...

This illustrates our load/boost issues... phila, can you comment?

Also, maybe need to log ldrlms_w, fupsrl_w, and pirg_w... I don't have locations for any of those :(

TIA :)

Note that ChargeLimit etc. doesn't have units... probably need to hand edit the ecu file to get it to report mbar... either that or the log is kinda screwed up so ECUxPlot is having problems extracting the units properly. Need the original .csv to make sure.

KFTARX and KFLDHBN

fldrrx_w is 1 throughout, I imagine coolant temp remains below 117* c, so all rlmxko_w intervention is from KFTARX.

There is also load intervention via ldrlts_w (KFLDHBN).

rlsol_w does need to be added to the config file though.

Relogged with the additional variable.  CSV in this folder, all 4 3rd gear pulls are relatively short.

http://www.myaudis4.com/tuning_images/nef6/

Charts of a few variables

http://www.myaudis4.com/tuning_images/nef6/charts.html

Made a couple minor timing adjustments that I will capture and add to this post later.

Awesome! Thanks Jeff!

Can you run the next set of logs out to redline? We need to see how close we are to injector max..

Here's the latest stuff including full csv.  There are a couple short third gear and at least one second gear pull.

http://www.myaudis4.com/tuning_images/nef7/

Here's some charts:

http://www.myaudis4.com/tuning_images/nef7/charts.html

The timing changes.  On the left side are the original tables and on the right are the current values.  For KFZW2 I lowered the values slightly to try and eliminate the ignition retard that was occurring around 2000-2500 rpm.  For KFZW I raised some of the timing values to try and have a smoother transition around 3500-4000.

(http://www.myaudis4.com/tuning_images/nef7/nef7_timing_before_after.jpg)

Acceleration comparison

(http://www.myaudis4.com/wp-content/uploads/nef7_acceleration.png)

Does anyone know what boost the APR tune pushes? ::)

Too bad there is nobody here with logging software that can log an APR car :P

Generally 15-16psi, IIRC.

You can log APR.

I believe this is as far as the Stage 1 tune will go.  The thought was to start a Stage 2 tune discussion and use that thread to address some of the more advanced topics that surfaced during the Stage 1 tune.

I'm swapping to different turbo's so I won't be logging the Stage 2 changes.

Uh. 11 psi vs 16?

I don't think we should stop here for stage 1.

IMO we need to massage HBN, bring up peak req boost to 15 or so, possibly iron out the VVT load->boost wrinkles (i'm sure purists will cringe, but IMO I don't gaf), work on fueling, then timing.

Here's a log from my allroad on APR 91 oct tune for comparison.

hmmmm i thought we hit 14 psi my mistake, ill have to take another look.......i agree with you otherwise bring load up notch

we were doing 14 psi or atleast mine was.

If we start breaking this down, is HBN an issue at this point? In Flyboy's last log (me7log_20141016_214741) I don't see HBN intervening. The corrected load follows rlmxko so it's most likely the KFTARX map, correct? Do we want to modify that map to at least keep our requested boost at lower temperatures? We should probably log frxta_w to see what it's doing. Even at stock boost levels I've had this cause throttle cut on cold days.

Agreed on all points, after reviewing that log. I think before messing with LDRXN we really should consider flattening all the KFP maps 3krpm and up... makes tuning LDRXN much more sane.

The requested load (rlsol_w) follows the corrected load (rlmax_w). It's included in the log I posted if you want to take a closer look. The math works out for the IAT correction factor.

I've noticed a number of misfires at cylinder 5 in Jeff's latest logs (up to 11).

Are these normal?

The misfire is not unique to the Nef tune.  I think it's caused by a poor fit on a phenolic spacer that I'm trying out.  Oddly VCDS calls it out on a different cylinder, 4 I believe, even though it is in the same cylinder that it is occurring.  It's most prevalent at idle.

Thanks for sharing the insight Jeff. It's good too know that it is not specific to the Nef tune.
I also experienced a very similar phenomenon on a L-box except on VCDS is showing Cylinder #5 where Me7Logger shows #4 consistently lol...

Would you say that it's most prevalent at idle AND only when the engine is warmed up / running in closed-loop?
And so I guess this is caused by some type of leak at the intake manifold?

Cheers!

So this what im getting when adjusting boost and fueling im not sure  as to why im not getting full boost but am absolute sure its not a boost leak (my other logs/tune show this) id like to continue to try to help but this is what im getting, if i load my other tune i see 17 psi and it follows requested......just trying to gain some momentum

Please log all the PID variables, including IMAX and WGDC pre lin

Looks like his I-limit is too low.

By that do you mean the input value of 191, or the axis at 191? If so how do you modify it ?

I mean KFLDIMX is too low.

You have a TC up to 4k rpm

Take a look at throttle angle in your log.
It clearly shows that you are pushing the pedal 100% but ecu is reducing it up to 4k rpm for some reason. It is TC- throttle cut.

There is no throttle cut in his last log?

new logs with a bunch of variables hopefully this helps.....let me know please somethings not right   >:(

i think throttle plate angle is what you talking about ....im not sure if thats a prolem

Increase KFLDIMX. It is too low. Try to envision the duty cycle it's going to take to hit the boost you want. Put those numbers in the last column (1000mb).

daz on my xdf it ends at 850mb......should i just load an mbox or another l box....or is this fine?

It sounds like your xdf is wrong.

this is the only real xdf i could find for lbox would you happen to have a better one? i understand if you dont want to give it up ......lol ill try an mbox for the time being as i hear thats possible on lbox.......

I have that xdf, it's not in there. .... i Just downloaded nyes hopefully that works out, i just adjusted kfldimx so I'll log tomorrow see what happens. .....

The L-box you are using (sw ending in 7004) is an exact match to the M-box xdf.

latest log i adjusted a few things timing, throttle plate opening, and i limit .....i still have work to do but atleast felt better and not getting the timing pull i was before, i think i limit is first though.....if i get this figured out ill post the exact things im doing .... as i think tuning a tip seems to be a little different then mbox

Your imax is a mess.... not sure what you did but it isn't good ;( It should be relatively flat throughout the entire rpm range

Also, whatever it is your are plotting, it is obscuring the details of what you need to be looking at. Concentrate on ONE thing at a time. In this case, you need to get boost straightened out.

yea nye i know i max is the first thing im going after in the morning then relog, but im just happy im atleast getting somewhere.....

Sorry for jumping right back, Ive finally got my car back together and have started looking at logging setup and stuff, but couldnt find a logging config on this thread.

Ive pulled the set of variables out of the CSV that ratherb1 posted above and pasted them into a log config, does this look sensible?

Tried that list, me7logger complained about a whole bunch of them not being supported by the mbox.

https://raw.githubusercontent.com/nyetwurk/ME7L/master/ecus/extras_8D0907551M.ecu
https://raw.githubusercontent.com/nyetwurk/ME7L/master/ecus/torque_8D0907551M.ecu

Thanks Nyet :) Is it common to have issues logging 2.7 A6s? The logger drops out when I start turn the key, it'd be really nice if I didnt have to hack into the wiring.

Change the baud rate to 38400, samples per second to 15. Should not have any issues. If you do, change connection to slow 0x00 and retry.

massaged i limit how does this look? i want to add a little more boost but want you guys opinion first.... i think i still have some work to do as far as i limit .....i want to get rid of the notch

I had read that Cam changeover maps should be set to equal values

I solve the notch issue on mine by copy last column from KFPBRK to KFPBRKNW

I am trying to understand how the boost is determined. I have read a lot and still seem to be stumped. I understand that dd used LDRXN_1_A map as boost in his xdf. I dont understand how he got boost out of that. To my knowledge LDRXN_1_A is the max the load/boost can be. And KFMIOP is what the driver requests based on hill throttle what have you. KFMIOP is the actual percent of KFMIRL that the engine ends up getting.

So we control KFMIOP as a percent based on pedal and hill or whatever, this is then multiplied by KFMIRL to equal our actual load requested as long as this output is below LDRXN_1_A.

I still dont understand how one can predict a specific psi based off this requested "load" business unless there is a conversion factor. I am also confused as to what the axis are for these maps I realize one is rpm but the other is something difficult.

No.. not really. It's complex. I have a description in the S4wiki but there is no way to summarize IOP in a single sentence like that


definitely not :(


The load->boost path is very complex.


The S4wiki tuning page attempts to explain all the various inputs to the maps.

Well im like a chicken with my head cut off now. lol i have no idea whats going on. can you explain how KFMIOP and KFMIRL relate? My waste gates are cracking at 15 psi so i want to get my boost set to 17-18 so i can plug my n75 in and get that dialed in first. Why i am trying to understand how load requests boost.

Sigh. Please reread my post again.


Sigh. Please reread my post again.

IOP is solely for converting load to torque. Its value in the map is based on rpm and load. How does the part of IOP that is above 60% throttle found? This is the part there is no restriction, i understand that but i believe 60% throttle can be applied to any value of rpm. Thefore what is that 60% part.

When load corrected is put into IOP it results in max torque. It goes on to say that max torque will always be calculated from the portions of the map with loads higher than the minimum load corrected from LDRXN. The first sentence contradicts the second. If Load corrected is put in Iop to get Max torque then the portions of IOP map that are being used are the Load Corrected portions. It doesnt say anything about LDRXN unless LDRXN is Load Corrected. I guess i dont understand how LDRXN effects Load Corrected.

Current torque must not exceed torque limit. It says that if IOP is tuned incorrectly such a result can occur. We can raise IOP to Raise Max torque. It hasnt said anything about current torque except that IOP calculates it.

I have been re-reading this over and over. drawing diagrams all sorts of shit, its not clear to me. I ask this question here because i am a noob and followed this stage 1 thread and the next step or question i have is this concept.

KFPED converts pedal to torque. Pick the highest torque value for anything under 60% pedal, and you can put whatever you want in IOP there.

rlmax (load corrected) comes DIRECTLY from LDRXN (rlmx)

PLEASE have the FR handy when reading this part of the Tuning page. There are NO shortcuts here. You have to put in the effort.


There is literally no "noob" friendly way to explain this, much like the load->boost calculation..

So its trying to say that LDRXN will always be the input value to IOP for max torque?


I thought IOP converts load to torque how can you input a torque?


EDIT***after reading a ton of posts it seems like no one knows fully what the hell is going on. people guessing IOP up or down or only playing with the last couple columns. It is separate from IRL but inverse functions lol some people code out the whole damn thing while others say they want the smooth torque delivery. The 60% throttle statement still makes no sense. I think im better off tweeking values and fixing mistakes than trying to really understand what is going on? IRL has to be raised to increase boost or load which in turn means IOP must be tuned. The last columns of IOP gives most people success.

Look at the FR...


Basically, based on IRL/LDRXN, what loads are only attainable at 60% ped or higher

I'd like to help get this thread moving again.   I have an '03 allroad, which is ME7.1.1, so I will need to use a different software base.  Do you think that will be an issue? 

I already have a decent stage I tune for it, but I am all for starting from scratch for the sake of helping others. 

not even a 6 speed allroad?  :)  I understand though as it's a completely different software base.

APB 6sp allroad would be fine.

Alternately, if you could supply ram locations for ALL of the torque variables I have for the s4 mbox, that would suffice.

In a couple of weeks I'll be putting my car back together and will be able to resume some logging, weather permitting.

I agree the discussion has gone into the deep end too quickly.

Alright guys, I'm finally about ready to get started. My progress will be slow as time and money are tight. Working with an '01.5 S4 Avant 6MT, stock except for an SSAC cat back. I've downloaded the M-box Hitachi .bin, and loaded it to ME7 logger - also tracked the changes made through the thread and made updates to the .bin file accordingly. Two questions...

Daz, your footnote in requested boost table says max at 14 PSI for stage 1 but Jeff's last requested boost profile maxed at 14.9. Where should I start once I flash the stage 1 file?

Nye, your list of logging variables contains things not in the bin file in the thread. Should I download a different bin file to create my csv from for logging or just use the variables that are there? I'm also assuming that the entries in your list that still have semicolons in front of them should not be logged. Please let me know where I'm off base, and I'll try to post some stock logs as well as logs of the file as far as it progressed by next weekend.

