NefMoto

Hello,

Can any one post up a Wideband Tuner Pro tuned / stage 2 Lamfa table so I can have a little look / plagiarize. :)

Many thanks in advance

Barks

Don't forget KFLBTS... you will most likely end up in there most of the time on an S3...
(http://i84.photobucket.com/albums/k30/tufftybloke/Random/Car/S3/Tuning/fuelling_T111.jpg)

<tuffty/>

@tuffty: With what TABGBTS values ?  :)

Instead of tuning KFLBTS on every RPM, just tune FBSTABGM and TABGBTS correctly, and you will get much better fueling.
The fueling need physically depends on cylinder filling and EGT, not on RPM.
My TABGBTS is 700C, I do my main fueling via KFLAMKRL.

Also, I would only use LAMFA at higher RPM (>4000), and only at 90%+ pedal, else you screw up fuel economy quite bad.

Example:

Doesn't work for 91 oct, which wants a lot of enrichment at low load to prevent knock.

This is just your opinion.
You can also run less timing on 91 oct, and leaner mixture, without much loss in power, and without causing fuel economy issues.
Of course if you run timing that you can only run at rich mixture (which btw may not make more power than less timing at leaner mixture, because of flame front speed changes and effective timing), you run into det.
Easy to get caught up on timing, and get to 10.8