Thanks and let's get this project going again!

Cragg

I'm not sure I like this approach, it is very confusing. I kinda think the req boost thing is a non-starter; we have to make people start with req load. for better or for worse.


Yea, there is a bunch of stuff that can be added by hand to the .ecu file which is not autodetected by ME7Info..

This is probably beyond the ability of a noob, but I'm including it here for completeness. Since this thread is SPECIFICALLY about the m-box, I suppose a complete .ecu file should be standardized on (and doesn't necessarily have to be the auto generated .ecu file)

I have a file with torque monitoriing vars and a file with misc extra vars here:
https://github.com/nyetwurk/ME7L/tree/master/ecus

Complete ecu/cfgs here:
http://files.s4wiki.com/ME7L-ecu/


Ya those are optional and are included for completeness in case we have to do further debugging.

BTW thank you for throwing your hat into the ring and contributing your time!

I for one have no prob with req load. I plan on going further than Stage 1 and have read enough to be familiar with the concepts. Kinda think if I can't grasp how it works I shouldn't play with it. Don't know if I should be worried about hurting something (fueling, exhaust restriction, IAT cut, ect.) over a certain load threshold. I know we went that high to get load actual closer to the target.


Much rather have complete info in logs than not, as long as my laptop doesn't choke on it!  ;D  I believe we stalled thinking IOP needed attention? Can I convert the .ecu files to .bin to make a .cfg file for logging or can you walk me through manual entry??

No, you just add them by hand to the .ecu file via text editor (or unix "/bin/cat" etc).

Learning that will take me a minute or two. If we are keeping Stage 1 noob-friendly (and what a noob I am!) and using ME7 Logger for logging results, we newcomers will be limited by the variables generated by the bin files available to us that we can import. That said, can we get done what we need to get done with the variables pulled from Daz's Hitachi M-box bin? If not I'll either need some pointed instruction on how to text edit in the variables we need or how to convert your ecu or cfg files to something ME7 logger can use.

Appropriate .ecu and .cfg files are linked in my post above.

Again, it is LITERALLY simply adding lines to the .ecu and .cfg files..

Bear with me brother - like speaking Greek to a redneck. Not a dumb redneck, just need a Greek friend to translate. If I understand, I can just add the variables to my .cfg file in ME7 logger in the order you listed them? The .ecu file though...never modified one with a text editor and I have minimal understanding of what I see when I open the .ecu file with Note Pad. If that's where you're telling me to go (hope you're not telling me to go somewhere else!)...

Again, the FULL FILES are in one of the links above, and you still have to edit the .cfg to add/remove which vars you want to log, so you already have things open in notepad :P

Alright folks - productive coupla weeks. Bench flashing setup done, Galletto cable ordered and received, C5 ECU converted to M-box and Stage 1 program flashed. Nyet, I couldn't find addresses in your files for 5 variables you listed: lamsoni_w, lamsoni2_w, uulsuv_w, ldimxr_w, and B_ldsus. They were not logged. Also, the address I found listed for B_ldipos (0x02) appeared to be incomplete - I included it in my ecu file at (0x0200). Please let me know if that is incorrect!

For those as green as myself that want to add logging variables in ME7L, here's what to do in plain English. In ME7L at "Create ECU File" select the .bin file you are running. That will create a corresponding .ecu file in ME7L. At "Create Template File" select the .ecu file you just made - that will yield a .cfg file in ME7L to select your logging variables from. In order to add additional logging variables you will need to add the new variable into both the .ecu file you created (with the address where it resides) and into your .cfg file to select for logging.

Full .csv files attached for two 3rd gear pulls at 2500 - 6500. Boost shown roughly correlates to 89% of LDRXN that IOP limits request to...where shall we go from here?

Yep, not applicable to Mbox


Honestly, not much more. We are almost out of injector duty cycle. We could sacrifice some fuel for boost but I'm unsure that is safe. I know other stage 1 files run a lot more boost at 14.7 .. but I don't know if that's a good idea.

If I'm reading this right should be safe in duty cycle at lower RPM? The duty cycle is high over 6K RPM, but looks like it would handle more boost down low if we keep boost taper the same. I would like to get rid of that load request drop at cam changeover, and it looks like there's plenty of injector cycle to play with enrichment under 6K. That would allow us to fine tune timing and fuel to keep knock away. I'm running 93 octane though, don't know how safe that would be for the 91 guys...

Why not do like most of the OTS tunes did and hit a higher PR at low airflow then taper off.

Yes, absolutely, we can request more boost down low. I thought he was referring to HP...

I think the practical limit for K03s is around 2.3 PR.

In real life (MAP) numbers, I am unsure of the pressure drop over stock ICs, but at low RPM hopefully it isn't terrible.

I'd be comfortable with 2.3 until 4000 rpm and then start tapering.  That looks to put the airflow in the vicinity of 400 cfm total, with stock IC pressure drop between 2-3 psi.

http://www.myaudis4.com/stock-ic-flow-test/

Unfortunately that puts us at 16-17 psi... which is close to where we are currently.

Much less if you assume intake drop of 1 psi (which puts us at 14psi)..

Unless my math is wrong.

https://www.google.com/search?q=%281000mbar-1psi%29*2.3-1000mbar-2.5psi+in+psi (https://www.google.com/search?q=%281000mbar-1psi%29*2.3-1000mbar-2.5psi+in+psi)

I'm comfortable with 17 psi prior to accounting for the pressure losses.  I think even if the compressor were seeing 19 psi at 4000 rpm it would still be on the map and operating at good efficiency. 

http://www.myaudis4.com/2014/05/03/stock-fats/

But if 15-16 psi MAP were all we wanted to do I'd be fine with that as well.  It just seems that stopping at 12-13 would be very conservative.

Works for me!

I'm working with a 551T a6 2.7t currently. I'm seeing issues with the ASR turning on limiting power until a hard ASR code pops on (ASR light permanent then power is consistent. Maybe RLVMXN/RLVSMXN is the issue? Although I didn't think 14psi would exceed torque limiter in those maps. I'm not seeing any Cel's regarding the torque control limits.

With the ESP button pressed (ESP disabled) you should not get any ASR.

Italian tuned KFMIRL to 110% of stock and used berTTos's KFMIOP interpolator to generate new map with bottom of X axis averaged at 13.05 and last 2 columns at 87% and 100% of max load request in KFMIRL. Matched new KFMIOP X axis with X axis in KFZWOP and KFZWOP2 timing maps. So essentially have generated a matched KFMIRL and KFMIOP set where the KFMIOP map requests the load that I'm shooting for (as a % of LDXRN map). Still need to back off timing to get rid of that knock at 5K rpm. Car feels great, like a different personality but still getting limited load result. Suggestions on where to go from here? Nyet, can you elaborate on your last suggestion of flattening KFP maps at 3K & up?

Your mifa/misol is pretty low, might need to tweak KFPED a bit, or bring up KFMIRL in the lower torque request areas...

Don't forget the INPUT to KFMIRL is mifa/misol.

You're only at a max of 75% on mifa, which corresponds to what you have in irl/iop and the load values you are seeing.  Log mrfa_w (Driver Requested Relative Torque) as well to see what that is.

This is a summary of my work thus far.

First thing was to raise the rev limiter, NMAX, to 7000 rpm and NMAXF (RPM DTC limit) to 300 rpm over NMAX:

(http://www.myaudis4.com/tuning_images/nmax.png)


Next I adjusted the Pedal Position for Desired AFR (LAMFA Axis).

Stock values are 0.5, 0.6, 0.7, 0.8, 0.9, 1.0 and they are changed below to 50, 60, 70, 80, 90, 97.5

(http://www.myaudis4.com/tuning_images/corrected_afr.PNG)

The desired AFR is altered in the last two columns (90, 97.499) from 3000 rpm and above, enriching the mixture from 14.7 to 12.5.

Next was modifying LDRXN to alter the boost profile, raising it slightly from the stock profile.

(http://www.myaudis4.com/tuning_images/nef5/nef5_boost.png)

Then some timing adjustments.  On the left side are the original tables and on the right are the current values.  For KFZW2 I lowered the values slightly to try and eliminate the ignition retard that was occurring around 2000-2500 rpm.  For KFZW I raised some of the timing values to try and have a smoother transition around 3500-4000.

(http://www.myaudis4.com/tuning_images/nef7/nef7_timing_before_after.jpg)

And lastly, a comparison of this tune versus the baseline stock file and a pull from a different APR chipped S4.

(http://www.myaudis4.com/wp-content/uploads/nef7_acceleration.png)

Thanks for summarizing your changes. Very easy to follow, and understand.

I'll need to add misol_w and mifa_w to my file and re-log.  Is there a common name to use with those variables?

This is the boost curve I'm thinking of shooting for, based off the MAP reading.

(http://www.myaudis4.com/wp-content/uploads/planned_k03_boost.jpg)

How did you end up correcting this? Can't seem to figure it out.. :S

I had thought the original values were percentages as they are written 0.5, 0.6, etc.  To get 50% I needed to change the values to 50, 60, etc.  Is that what you are asking?

The XDF Daz posted will enable you to edit the axis values.

Yes sir, I have been messing around with Nefmoto stage1.xdf and 8D0907551M-20120516.xdf Keep on getting .5 .6 .7 ect.. :S  (would you be willing to share yours with the changes? Would be greatly appreciated! :)

That's not the XDF I was referring to.  The one Daz posted http://nefariousmotorsports.com/forum/index.php?topic=6695.msg62391#msg62391 (http://nefariousmotorsports.com/forum/index.php?topic=6695.msg62391#msg62391) called the Nefmoto stage 1.xdf has been tailored for this thread to just have the content you need to make the changes mentioned and it uses plain language terms.

Also i forgot to mention, we don't actually have a milsol (not to be confused with misol) location :( So tracking the rlmax path is somewhat difficult... but you can just sort of assume it from the highest torque cells of IOP

Yup that is the one I was messing around with, Loaded a M bin and when I look at the accelerator scaling I get the .xx %  Can't wrap my brain how to change it to display it in xx.x %  Sorry, should have been more specific. (Plus im kinda new to this whole thing).

Ah had to re read this, but yes the .50 .60. 70..  Do I just enter 50 in place of 0.50 and so on?

Yes

Oh boy do I feel dumb! lol thought id have to adjust some formula spin around the room twice and cough to get this correct... (usual way of getting Audi stuff to work) :D

(http://www.myaudis4.com/wp-content/uploads/mifa_rlmax_plsol.png)

got the actual log handy?

File is here:

http://www.myaudis4.com/tuning_images/logs/ (http://www.myaudis4.com/tuning_images/logs/)

Odd. rlmx looks fine, ldrlts looks fine... rlmxko looks low

That leaves frxt/frxta and ldrlms..

can you log

frxta
frxt
ldrlms

Aha. It is very cold.

I think you want to flatten KFTARX so it doesn't pull load under 20C (note that this isn't a good idea when running near map limit, because the load->boost temp dependent calculation can send your req boost over the MAP limit if you aren't careful when it is cold - but for stage 1 it is no big deal. if anything you'll get more hp when it is cold)

(http://nyet.org/cars/images/me7log_20150327_193255.png)

This was last fall when the outside temp was 61F.

(http://www.myaudis4.com/wp-content/uploads/fall_2014_rlmxko.png)

Post adjustment to KFMIOP and KFTARX.

(http://www.myaudis4.com/wp-content/uploads/post_kfmiop_kftarx-e1427563140130.png)

Can you bump LDRXN a bit more? Maybe something like:

109.99   120.00   132.75   145.81   152.46   176.93   176.91   176.91   174.40   171.89   168.38   162.87   153.94   146.91   142.43   131.93

I think this will hit our target boost.

Done.  Results are:

(http://www.myaudis4.com/wp-content/uploads/nef9_new_ldrxn-e1427592519221.png)

I added this log to the folder shown above.  It's the 3-28-2015 csv.

Time to start logging PID variables.

Looks like not enough i max.

Please review the s4wiki and how the PID functions.

What Nye is trying to say is that we have increased the amount of boost that's being requested, but not told the ecu what it's going to take to achieve the new goal. The next step is increasing the values in DIMX to reflect what it'll take to hit the boost we want.

I might be late to the party, but the last log shows boost actual following boost desired quite closely.  DIMX and the other PID tables are fine how it is.  There's a load limiter that's being hit.

Here are the changes made to KFTARX and KFMIOP.  Left is the delta from a stock tune for the values that were changed, and the right are the resultant values being used.

(http://www.myaudis4.com/wp-content/uploads/kftarx_kfmiop.jpg)

And this is the boost data from the last log, which contained the modification to LDRXN that Daz mentioned making.

(http://www.myaudis4.com/wp-content/uploads/nef9_boost-e1427766662743.png)

Forgive me, I was looking at the wrong lines.. Yes... something needs to be tweaked in the PID tables to make your actual line up with desired. 

What does your WG duty cycle variables look like in the log?  I usually log ldtvm (WastegateDutyCycle) and ldtvr_w (WastegateDutyCycleBeforeLin) to determine if I should change something in DRL, or in DIMX.

Here's the change I made to KFLDIMX.  On the left is the changed values relative to a stock baseline, and on the right is the resultant table values.

(http://www.myaudis4.com/wp-content/uploads/kfldimx.jpg)

The resulting data for a 2nd and shortened 3rd gear pull are here with a 20150331 filename:

http://www.myaudis4.com/tuning_images/logs/ (http://www.myaudis4.com/tuning_images/logs/)

I say shortened because around 4500 an anomalous event occurred and I am still uncertain as to what it was.  Need to see if I ran something over, popped an IC hose, or something else, but WGDC instantly dropped to 0.  It probably won't be until tomorrow evening that I can get around to investigating further.

Possible positive deviation limp?

In any case, before you continue logging you need to add the various PID variables to your setup, particularly I max and the various I limiter adaptations.

Keep up the good works, Jeff. I hope everything is ok with your car.

You contributed a lot and for a newbie like me I have been reading this topic like a best seller :)

Daz reminded me that early on we had raised KFMLDMX in a discussion outside the thread.  These are the changes that were made and that are in now place with the development tune:

(http://www.myaudis4.com/wp-content/uploads/kfmldmx.jpg)

My progress on this tune is going back on hold while I have a precautionary rebuild done on the K03's. 

K03 Troubleshooting (http://www.myaudis4.com/2015/04/05/even-more-perplexed/)

I've read this thread over and over and decided to start building my own tune. I've got a 6-Speed, M-Box (converted from A) S4, exactly what this thread was intended for. I'll throw a kink into the mix and say that I'm tuning at 9000 feet elevation, which I believe has effected my tuning quite a bit (especially for these Boost (in PSI) Calculations). I've coded EGTs and Rear O2s out per the wiki.
Car info:
-2000 S4; 154K, Engine refreshed and resealed at 146k) 6 Speed, converted to MBox with Bosch MAF.
-Piggie DPs (rear O2's removed)
-2.5" custom catback with magnaflow muffler
-Custom 3" Intake (Bosch MAF)
-RS4 Clutch
-710N Diverter Valves

As a newbie tuning my S4, this is what I see.. So I do know that these desired boost numbers in PSI are just estimates. So I began with logging stock and moving along with the thread and then I kind of deviated with my requested load and boost numbers. I decided to do this because when I actually log, my requested boost levels are much lower than the "requested boost (in PSI)" within Tunerpro. I also noticed my engine load is not nearly what is actually being requested.. I'm not sure why this is happening. I was thinking maybe it had to do with some of the requested load tables etc.. but none of them are even close to the loads I'm getting..

I've attached the CSV.

Don't worry about actual load not meeting requested. BTW I dont see any requested load in your logs...

You can definitely get away with a bit more req load.

Ah I must have missed it. My LDRXN looks like this. I know these are maximums, but I'm not getting anywhere close..


111.35   123.40   135.47   166.50   173.39   176.84   178.57   173.39   171.68   170.98   170.30   168.21   161.68   155.11   150.99   144.78

I did some tweaking of KFMIRL and used berTTos interpolator to get KFMIOP. Hoping that will help a bit.

I've attached my most recent log. Some things of mention:

-I'm getting timing retard, namely on cyl 4. I will address this as soon as I smooth out the boost.
-Seems like my requested vs. actual boost is pretty close per my logs.. Although it's no where near whats being requested per LDRXN
-Requested load vs. actual load is close.. but I can't seem to pinpoint exactly where these are coming from. I know its a combination of things, but how do I increase that to say follow closer to my LDRXN. Per the tuning wiki, at WOT the ECU typically follows those values. I think I'm hitting a load limit somewhere, and I can't seem to pinpoint it..

Look at KFTARX, and KFLDHBN. Finally, KFMIOP. Your limiting factor will be in one of those.

Stock KFMIOP is fine for stage 1. However, KFMIRL needs to request enough load at 80% torque request.

Here's the current boost on a couple abbreviated pulls.  I'll get more/better data later, just posting as I'm back on the road.

(http://www.myaudis4.com/wp-content/uploads/nef10_boost.png)

Looking good, but that much underboost might cause a positive deviation code at some point.. will need to bring up IMX unless your wg is already at 95%

He won't get deviation from that. Not unless the MAF reading is way too coincide with actual boost.

I'm pretty sure wgdc is not near 95%, but let me get a better set of data before taking more action.  I had upped IMX previously, mentioned in this post (http://nefariousmotorsports.com/forum/index.php?topic=6695.msg73090#msg73090) and shown below, do you think it needs more work potentially?

(http://www.myaudis4.com/wp-content/uploads/kfldimx.jpg)

Log WGDC on one axis, boost (desired and actual) on the other. Post a screenshot.

Boost from above post with wgdc added.

(http://www.myaudis4.com/tuning_images/nef10/nef10_boost_wgdc.png)

Need more I-limit. Raise DIMX

Looks to be 83-85 percent in the undershoot areas.

Post logs with I contribution and I limiters...

If you guys could include the 'why' it might make this discussion more educational.  Just saying raise this, log that, doesn't tell the reader the reason for doing these things and leaves a person with just following the step-by-step without the benefit of knowing why things are being done.

Post logs and I will explain with graphs. Also, review the tuning page:


Once D is zero, and P is no longer changing, I+P is steady state. If I is too low, you'll have persistent underboost, regardless of P, since as it approaches 0 (lde=0) I represents the sole contributor to wgdc.. Always post the logs, and I can post graphs showing what is wrong. Also, google "integrator windup" in the PID control context.

The main thing to realize is that the PID trims are all done via altering the I-limit... which means the I limiter (along with the linearization map KFLDRL) is the primary place to do PID tuning.

Thanks Nye.

Here's a better set of data, hopefully with pertinent variables.

Nef10 CSV File Log (http://www.myaudis4.com/tuning_images/nef10/me7log_20150601_Nef10.csv)

Hey guys just wanted to throw my hat in for the effort.

2000 6mt car running the Bosch mbox software daz posted. So far I've done the AFR tweak, RPM tweak, and prelim boost tweak, I've been using the xdf Daz posted for Nefstage1 also did the post cat o2 tweaks from daz xdf as my car has a downpipe back exhaust with no cats.

I haven't tweaked Timing yet, but I'll be contributing my logs from here on as well.

I'm in Central FL, so we should be able to get some good warm weather info.


BTW just a dumb question but is it OK to switch xdf for the same tune IE: start on Nefstage1 then progress to a more open xdf once I need to explore other variables?

Absolutely. As long as the xdf is for that software!

As you can see, I-Max is completely inappropriate for requested boost. It should (roughly) be the dc required to make requested at every point. Req boost is rising until 3000rpm, but i-max does the opposite. Then it starts tapering, but I-max is increasing.

So:

By 2600 rpm, we should be seeing requested boost, which is around 16psi.

If you look later in the graph (at 3100rpm) , 16psi requires around 82-83 wgc. At 2600, I-max is around 82, and given that Q2 is going to result in at least -3 or -4, we'll probably need an I-Max of around 85-86 at 2600 rpm.

By the time req boost is at its max (3000rpm), I-max is almost at its minimum! Given that we probably need at least 87ish dc to make 16 psi, I-max should be rising to match the requested boost.

Also, the requested boost should really not be increasing during late spool, it should really be flat (or tapering). This will make tuning the PID easier (since I will accumlate faster, it is more likely to follow i-max).

Finally, I have omitted linearization for the time being since it only complicates the discussion: suffice it to say the purpose is to have a normalized dc->actual boost curve.

That is to say, you need more DC up top for a given boost, so LDRL should be tuned such that a COMPLETELY flat pre-lin DC should result in a post-lin DC that is increasing sufficiently to keep the boost at a given PSI.

Ideally, when tuning i-max, you should be thinking about corresponding changes in LDRL accordingly. If you are finding you are having to have an increasing i-max to keep up with a flat boost request, instead of having a rising i-max, consider a rising LDRL instead.

Thanks Nye, great explanation of what's going on with that graph.  Funny part is that when I just looked at the boost curves I thought they didn't look too bad.

Going back over the past few comments it sounds like refining DIMX and KFLDRL are the way ahead?

Yes, although I think we may ALSO want to get req boost to be flat during spool instead of that odd slowly increasing ramp, to simplify PID tuning...

Now we're talking about URL....

I don't want to try and change more than one thing at a time, how would you suggest I proceed?

Edit: How about trying to flatten the requested boost curve first?

therealnap0le0n, let me know if you are staying up with the changes, I think it would be best if we move along together making the same changes so as not to confuse the thread with multiple tune variants.

I got left behind at the first boost correction but that's ok I can catch up pretty fast as I'm getting more familiar with this it looks less like chinese. That being said it seems that we've exhausted the capability of the NefStage1 XDF and should proceed with a more complete one. I think at this point I prefer an XDF that is going to be using the bosch terms like you are using anyhow.

At the point I am at right now I'm trying to get all of the correct variables in my logs, and I've discovered a couple of weird instances.

1) I have logged using the ME7logger GUI and its boost Act seems to reflect what my PSIg ( The guage reads about 13psi )

2) Logging using VisualME7logger Boost Req and Act is reported a few PSI less than my PSIg (it reports about 10 - 11 psi with the tune set to 13.5psi~ )

3)I have no cats and with my factory tune had a CEL for 2ndry cat warm up/efficiency codes (duh), now that code is gone but i have 17522 oxygen sensor B1 S2 resistance too high intermit, could this be related to the emissions options I selected in the tune? I was under the impression for catless i should have converter diag unchecked, rear o2 unchecked, and post cat fueling correction checked based on the tooltips for these items.

4)At 12+ PSI the car starts to have a multiple miss fire, I need to check my plugs I think they are bkr6e with about 10k miles and I think a .025 gap

its been raining non stop so I'll pull my plugs either Tom after work or Thurs

here is a 2nd gear pull from this morning with VisualLogger

-----------
Edit: I think there's something wrong with the log that I provided because it seems to indicate my Abox files even though I definately used my M box files....

May as well do both, either way you're going to need more wgdc before the current peak req boost. Besides, I almost always tune imax, req boost, and ldrl in concert since they are intimately connected.

It is going to take very very long time if we can't multitask a bit. As it is, this is taking much longer than I originally envisioned.

Ok, I'll try messing with all three.  I'll make some proposed changes and post them, unless somebody else wants to make a suggestion.

Pulling my engine twice along the way has been a distraction, I think when we've been able to make progress it's been reasonably well paced.

Ok I think I figured out my logging issue. Here's a 3rd gear pull from this morning, a bit of miss firing still have to look at the plugs

Sounds good to me, I will definitely try to keep things rolling as best as I can.

Thanks again for the time and effort you are putting into this!

That has a very different req boost profile from flyboys' logs.

Lets concentrate on a single file here. If you want to start another thread, do that...

I need to re-review the thread I was 99% sure I was using the same values as flyboy. I'll take the time to go over things again and become current tonight.

If you want, email me the bin and I'll check it against the changes I made with mine.  You're still using Daz's XDF?

The thread is kinda hard to follow in the middle but thanks to your screenshots im actually caught up to the part with imax, boost req, and ldrl tuning so Im 90% sure im concurrent. PM me your email and I'll send you what I have.

Also I've since left Daz's XDF for Nyets 8D0907551M-20120516.xdf as it is more complete.

My email is shown in the envelope in the signature area.

I've also been documenting the bare bones steps of the Stage 1 phase here (http://www.myaudis4.com/stage-1/) and now Stage 2 here (http://www.myaudis4.com/stage-2/).

I found that the last tweak to LDRXN had not been incorporated into his file.  Perhaps that is the culprit?  Aside from that all the other changes match up with what I have.

For flattening the boost curve, do you suggest raising ldrxn at the lower end to bring requested boost up?  I'm also thinking that holding a bit more boost mid to upper range is possible, near the top the injectors are close to maxing out so I'm fine with dropping down to around 10 psi in the 6000 rpm range, but it seems like that middle upper area could support a little more boost.

I think for the sake of safety we should leave a buffer of injector duty in the top end, even then the "fun" areas of driving tend to be the mid ranges.

one thing to remember though is tons of boost at low speed low RPM will strain the motor and ignition more. But I'm sure we can squeeze it a bit.

Well I made a stupid mistake and bricked my ECU.... So it's gunna be a little bit before I can get going again.
-----
Edit: Spare A-box will be here Monday so I'll be back on it then. Afterwards I'll have to look into unbricking the current ecu I have.

Candidate changes for LDRXN, red line is existing and black line is proposed new curve.

(http://www.myaudis4.com/tuning_images/nef11/candidate_nef11_ldrxn.png)

and KFLDIMX, values are increases over the existing values.

(http://www.myaudis4.com/tuning_images/nef11/candidate_nef11_kfldimx.png)

After further comparing bins I discovered a change that was made as a result of this discussion (http://nefariousmotorsports.com/forum/index.php?topic=6695.msg63740#msg63740) but did not get posted to the thread.

KFLDHBN (max requested PR) was modified as shown below.  On the left is the resultant table I'm now using, and to the right is the delta over stock.

(http://www.myaudis4.com/tuning_images/nef10/nef10_kfldhbn.jpg)

You should request more load even earlier es k03 comes on boost early. Just a few 100rpm.

Earlier than 1600rpm?

Yes, or twiddle with the KFURL path (load->boost) to make req boost flat with respect to req load.

In any case, cleaning up IMAX is way more important I think..



I have yet to overlay your data with the compressor map, so I'd hate to make a guess at this point...

I think there's room to have a little more boost around the 3500-5500 rpm range.

(http://www.myaudis4.com/tuning_images/nef10/nef10_compressor_map.jpg)

This is my new proposed curve (black) versus the existing curve (red).

(http://www.myaudis4.com/tuning_images/nef10/nef11_proposed_ldrxn-2.png)

Looks good to me!

I would have max request at 1600-1700rpm.
He has it at 2000. Bit late imo.

Two data sets with the recent revisions incorporated.  Was a bit warmer today, around 82F at the time of logging versus mid 60's last time out.

Dataset 1 (http://www.myaudis4.com/tuning_images/nef11/me7log_20150606_nef11_a.csv)

and

Dataset 2 (http://www.myaudis4.com/tuning_images/nef11/me7log_20150606_nef11_b.csv)

what we're the ldrxn values you used to create the new curve? I would like to be on the same page value wise but i cant get exact values from the curve

Based on the injector DC from the last logs I'm going to pull back on the top end boost a little more.  New curve for LDRXN in black, last is the red line.  New values shown in the table.

(http://www.myaudis4.com/tuning_images/nef12/nef-12_ldrxn.png)

Updated the DIMX values slightly as well, dropping them a bit.

(http://www.myaudis4.com/tuning_images/nef12/nef-12_dimx.png)

Got my ECU in and flashed the tune with the latest revisions.

3rd gear pull is here.

Note: When viewing in ECUplot, if I leave "filter data" unchecked the data reflects what I see in car ie: large boost spike at first throttle, if I filter data the spike disappears and my boost data differs from what my boost gauge reads.

Also unfiltered my Boost Act follows my Requested very closely, when I filter it the Act is much Different than requested.

You need to pay attention to the scales.

Additionally, start the log at 2000. DO NOT go wide open UNTIL it says "Press C to stop logging". This ensures that the logger has a connection, and is working when you go WOT so as not to miss the first part of the log.

Scales in ecuplot or tunerpro?

Also I'll log using The regular ME7logger as I think the VisualME7 logger is pulling the data differently.

That looks UTTERLY different from Flyboy's plot.


That is because overboost is causing throttle cut, so the filter is cutting the run short. in 'configure filter' set "Minimum Throttle" to 20.


I max is WAY too high for your car, which is terrible, because it seems your hardware is VERY different from Flyboys. Not good.

Also, I am unsure of the source of that huge boost request spike.

Looks like too MUCH I Max this time... but the IAT's are different, so some changes in KFTARX might be required here, unfortunately...

Look at the difference... very very different wg settings to produce the same boost in the different cars

OR, LDIATA.

Yes, this.

I've not worked on ANY stock k03 cars: do wastegate behaviors really vary this wildly? 20%  :o

Perhaps we each do a pull with N75 electrically disconnected to see what boost level results?   Make sure the difference is not attributable to hardware.

His exhaust does not have the catalytic converters where I have them, though mine are aftermarket.

wouldn't we need to leave it electronically connected and physically disconnect it? I'm not opposed to that.

Also did a pull this morning getting on the on ramp to work with the regular ME7Logger this time. Got a huge throttle cut at 5000 RPM much warmer this time too, I suspect my car is not going to make this easy on me.... but hey thats half the fun eh?

One more time, read the tuning page :)

http://s4wiki.com/wiki/Tuning#Calibrating_KFLDRL


That sounds like one likely culprit.

I'm thinking of checking to see how the wastegate actuators compare.

KFLDRLAPP is PERFECT for this!

http://nefariousmotorsports.com/forum/index.php?topic=517.msg4028#msg4028

Yes. It's VERY easy to make a stock turbo file.

It's VERY difficult to make a stock turbo file that works well on EVERY car.

Tips are even worse.

I think we could definitely use more cars to pull data from.

Plus i think environment and car variables are coming into play, at this point I need to figure out the boost spike on first throttle tip in.


For the record:

User: TheRealNap0le0n (Alex)
Car: 2000 S4
ECU: A-Box running Daz's 551M - Bosch Software
MAF: Bosch
Injectors: Stock
Plugs: NGK BK6RE .028 Gap
Motor Mods: Stock
Turbos: Stock as far as I know
Diverters: 710N
Intake: Stock/DARIntake Mod
Exhaust: Catless DP's, Single 3" Single Muffler
Fuel: 93 Oct
Location: Orlando, FL

Too much IMX, as i've already explained :P
Please reread my posts.

You have a non-stock exhaust, which means you can't use the same imx as a bone stock s4 would. You are on your own. If you like, start a separate thread for remote tune help.

yea will do, thanks for the help so far guys

Nye, I'm on a non-stock exhaust too.  Realistically I don't think anybody with a bone stock S4 is going to be interested in this site.   A stage 2 S4 usually implies some exhaust upgrades if not IC's.  I'd like to see us try and work through it, and if it proves too difficult then branch off, but companies like APR and GIAC had to account for different hardware when they did this, I think it is at least worth a shot.

Here's the change I made to KFTARX, highlighted in red.  I split the difference between previous value and the column to the right, moving left, to scale the values down so as to have less boost with higher temps.

(http://www.myaudis4.com/tuning_images/nef12/nef-12_kftarx.png)

and the results with these last three changes above [LDRXN, DIMX, KFTARX] incorporated.  78F outside.

Nef-12 Datalog (http://www.myaudis4.com/tuning_images/nef12/me7log_20150609_nef12.csv)

Interestingly the car felt stronger at the top end despite not showing much difference in the acceleration curve.

Apologies for leading you down the garden path - I'd rather not mess with KFTARX just yet.

Hows this: Let's get LDIMX dialed first... bring it back down a bit so you don't get all that overshoot. Put KFTARX back (sorry, my bad) and tweak LDIATA as daz suggested once we get a sane IMX at a given operating point.

given safe limits I'm all for testing the community tune so we can come to a "suitable" for most tune, and then through the learning I've done with this I'll hopefully be able to massage my personal file to my cars needs.

Let's work on LDIMX first and not get too ADHD on the other variables.

A slight digression here, but I think it is relevant to the discussion:

Their stage 1 was ALL over the map, and if you'll recall, people had all sorts of throttle cut issues when they deviated from the stock exhaust (piggies, high flow cats etc) even a little (APR and GIAC had no idea what KFDLULS was, lol).

Hence, "throttle cut defender", the stupid accel pedal stopper, etc.

The IMX adaptation system was never good enough to compensate for these huge deviations...

I don't like the idea of dorking with KFTARX at this point (keep it flat, all 1's IMO). I'd use LDIATA if we're sure it's temperature related, leave request alone, just fix the PID.

The problem with a flat KFTARX is that you'll have rising req boost as temps go up...


Agreed.

The changes I made by and large were a shift toward 1, so I'll leave KFTARX be for now. 

Next change I'll make will be to drop DIMX a little.

Feel free to do several iterations. IMX is relatively safe to move around wherever you want at these boost levels.

Here's with DIMX lowered a little and I also slightly altered the LDRXN to not request full load quite so fast.  I'll post those changes soon.

There's three partial third gear pulls, the last is interesting in that it starts at low rpm and boost never overshoots but the two prior that start at a later rpm do.  What's the recommended way to counter that?

Nef 12a data (http://www.myaudis4.com/tuning_images/nef12/me7log_20150611_nef12-a.csv)

Tune Q2... now, Q2 is a tricky beast, pay close attention to the axis...

The RPM axis is clear: if you get different overshoot depending on RPM position, vary Q2 with RPM

The other axis (lde, iirc) is not so clear: it allows you to schedule the Q2 gain depending on lde; that is to say, you can make Q2 bigger or smaller long before actual meets requested, or JUST as lde nears zero..

But the D component from Q2 is already proportional to how "fast" lde is changing (e.g. spool), so counter intuitively, if you have a lot of overshoot on fast spool, but not slow spool, you may want to increase Q2 for LOW lde.

Conversely, if you have overshoot on slow spool, but not fast spool, you want to DECREASE Q2 for LOW lde.

You can move the whole thing up and down by toying with IMX, of course. Typically, it is easier to just make IMX very conservative so you dont overshoot, so you can keep Q2 low. You might give up a bit of spool, but it is likely better to have undershoot than uncontrollable boost oscillations (which you will ALWAYS have on overshoot.. that is the nature of PIDs).

Now, keep in mind that on top of this, you also have the IMX trims going on.. but if lde is never that big for extended periods of time, the ECU should not be trimming IMX much.

Now that I look at your graphs though:

You can see that D is basically zero when we most need it to control the overshoot....

this is because your req boost is rising just as actual is about to meet requested. That means the slope (derivative) of lde is smaller than if req boost was flat.

This is why it is almost always easier to tune Q2 when req boost is flat near expected overshoot....

Ok, it sounds like the first order of business is still to try and get that requested line to flatten out.  I'll keep working LDRXN to make that happen and then revisit the other changes to DIMX, and possibly Q2.

Please let me know if anything I'm explaining is unclear; it is a complex topic and I am trying hard to make it as layman friendly as possible.

Tuning PIDs is NOT an easy thing; in fact, it is very rarely done by hand... usually automated systems are used...

Below is the change to DIMX referred to yesterday, and also the most recent tweak to LDRXN.

(http://www.myaudis4.com/tuning_images/nef12/nef12-c_ldrxn_dimx.png)

These are the results produced.  Looks to me that I need to raise LDRXN around 2500 rpm some more.

Nef 12c data (http://www.myaudis4.com/tuning_images/nef12/me7log_20150612_nef12-c.csv)- just boost onset

Here's how AFR is looking with the current tune and weather conditions:

(http://www.myaudis4.com/tuning_images/nef12/nef11_afr.png)

Almost there with the requested boost line.

(http://www.myaudis4.com/tuning_images/nef12/nef_12_d_ldrxn_result.png)

Looking good! BTW instead of fiddling with IMX to clean up the overshoot, you could increase req load until you get a good req boost taper from 16 psi at 2700 on down, etc. Not sure where that is on the compressor map, though.

I agree. Don't trim DC, rather bring up request, but even so, that looks pretty solid.

By continuing to raise the requested line that much I'm not going to cause actual to start rising as well?

Well yes, a little bit. Are you good with 16psi? I'd shoot for 17, but that's just me.

I'm comfortable with that boost level at low airflow.  I'm hoping we are going to be keeping the tune smooth, I think there's some benefit to showing how to develop a refined tune versus one that aims to max out the turbo's and fueling.

What would the highest request be on stock set up while remaining efficient? I think 17-18psi max without overworking the turbos

We would be trying to maximize low airflow/ low rpm power due to hardware limits up high

I'm currently using the stock IRL values, I think they'll need to be raised.  Does that sound right?

If you have verified that is the limiting factor, then yes.

Not yet.  LDRXN is below the max IRL values, but if I am to keep pushing the requested boost up it looks likely that LDRXN will exceed the max IRL values, and I thought they were the cap for LDRXN.

Again, it isn't that simple because torque request input to IRL isn't generally 100.

I bumped up LDRXN a little more and dropped DIMX some.  The resulting req and boost curve look very much like the last one I posted.  Thought I'd give that route another try before abandoning the methodology and strictly try to raise the requested line to meet actual.  Since the results weren't very progressive I'll look at a more aggressive LDRXN curve around the boost ramp range.

I haven't been keeping up with your latest changes Jeff since last time we spoke (far too busy with work/family), but just for reference I was having that 20PSI spike on initial throttle, I dropped I max down to stock and my car now goes to 16psi ( +1 psi of request ) very briefly, and then follows requested to the dot until about 5500rpm where the taper continues to follow requested at +1 psi again. You may not need as much DIMX as we intially thought.

I'll keep that in mind as I progress.  For the moment I'm trying to up the requested line as Daz & Nye suggested.  This'll be my last update for a couple weeks as I have some other activities to attend to.  Here's the new LDRXN, black line, and below it the results (csv file) it is producing.

(http://www.myaudis4.com/tuning_images/nef12/nef-12-f.png)

Nef-12-f data (http://www.myaudis4.com/tuning_images/nef12/me7log_20150619_nef-12-f.csv)

Request: can you add rlsol_w to the log?

Need to figure out if rlmax is capping rlsol or not.

Sure, I'll include it with the next log.

and I'll add misopl1 too. 

Here's a slightly further tweaked LDRXN.  Black line is new, red is the last revision.

(http://www.myaudis4.com/tuning_images/nef12/nef12g_ldrxn.png)

and the results in 76F temps.

(http://www.myaudis4.com/tuning_images/nef12/nef12g_boost.png)

Here's the Nef 12 g data file (http://www.myaudis4.com/tuning_images/nef12/me7log_20150702_nef12g.csv), (with rlsol_w, but not misopl1 added).

Did you ever make changes to DIMX you mentioned?

That looks FAN tastic! Good job dude.

Looks like we may be able to add a bit of timing even!

Although I am concerned about the injector duty up top, we are plumb out of fuel. Any chance you can get a wideband?

I have not messed with DIMX since revision 12 A.  I'm going to see if I can rein in the overboost from 4000-5000 rpm by dropping DIMX a little there.


Thanks, that change worked out pretty fast.

I have a wideband wired up, I just did not log it on this latest time out.  The AFR has been quite rich, I was wondering if the ambient temps were playing into that.  I'll log AFR next time out and post the results.  It looks to me like some fuel could be pulled out.  Also the boost is dropping pretty rapidly after 4000 rpm, with the intent of not running out of fuel.

It looks like BTS kicking in, but need to log lamfa just in case. Nothing wrong with 11.5 .. if anything it means we can add some timing.

fyi, in case you are missing any ram vars:

http://files.s4wiki.com/ME7L-ecu/my_8D0907551M.ecu

Don't touch DIMX.  Your last logs look close to perfect as far as the boost profile goes.  I wouldn't touch a thing, but if you must, just pull a little from kfldrl and call it good.

Thus far we have not touched KFLDRL (Map for linearization of boost pressure), not even mentioned it or what it does.  Could you explain what it does and why it would be the table to change, as opposed to DIMX, and how we might go about modifying the table?  My hope is that we can educate along the way, not just show a step by step to making a tune.

Stock KFLDRL should be fine, it is calibrated for K03s with stock wastegates. Discussion of KFLDRL is already all over the rest of this site (not to mention the wiki, and any other place which discusses PIDs) and IMO beyond the scope of this thread.

I doubt AFR is much different now than in this post (http://nefariousmotorsports.com/forum/index.php?topic=6695.msg75842#msg75842) since all I have been working on is the initial boost rise.

Assuming that is the case, some further changes to timing are in order?

Need to look at I-limit adaptations to determine if this statement is correct. They can disguise very large discrepancies.


Negative. DRL represents an ideal calibration for the stock wastegates. It should not be touched in this case.


Lean it out, add more timing. If we can do that EGT's will come down, we can leave TABGBTS relatively low (I'd recommend 750, maybe 800C) and maintain enrichment, but only as needed. I'd also suggest adding a few more variables to the cfg file:

tabgm            ;{EGTModelBeforeCat}                ; {Abgastemperatur vor Kat aus Modell}
b_tabgbts       ;{CondEGTForPartProtection}        ; {Bedingung Abgastemperatur für Bauteileschutz überschritten}

Are you referring to the exercise we did comparing I-Max to WGDC?  Continue to refine that?


1. Lean it via LAMFA - Try to get closer to 11.5 on the wideband?
2. Increase timing via KFZW/KFZW2 - Do this until I see some timing retard and hold there or pull back slightly?
3. TABGBTS (EGT Threshold for component protection) is not something we have touched thus far.  Current value is 499.99C, what's the reason for raising it?  If EGT's are going to come down with the changes to fuel & timing why would we need to raise the EGT threshold?
4. I'll add tabgm, tabgm2, and B_tabgbts to the config file.

If we are leaning fuel afr should go up not down.

Last I saw we were targeting 12.5 lamfa, 11.5 would be richer.

That's what the value in the table is, but at the tail pipe it's much richer.  I'm going to modify the table to get the actual reading closer to 11.5.  If things look good at 11.5 I can try and lean it out to 12.5.

No, don't modify LAMFA from 12.5.

Raise TABGBTS to 750, log it again adding the variables posted.

This is with TABGBTS raised to 750C.  Also did some more tweaking of LDRXN and DIMX.  I also complicated matters by putting higher flowing aftermarket IC's on in place of the stock ones.  They don't look to have messed things up.

Here's the current AFR at tailpipe:

(http://www.myaudis4.com/tuning_images/nef12/nef-12h_afr.png)

And the latest log, Nef-12H (http://www.myaudis4.com/tuning_images/nef12/me7log_20150705_nef-12h.csv),  with the additional variables added.

No. He's referring to the adaptation channels which modify the i-lmit.


They're like fuel trims.


No. But might want to numb BTS first to see where KR goes. Logging req AFR is hopefully accurate enough for that activity, and making sure IDC doesn't exceed 95.


Yep. This. But only after we have fuel where we want it.


Well, unfortunately, EGT and timing are interrelated, so we kinda need to numb one (possibly temporarily) so we can tune the other, then see where we end up after.

I made it to the second post here (http://nefariousmotorsports.com/forum/index.php?topic=141.0), and get that using BTS for adjusting fueling is wrong.  I've not waded through the following 29 pages of posts to see if there are dissenting viewpoints.  Took a look at the S4 tuning Wiki and it's not real clear there, seems to only refer to BTS when needing more fuel, not less.

How do you suggest using BTS to lean the AFR slightly?  It looks to be just past 4500 rpm where I need to make some adjustments.

TABGBTS. as dd suggested earlier

It was raised for the data I posted above.

You should add requested AFR to this plot. It seems real AFR is too rich between 4500 and 6250 RPM (everything richer than 11:1 is just waste of fuel).

EDIT: When you compare requested AFR with real values you can decide what should be tweaked - maybe it is not requesting that much fuel but it is just poorly tuned from factory?

Lambda calculated from AFR by dividing by 14.7

(http://www.myaudis4.com/tuning_images/nef12/nef12-h_afr_vs_req.png)

It is clearly BTS

Is raising TABGBTS to 800C the next step?

750 would be ideal.

So leave it where it's at now?  I set it to 750C here (http://nefariousmotorsports.com/forum/index.php?topic=6695.msg76759#msg76759).

If that's the case, do I start modifying the KFLBTS table to alter AFR after the EGT protection (above TABGBTS threshold) kicks in?

I do not like those plots with multiple runs through rev range as I find them unreadable.
Also I am not so sure about it being purely BTS related as BTS could be responsible only for 5000-5500 RPM range (from Nyet's plot) while enrichment starts at 4500 and still takes place up to 6250 RPM.
To be honest I would try something else - disable BTS (setting TABGBTS to really high values) and set constant 0.75 lambda (of course it is rich, but as you disable BTS it should be set to safe value) in LAMFA for >90% torque request. Then I would tweak KFKHFM to get real values as requested (and reiterate for lower loads limiting maximum load). Then I would re enable BTS or any other fuelling method you have chosen to use. That way you know your real AFR will follow your desired settings for every engine speed and load.

It could also be KR enrich. In any case, I'm not sure further changes make much sense at this point; we're getting good timing, and seeing a bit of correction up top.

I don't see anything wrong with letting BTS do its thing either.

We are out of injector though, so perhaps dialing back the boost might be prudent. Then again, we have just enough to get a bit of effective BTS (although not all of it), so maybe it doesn't matter.

I also think the safety features (KR/BTS enrichment) should be left working in tune normally used but for proper fuelling calibration they only add complexity and we are not sure if fuelling works as expected. That is why I suggested disabling BTS (as we are not sure if it is due to BTS or something else) but use useless normally pig rich LAMFA fuelling and do some logs to be able to find if KFKHFM is properly tuned in stock form. AFAIK KR is not used in stock form so it should not affect fuelling but to be sure we could disable that too. After calibrating fuel so it does what is exactly requested in software both should be reenabled/tuned to suit our needs. IMHO anything richer that lambda 0.75/AFR 11:1 should be avoided in normal driving areas - such high enrichment might be only set in BTS for really large modelled EGTs (it also could be set as ATR last resort security value).

I do not agree as it is running IMHO too rich now. So if we set fuelling properly there still will be a margin for KR/BTS/ATR effective enrichment.

EDIT: Having so good logging capabilities we may just log all lambda values to be sure what is affecting real fuelling and from that work out why it is so rich now. That way we could also calibrate fuelling without disabling any security functions. IMHO it is just important so real AFR will follow requested value (it does not matter if it is from LAMFA/KR/BTS/ATR). Next exercise would be to check how far is modelled EGT from real values...

EGT values are real, I'm using the RS6 EGT sensors.

Yea, i thought so.. Really great to see the logs reflect that modeled EGT is pretty damn accurate!

I know nobody wants to see 11 or lower AFR, but with EGTs climbing towards 800, it may be prudent, especially on pump gas on very low timing.

In any case, i wonder if we're seeing REAL EGT corrections from the RS6 sensors (not BTS but ATR). I've never logged a car equipped in this fashion, but it would explain why enrich seems to be happening sooner than expected.

If so, we are now out in the weeds and off the stage 1 reservation, unless we agree numbing ATR is relatively harmless for those running stock NB EGTs... and, incidentally, numbing ATR is one way of coding out NB EGTs which are notoriously unreliable and prone to failure.

Comments?

What could cause (high injector ms/ low DC) at lower rpms? My injector logs look so different then Flyboy

After 5750 rpm the desired AFR starts climbing, which looks to be what would be expected based upon the KFLBTS table.  Would setting the target AFR to 11.5 between 4500-5500 keep it from going as rich as it is?

Have you made more changes to your bin since you sent me one?

Just a bump in ldrxn mirroring yours and drop in DIMX due to the massive overshoot I was getting. Although my timing logs aren't as clean looking as yours either, my car is displaying more knock retard than yours and cam angle isn't showing a flat line with a pronounced change over either.  It's amazing how differently our cars have reacted to the same changes

I give you guys a lot of credit, refining these tunes is not easy, and I fear when my turbos decide to eat it and I go frankenturbo I'll be paying one of you guys for your time vs trying to build a stage 3 tune myself. My confidence is just not that high

So if modelled EGTs are close to real than IMHO BTS should be retuned as 800*C is just to early for such big intervention. The other thing which should be done is to try to readjust (I mean advance) timing and check what will happen.

I do not think so as ECU do not know it is using wideband EGT sensors IMHO (or I just missed the fact ATR was adjusted). Still 800*C is too early to drop AFRs below 11:1 IMHO.

If narrowband EGTs are so bad maybe it would be better for those who still have them to deactivate ATR but make BTS over reactive for really high temperatures (like setting to FBSTABGM to > 1). Still I would try to log all fueling request to find what is going on before taking decision what should be done. If in fact it is due to KFLBTS I would try setting it to more "sane" values and use FBSTABGM to alter enrichment according to EGT if you really need TABGBTS set to relatively low (and safe) value.

I think I agree


Not going to happen on pump gas if we are already at 11 and correcting.


Sounds reasonable to me.

1.  Some more minor adjustment to LDRXN, raised a little at the start to try and get requested to match actual.
2.  Bumped up timing a fair amount (KFZW and KFZW2)
3.  Increased TABGBTS to 800C.
4.  Ambients temps were a good bit cooler this evening 66F, versus the 76-82F it's been recently.

Here's AFR:

(http://www.myaudis4.com/tuning_images/nef13/nef-13_afr.png)

Here's the Nef-13 data log (http://www.myaudis4.com/tuning_images/nef13/me7log_20150707_Nef13.csv).

Welp. It's definitely enriching at 800C :)

You can numb BTS further ... or.... not sure where to go.

Me, I have BTS set to not go below 11.5.. but I hesitate to unilaterally recommend that.

So it seems FlyboyS4's changes dropped EGT a bit (due to advanced timing I suppose).
Flyboy, how is your FBSTABGM set now?
Perhaps the proper solution is to leave KFLBTS as it is but scale it by FBSTABGM (set to < 1 for 800-850 or even wider range of temperatures)?

EDIT:
It all depends on how you set your fuelling. If BTS is used purely as last resort component protection it could be left even at lambda 0.7 (0.75 might be better if ATR is still used, because otherwise ATR will be unable to do anything). In this situation if you use KR based fuelling it might be wise to set it to lambda ~0.8 for such high load and high engine speed areas (I would not set to anything higher than 0.85 for lower engine speeds and no ignition advance correction).

FBSTABGM ought to still be stock, at this point I don't know what it is, so we've not modified it.

I'm puzzled why BTS is intervening with the AFR to such a great extent now, but when it was set at the stock 499C level, like back with the Nef 4 revision (http://www.myaudis4.com/tuning_images/nef4/relogged/), it wasn't.  The EGT's were above 500C from the start of the pull.  Is it the lower engine load?  If that were the case it seems as though the exhaust gas temperature is not the primary concern for driving this behavior.

I understand it might be questionable to tell people to change FBSTABGM in Stage1 tune but as we already changed TABGBTS IMHO we should tune BTS to our needs fully.
So what is FBSTABGM? FBSTABGM is a map (4x1) which is used to scale KFLBTS enrichment according to modelled EGT temperature.
So you will find ECU requests lambda values set in KFLBTS only if for given modelled EGT FBSTABGM is set to 1. If its value is lower or value read from the map is lower due to interpolation requested lambda will be higher. If FBSTABGM values are > 1 requested lambda is lower.
For example:
ECU would not use KFLBTS enrichment up to 800*C despite having TABGBTS set lower. Between 800 and 850*C there will be slight enrichment depending on exact modelled EGT. At 850*C it would use exactly half of value in KFLBTS, so if for given load/engine speed KFLBTS has value of 0.8 (lambda) it will request 1-((1-0.80)*0.5) = 0.9. Between 850 and 900*C enrichment will be scaled between 0.5 and 1 (for 900*C it will use exact KFLBTS values) and above 900*C enrichment will be bigger than KFLBTS values (from 950*C it will be 1.25*KFLBTS value, so for KFLBTS value of 0.8 it would request 1-((1-0.80)*1.25) = 0.7).

That is why I opted to use leaner values in KFLBTS and enrich further basing on modelled EGT temperature as you found it is not far from real EGT. FBSTABGM is IMHO good to use this approach.

EDIT:
I am still asking if we are sure it is in fact BTS enrichment. If I were you I would double check ATR and KR enrichments.
Have you logged all lambda requested values? Could you plot them?

This.

They're all documented in the FR, the tuning wiki, and in my .ecu file which I have posted several times in this thread. Don't make me dig for all three yet again.

Also, obviously we need dlamatr and dlamatr2 to see if BTS is truly from BTS or ATR

What I have logged is contained in the data file linked above, I can add more variables if there are specific ones needed in addition to what I am already logging.

Which ones do you think would be appropriate, if we are looking to log BTS?

Looking at your old logs, compared to one of my logs, I cannot figure why for the same IAT, and MAP, your MAF reading is much lower.

EDIT: I wonder if the airflow difference traces back to the Bosch (yours) versus Hitachi (mine) MAF sensors?

Leaned about a half a dozen cells in KFLBTS, new AFR curve looks like:

(http://www.myaudis4.com/tuning_images/nef13/nef-14_afr.png)

couldbe the Bosch VS hitachi MAF...

I've spent some time looking at your log Nef 13 vs the last one I had taken comparing variables one by one and it seems that my results are more similar than previously thought. The auto range of ECUxPlot was making it weird for me.


I'm trolling the previous pages of this thread to incorporate  the last changes.

Side note, my car still uses stock EGT sensors if this info will help our current developments

You can load two log files in one graph, you know.

BTW please, in the future, can we put these digressions in a different thread?

Yes, I did know, I'm not sure what had confused me previously. And I apologize I'm just trying to keep up with the thread to keep my revisions current so I might be able to learn/contribute what I can. I think I'll keep the sidebars off to emails with flyboy.

again im sorry

Is it any different that before? It is still pig rich from 4500 RPM so your fuelling issue should not be due to KFLBTS but it might still be BTS related...
We really need all lambda requestes logged and plotted as for now we are tuning blindly IMHO. I am quite sure that after checking which lambda request is responsible for this situation further logging would be needed to find real cause of that issue.

It's about a half a point leaner than before, which is about what I would have expected given the minimal degree of change that I made.

Tell me the variables to log and I'll include them.


I don't think the discussion is a digression to the topic.  It's raising other considerations related to trying to modify a tune.  I hope we're aren't trying to make a tune that works only on my car, because that's a waste of time.

Which variable is lamda request from BTS/ATR?

lambts_w   ?

and lamfawkr_w  is for KR?

Leaned out the KFLBTS a little more.

(http://www.myaudis4.com/tuning_images/nef14/nef-14a_afr.png)

Have also:
1. Upped LDRXN a little more at the start.  Think this is about it for changes to this.
2. Altered DIMX pretty significantly.  Is this where I start modifying the Q2 term to control the overshoot?
3. Added some timing to lower engine load regions.
4. Logged the two variables I mentioned above, are those what should have been logged for the BTS work?
5. Accelerated in 6th to about 3000 rpm to see what would happen.

Nef-14a data file (http://www.myaudis4.com/tuning_images/nef14/me7log_20150709_nef-14a.csv)

No. IMX is still too high.

Bump

Thanks, I'll add those variables.

Regarding LDRXN, below 2000 rpm it's almost stock like, I'm wondering if the abrupt spike upward is causing the lower rpm part throttle drivability to suffer.  Part throttle acceleration is not particularly smooth.  I'm wondering if I raise LDRXN a bit more in the region below 2000 rpm if it might smooth out some of the abruptness?   

smooth is always good, but I doubt we're ever near LDRXN <2000rpm

Agreed, just tossing ideas around for how to get it a bit smoother.  I may give it a try to see what happens, but any other suggestions for routes to take that are more likely to work I am open to trying.

Here's the Nef-14a vs Nef-4 AFR's.

(http://www.myaudis4.com/tuning_images/nef14/vs_nef4/nef14a_vs_nef4_afr.png)


What's I find interesting is that the EGT curve is nearly the same for each tune, so apparently something else is also coming into play to drive the AFR rich with the most recent revisions.  I guess that would possibly be the higher engine load.  BTS variables have been added to my config file for the next time out.

Nef 4 vs Nef 14a additional charts (http://www.myaudis4.com/tuning_images/nef14/vs_nef4/)

Are those curves logged from your wideband?

Yes

Nef-14b data file (http://www.myaudis4.com/tuning_images/nef14/me7log_20150710_Nef14b.csv) includes:
1. Revert back to the Nef-12G DIMX profile.
2. Increase of LDRXN below 2000 rpm. - This change does seem to have smoothed out the part throttle acceleration at rpm's around 1800-2000.  Only drove a short time but that's the initial impression over the standard route I take.

Here are screen shots of the current settings:

(http://www.myaudis4.com/tuning_images/nef14/nef-14b_ldrxn.png)

(http://www.myaudis4.com/tuning_images/nef14/nef-14b_dimx.png)

(http://www.myaudis4.com/tuning_images/nef14/nef-14b_kfzw.png)

(http://www.myaudis4.com/tuning_images/nef14/nef-14b_kflbts.png)

TABGBTS is set to 800C.

Interesting. I don't see any ATR but BTS is triggering pretty low (725C).

i did a pull with these settings here are my results, it seems that I have a bit of timing being pulled to knock and still seem to have a severe boost spike with DIMX settings above stock

this is with the current community tune settings Jeff posted

And interestingly a lot of real EGT activity which isn't adding fuel via ATR

Also I max is high, even on flyboy's car, so it is no surprise you are overboosting even more.

yea, I will pull imax down uniformly. I did notice the difference in bts enrichment in his car vs mine. As of right now I'm running an mbox bin that has the MAF adjustments for a Bosch sensor but it is not the bin Daz posted in the beginning of the thread. I do have that bin and can compare its egt enrichment tables to mine .

Your DIMX is too high everyware. What is the baropressure when those logs are taken?? If it is low - much under usual - you can expect when baropressure raises your overboost is gonna be even higher and car gonna run leaner.
Turn down DIMX from 2700-5000rm you will gain some fuel IDC and overfueling is EGT protection. As suggested BTS upp to 800 and do LOG again.
Also i would add more fuel in spool.

My car is not overboosting with these settings, it's therealnap0le0n that is seeing the overboost. 

I have TABGBTS set to 800C.

What method do you recommend for adding fuel during spool?  Thus far we've addressed fueling via LAMFA and only just started to consider possibly using BTS.

IIRC your last few logs had significant overshoot.


Unfortunately BTS can't really be used during spool.. your only choice is LAMFA.

You do overshoot all over the place. And it leans your AFR. Use FKKVS for adding fuel before and under spool.
Look at your injectors  where in ms it comes in time. Then you tune FKKVS in those regions.

Ok.  Like I said above, I went back to the Nef-12G DIMX settings that got this feedback (http://nefariousmotorsports.com/forum/index.php?topic=6695.msg76624#msg76624), but I can resume making changes if needed.


Ok.  I'll be unable to do any updates for a couple of weeks but will plan some additional changes to implement shortly.

you're correct but I've been working on my tune as we go along and have managed the overshoot a bit with stock DIMX settings. I've been hesitant to post logs of those runs as to not clutter the thread with alternative settings from flyboy In order to really compare a "universal" community tune.

 I think flyboy will be taking some time for personal things so I'll try to continue where he left off and between the two cars come up with a happy medium. I'll take some time tonight to reread the boost PID section so I can tackle that,

now that this is getting really in-depth what are we considering "stage 1" and "stage 2" of the community tune? And are we going to take on a "stage 3" major hardware upgrades tune ie: turbo, injector, etc?

BTW I don't think we should raise TABGBTS any higher. May need to drop to 11.7-11.5 during spool all the way to redline... which means we will also be out of injector.

You really don't want high egts on those K03s, esp on a short pull like that. Row through the gears and .... :(

Look at the BTS maps... Load is obviously a big part of it...

What I'm not understanding is how the EGT's are now being characterized as high, but I was seeing the same EGT's at the same rpm's when the tune boosted to 11-12 psi.  It seems that having high EGT's are o.k. if the engine load is low, but if the engine load is high the same EGT's are bad.  I don't see where the distinction between engine loads alters the real world physical outcome, aside from the table being setup to respond differently for different engine loads which creates the different outcome.  I'm apparently missing something here.

EGT's are not high, they are  more than acceptable. Don't touch FKKVS, that's terrible advice.

IF we have room for more advance, EGT's will drop. If we don't, It is what it is.

Then we will have to modify the BTS maps, which I was hoping to avoid.

No it is not.
Bump fkkvs 1% in needed regions and its goona be perfect.
Fkkvs is the best map for small corrections.

Why touch fkkvs if actual is following requested just fine?

You should not be touching FKKVS on stock fueling (unless of course you use it to compensate for meth, but that's a whole new ballgame). If actual~requested, and you want enrichment, request it. What you're talking about is only suitable if actual!=requested. In all other applications, this is still questionable. Saying FKKVS is best for small corrections is shortsighted. Find where the error is coming from, fix it there or you very well may end up with a compounded error elsewhere.


You wouldn't. I can only assume he's getting hung up on his own (limited) experiences rather than practical application.

I'm connected via an emulator to a B6 with a GT3076 right now.

On page 35 there are wideband logs posted, but they are comparing revisions of the file. Not actual AFR vs requested.

I know that.
Thats why i said comapare it too his CSV posted not to each other.
Anyway, i am glad you see some sunshine :))

The topic here isn't to correlate actual with req AFR, but rather, trying to decide on what kind of EGTs are ok.

1% here or there isn't going to change EGTs any, especially if we are talking about moving requested AFR around by far more.

Ofcourse. That is correct.
I was focused on spool inrichmant. That has nothing to do with high EGTs.
Anyway. I did read the whole thread. Have you tried advancing timing??

we have a ton of timing advance over stock values especially up top

What's the goal of enriching AFR during spool in my case?

You mentioned this a few times now:



Where are you seeing the AFR being too lean?

Again, there is no reason. If anything we want to retard ignition a bit and lean it out to keep egt's up to aid in spool. This isn't needed for K03s, and we're straying off mission here-lol.

I'll defer to dd on the egt issue. I'm fine with numbing BTS to keep AFR from dipping too low if he is.

Other than that i'd say we're pretty close to done.

What would be nice is a dyno at this point. stock vs tune. same dyno, same day.

I see no problem with it. Best chance we have of making something fairly universal.

I'll try to address Imax Tom and based on the feel of the car I'd say it's good as a "universal community tune".

While reading the wiki i noticed this, could this be the reason flyboy has BTS enrichment and I do not?

I would aim at 12AFR at spool and raise it gradually to 12.5 all the way.  In a sence coming in cooler throu the rpms.
I have tried both lean spool and rich, and rich has worked better for me as lean cosed loosing Power.
The Point is - there is more ways than one doing this.

I dont see your AFR leaning out now, but my Point was as i asked:  What was baro pressure when you took these logs?? Medium for where you live, high or low?
I am at almoust sea level where middle baro p is 1000mbar. If i tune a car on low baro pressure ex 980 mbar, i have too consider that it will be Days that baro is 1020mbar. That will couse a lot of overboost if i tuned the car with some overboost at low baro pressure day. Bc the N75 takes the ATM p as a refference.
This will couse AFR the lean out of course.

Cheers

Mocke, presumably you understand how turbochargers operatePlease explain to me how running richer.
Agreed. Many different ways, but very few of them are correct.

I'm going to split all of this into a separate top

On a side, note could the reason Flyboy and I have different BTS enrichment be due to him having a Hitachi MAF and I have a Bosch, thus the EGT model being off?

Unless your load readings are drastically different (they shouldn't be), no

Not at all man. No egos or feelings hurt. We've spoken quite a bit, you know I have nothing against you.

The issue is this thread is to teach good practices. The advice you are giving is hovering somewhere between uninformed and dangerous. Not to mention it's off topic.

OK so to clarify the s4 wiki, the BTS offset for MAF scaling would be in the case of scaling a larger housing?

I have never seen BTS be affected by MAF scaling. Does anyone have the source of this?

I have seen MAF scaling completely wack timing tables, which affects EGT's, which then affects BTS, but I don't think that's what's being referenced here.

Nothing is of topic here. You are blowing things way out of proportion. My advice is nowhere near dangeous in any way.
You are free to delite my inputs and dont forgett a couple of more karisma minus.

MAF scaling will also move load around, which also changes the EGT model.

Plus obviously load is an input to every BTS table :)

If you are saying that I have smited you, You are incorrect. I don't smite anyone. Only applaud.

And like I said, not deleting anything, just moving the talk about spool characteristics and methods out of the stage2 thread. K03s already spool instantly, there is no need to discuss that here.

From the S4 wiki, maybe I am misunderstanding

Well as FR says gas for bts will be scaled by mass air flow (or calculated) if u r running wot. So miss scaled maf will affect bts badly

Just to close the loop on this, I've not been logging barometric pressure, in the past it looked to me that at idle the boost pressure desired was around barometric so I stopped logging baro.  I did about a dozen logs over the two week span so I would expect the barometric pressure did not remain unusually high or low consistently through that time period.

Here are the results of my latest revision, I lowered DIMX a ton, I've actually been massaging the lower ( 1750 - 3000 rpm ) DIMX regions up to improve spool a bit as I have eliminated my overshoot but am also not meeting request until later. Will log Tom morning and see how that goes.

Can you guys take a look at my timing logs from this as well and see if they are within acceptable levels?


Just to keep in mind, I live in FL so my elevation is very low and the air is VERY thick and muggy right now. This log was taken in the afternoon about 90 degrees F out.

Pull timing. Also one of your O2 sensors isn't looking so good.

Do you have a wideband?

OK thought so, and no I don't have a wideband unfortunately. I have my rear o2 coded out everything but the eskonf. I'll have to do that to use those bungs for wideband or could I just ziptie the o2 out of the way but still plugged in?

I'll check the finances but I think I can swing buying a wideband I know the innovate isn't too pricey

I'm puzzled by the very different numbers being shown in the Neftune5 data versus the Nef14b data that I logged.  Boost level is nearly the same, IAT's are close, yet MAF numbers and IDC are quite a bit higher (around 20%) for my car, the resultant acceleration curves are very similar.

It seems that one impact of the higher fueling is slightly greater but less varied timing curve.  If the MAF scaling is possibly the cause of the difference I do have an RS4 MAF that I could try on my car.

(http://www.myaudis4.com/tuning_images/nef14/neftune5_vs_nef14b.png)

Where do you live? Elevation maybe?

I don't think it is that, I'm at most driving around at two or three hundred feet.

I pulled timing in KFZW back to basically stock levels. another hot day about 95 degrees F

Replace that O2 dude

I know this thread is further from a basic stage 1 tune than I'm going to post here, so if you want to move this, go ahead.

here's the raw CSV from my first run after flashing my stage 1.

Critique's definitely welcomed, first timer and I followed flyboy's website for the basics.

I couldn't upload a file for some reason.. but here's a link

http://streetandtrack.ca/afterhours/tuning/s4/run2-jul26.csv

Thanks for jumping in and contributing.  So I understand where you are at, did you get up through the steps summarized in this page 22 post (http://nefariousmotorsports.com/forum/index.php?topic=6695.msg72750#msg72750), and this Stage 1 tune page (http://www.myaudis4.com/stage-1/)?

This would include:

• NMAX/NMAXF
• LAMFA/LAMFA Axis
• LDRXN
• KFZW2/KFZW

By the looks of the boost profile I'm guessing that's about where you stopped.  You may want to press on and start adding the changes that I put on the Stage 2 page (http://www.myaudis4.com/stage-2/).  For a bit there at the end of the Stage 1 phase and shift to Stage 2 I lost track of the sequence that additional tables were being changed.  There are probably a couple of tables, like KFMIOP, KFLDHBN and KFMLDMX, that were altered in the Stage 1 phase.

I need to compare notes with therealnap0le0n and make a post summarizing all of the changes made thus far.

Looking at your csv file it looks close to what I logged at that phase of development.  I noticed in the csv header that you have the Bosch MAF, and interestingly the MAF and IDC curves in your file are more similar to therealnap0le0n (Bosch MAF) than mine (Hitachi).  The only odd thing I noticed on first look through was some small intermittent spikes in both O2 sensor readings.  Nice job incorporating all the changes, nothing sticks out to me as being offtrack.

If you decide to press on feel free to contact me for help, I've noticed a few instances where the changes flyboy made didn't suit my car and having a 3rd vehicle will be great for a homogeneous tune as we will have more data points to compare.

I've been slammed at work but I'll see what my latest log shows and see about screen shots of my bin settings

I should note that my chassis is a B6, with a 2.7t swap.  It might change some values somewhere in the car profile of ecuxplot but I have not made those changes yet.

FlyBoy: that's exactly where I'm at as I wanted to get some real logs before pushing it further.  I have some "before" logs of 4 pulls I made on the stock modified M box file that daz provided on page 3 if those would be of interest.

nap0lean: I will be pushing forward with this, I still have all 4 cats and stock dp's but straight back after the cat's, which I don't think qualifies me to move to stage 2 just yet, so I'll be working on the stage 1 for a while.

I will make those changes on your stage2 page and get some readings.

*edit* I just had a look at the stage 2 changes and referenced my bin, I think I'm going to need nap0lean's input if those tables are related to MAF, considering my values are very very different from Flyboys who has 0's in lots of the cells.  I will keep reading as I have been doing and catch up on these changes and what they actually do, maybe it will all "click"

http://streetandtrack.ca/afterhours/tuning/s4/KFTARX_s2.png

On that Stage 2 page the KFTARX table to the left only shows the delta values from stock.  The zeroes are where I did not make any changes from stock.  Only the locations circled in blue were changed.  The table to the right shows the KFTARX values I ended up with after applying the adjustments from the left table to the stock values.  Your screen shot looks to be the unaltered stock values.

That's right, I did not change them, I understand now it's the changed values, I will proceed.  Thanks!

I made the changes to my tables last night based off the Stage 2 portion of your website Flyboy, I just need to finalize the installation of my AEM wideband and boost gauges before I go any further as safety precautions.

Is that as up to date as it is right now at 15psi and modifying those tables? 

I'm going to get some more logging done to make sure my runs are consistent as they stand.  I went out for a drive last night and intended to do some logging, but was happy enough just to drive it so I didn't grab anything more.

My summary on that page is lagging, there have been about a dozen tweaks made since then.  The closer you get to where we are currently at the more fluid the changes become as we are trying different things to see the results produced.

This post (http://nefariousmotorsports.com/forum/index.php?topic=6695.msg77016#msg77016) has the last changes I made to LDRXN (boost profile) and a few other tables.  The discussion has wandered a bit since then so there are a lot of posts but not much change.  You may want to leave TABGBTS at the stock level as that is an area of uncertainty presently.  I think therealnap0le0n has incorporated these changes, with the exception of a lower DIMX setting, closer to stock.

Yes the only deviation I have from Flyboy is DIMX is lower almost stock, and my timing tables are basicly stock as well. For what ever reason my car overboosted on initial ramp with extra DIMX and I had a ton of knock with altered timing.

As of right now my car decided to blow the lower control arm ball joint apart so its garaged for now, I bought a 2011 q5 as a daily two days ago. When I get the control arm in I can begin testing again

I'd like to try and understand why the MAF and IDC values are significantly different between my car and the other two presently active cars.  To bring our cars closer in setup I can swap back to the stock EGT sensors, and try a Bosch RS4 MAF.

I think I got caught up to date on my bin file with all the changes from the later portions of this thread.

I'm going to grab some more logs tonight after work of my current bin just for historic sake, and flash the new bin and grab some more.

So, I grabbed some logs with my new bin using all the values changed:

http://streetandtrack.ca/afterhours/tuning/s4/log1-aug4.csv

That's a huge boost spike, nap0lean is that what you were talking about?

I then re-flashed back to my old bin and took logs again:

http://streetandtrack.ca/afterhours/tuning/s4/log2-aug4.csv

What values did you use for LDRXN?  Your requested boost line is well below what I've got.  We've been pushing LDRXN up to try and get the requested line to match the actual boost.

(http://streetandtrack.ca/afterhours/tuning/s4/LDRXN_Aug4_a3.png)

This was one of the last things I modified because it was one of the last posts I found (I ended up going through the full thread and tracking down all changes and adjustments)

The most recent settings are:

116.58   133.64   155.16   180.43   191.79   197.68   198.05   191.79   185.00   175.01   170.70   163.97   155.16   146.91   142.43   131.93

Can you add B_tabgbts to the variables you are logging?

Can you keep the first message updated with the latest bin? It is impossible to track what is going on right now for people new to the thread.

Sure, I updated the first post with the bin that has the last changes.  I did notice the KFKHFM table is set to ones, which I don't see mention of in the thread, probably was a result of an off-forum discussion.

Thanks for posting it, I downloaded and modified the EGT and MAF sensor locations back to Stock and Bosch.  my car did NOT like that bin.

Started up fine, idled fine, I gave it a bit of gas to make sure it would respond nicely, and it stuttered and bucked a bit... took it around the block, same thing, 2nd gear a bit of gas to just get moving and it bucked HARD.  So I pulled back into the garage to reflash it back.

After that, I checked your latest LDRXN values and modified my latest file (the one with the big boost spikes that I posted) and reflashed and took some logs.  It feels smooth and strong (other than some weird issues where if I'm kind of half rolling it on and it doesn't bring boost on, N/A pull lol, need to do more logging and general driving though to begin to track this one down, happened 2 times in the span of 5 minutes of driving).


Here's a pull from the night, and I added that B_tabgbts to the logging.
http://streetandtrack.ca/afterhours/tuning/s4/log2-aug5.csv (http://streetandtrack.ca/afterhours/tuning/s4/log2-aug5.csv)

I noticed you don't have tabgm2 in your config.

In general your results look close to mine.  Intake temps were greater for you. 

I think I need to return the KFKHFM table I'm using to the stock values before I make any more judgments about how my setup compares to yours.  You may want to follow therealnap0le0n's suggestion for DIMX and see if that doesn't help with the initial overshoot.

I modified my config to include tabgm2, I guess it's safe to just keep enabling items and never disabling them with new bins / configs? is there an upper limit to the number of items it can log at a time?

As for my intake temps, I'm not running the stock airbox from the b6 a4.  There's a "ram air" filter attached to the corrugated intake piping and it's sitting where the piping stops and probably sucking in hot air from the rad.  I'll look into my options to get cooler air from below into the intake.

I've lowered my DIMX back to the original values from the Mbox bin provided.  I will flash it this morning and grab some logs and upload before I leave town for the weekend. For transparency, these are the values that were suggested previously:

(http://streetandtrack.ca/afterhours/tuning/s4/KFLDIMX-a3.png)

And these are the values from the stock MBOX that I'll be trying:

(http://streetandtrack.ca/afterhours/tuning/s4/KFLDIMX-a4.png)

My boost profile returned to what it was before, where the requested boost was higher than the actual boost with only a slight spike at the peak.  I'm going to split the difference across DIMX in changed cells and log again, unless I should try something else?

Here's my logs for today, not sure what happened in a4-log1-aug7 with the power and torque curves, I ended up enabling one of the speed values to be logged and it whacked the whole thing.  Boost profile looks nice and I'll work on raising the actual boost up.

Today was a bit cooler in the morning than the last evenings I tested here in Ottawa, Canada for me, at 20C(68F) and 58% humidity.

http://streetandtrack.ca/afterhours/tuning/s4/a4-log2-aug7.csv
http://streetandtrack.ca/afterhours/tuning/s4/a4-log2-aug7.csv

I try to be mindful about not logging a lot of variables unnecessarily.  Stuff in your file like the coolant temp, maf voltage, airflow over the throttle body, cpu load, I'd drop those unless a reason comes up to log them.  It looks to me that during your pulls the selected gear is not reading correctly, you'll start in 3, it'll suddenly shift to 2, then a while later back to 3.

I think trying to massage DIMX to get the actual boost where you need it is a good path for now.  I've still been trying to get my setting just right, but the boost curve is doing what it's intended to.  There seems to be a difference in how your two cars boost responds with the same settings I'm using.

If you want to email me the bin you are using I can check to see where the differences are at.

Hitachi vs Bosch? Not that it should make much difference.

I've emailed you the latest bin that was flashed to my car (pre-checksum fix), thanks for taking a look.

I'm going to adjust the DIMX values tonight or tomorrow night and re-flash my car to see what happens.

So I was looking at DIMX tonight, and compared it with your latest revision I had, and noticed that you had adjusted values in spots that I needed adjusting in to force more boost.

I ended up copying your values over to my bin and giving it a try, here's 2 runs:

http://streetandtrack.ca/afterhours/tuning/s4/a5-log1-aug12.csv
http://streetandtrack.ca/afterhours/tuning/s4/a5-log2-aug12.csv

Overall looking better, that spike isn't as high and it doesn't dip down as low afterwards.

Progress in a way.  I've been trying to chase down why there's a difference between the initial boost responses the three cars currently active in the tune have shown.  Looks like the N75 valve adjustment (http://www.myaudis4.com/2015/08/18/n75-valve-tuning-investigation/) may be the source.

Update: Adjusted DIMX with the new N75 setting.

(http://www.myaudis4.com/tuning_images/nef_n75-2/nef_n75-2_dimx_table.png)

and result attached.

Nice find!

BTW your I-limit still has a pretty odd shape compared to req boost.

Reposting so you know what I am referring to.

Thanks for confirming my hunch that the DIMX still needed some work.

I also noticed, though have not posted a chart of it, that my desired AFR is going richer than the other cars.  That seems strange because I have gone back to the S4 EGT sensors and returned the TABGBTS and KFLBTS to the stock values.

Nef_N75-2 data file (http://www.myaudis4.com/tuning_images/nef_n75-2/nef_n75-2.csv)

There are so many variables that go into the BTS fueling (and the EGT model) that I am not surprised. You'd have to flatten them all out (or log a lot more stuff) to nail down the root cause. Also, I don't think the enrichment you are seeing is from ATR anyway, just BTS from modeled EGT.

By the time ATR triggers, it is WAY too late...

BTW I don't see anything wrong with letting the stock BTS do it's thing. if you want, you can prevent it from going much below 0.7 by flattening the bottom of the BTS map, but I wouldn't recommend it. You can also numb DLBTS (set to 0 <=6%) if you want.

A couple of things I've noticed differing between the Hitachi and Bosch bins:

Variable/Hitachi Value/Bosch Value

CLAATR / 19 / 0
CLAATR2 / 19 / 0
CLAATS / 19 / 0
CLAATS2 / 19 / 0
TABGSS / 979 / 1229
TABGSS2 / 979 / 1229

Should they be different?

Dunno but I'm not surprised, there are a ton of other changes between the old bosch files and M box.

From the S4 tuning wiki those are all related to EGT sensors

Aside from DIMX, is there anything else you would advise improving with the tune?  If I accept that the AFR is where it should be, the timing also looks good.  I don't know if there's anything left to do.

Then this is not a stock bin.

I was comparing using the bins you posted on the third page.

I flashed those DIMX changes and I've grabbed some logs.  I will have more tonight and upload them, the boost spike is not so spikey anymore.

Here's results using an RS4 MAF and a further updated DIMX table.  I thought the last revision reduced the boost onset too much at the peak, it was slowing boost too much prior to hitting the desired boost level.  This update is a little more aggressive, but went too far, overshooting more than I'd like.  I'm making another change to try and reduce the overshoot.

Also, even with the Bosch (RS4) MAF I am seeing a good deal more g/s and higher IDC than the other two cars.  I'm stumped as to why that it.

I also noticed that with the stock TABGBTS the modeled EGT is above the threshold immediately.  Apparently around 3500 rpm the BTS fueling is kicking in.  I've reverted to the stock KFLBTS table, but my car is still requesting a richer afr than the other two.

Nef RS4 MAF data (http://www.myaudis4.com/tuning_images/nef_n75-2/nef-rs4-maf-2.csv)

Here's my latest log, with the original revised DIMX values.

Car feels very smooth and have not had any hiccups with it.

http://streetandtrack.ca/afterhours/tuning/s4/log1-aug23-a6.csv

These DIMX settings don't slow the initial boost rise as much.  I don't know that I can do much better with the boost onset if DIMX is all I have to work with.  Next I'd try and bring the mid and upper rpm boost level down to the desired line.

(http://www.myaudis4.com/tuning_images/nef_n75-2/nef-rs4-3_dimx.png)

(http://www.myaudis4.com/tuning_images/nef_n75-2/nef-rs4-3.png)

I'd say you're done with boost control.

Sounds good to me.

Fueling seems ok unless I want to start messing with the BTS table (I can't find the FBSTABGM in the XDF I've got), which I'd only do if I put the RS6 EGT's back in, which makes the tune work I do pointless as a community project at that point.

Timing also looks good to me, (I'll post up the rest of my data shortly).

Unless there's something I'm missing I think it's time to recap what's been done and call Stage 1/2 tune complete.

I'd say that's fair.

This has been a great thread. I'd like to applaud everyone involved. Did we come to the conclusion that the adjustment (to the flathead screw) of the N75 was the reason for the difference in boost profile @ spool-up between the cars?

Yes, the N75 hardware setting caused the difference.

I reallllllllyyyyyy

don't like that we touched the N75. I feel as though people are going to do that now.

lol.

It was very very far from a few different stock n75s he compared it to.

IMO the correct recommended action should be to purchase a brand new stock n75.

To the best of my knowledge neither of the two other car owners had changed the setting on their N75's.

The two devices I have are each set differently, and I know that is how they came from the factory.

I contemplated purchasing a new one, with the hope that it *might* be set like the other cars, but given the range of settings I have seen on these out of the box, I felt the best way to ensure my setting was equal to the other cars was to set it myself.

It would be interesting for you to purchase a new one and note how it is set compared to these others.

Can't know they are equal w/o testing the solenoid manually... there is a reason there is a screw - manufacturing variability of the valve as a whole (including spring).

I was thinking of *setting* to be what was needed to get an equivalent boost onset.

Here's a compilation of the tables I changed and values I used upon completion of this phase.
Boost rises to around 17-18 psi and tapers to 10-11 by 6500 rpm.  Target AFR is 12.5 and IDC topped out around 90%.

Looks good, but is there a reason to mess with KFMLDMX at all for a stage I?  I don't recall running into any issues with MAF over the limit in any of my stage I/II implementations.  Perhaps this was covered earlier in this thread.

It came up during a side discussion (http://nefariousmotorsports.com/forum/index.php?topic=6695.msg73125#msg73125).  I don't recall if it made a difference or not, but I left the change in there.

DMX should exceed your max measured MAF readings by approximately 30 percent.

Dyno results:

Nef community vs stock tune (http://www.myaudis4.com/2015/11/13/end-of-the-k03-journey/)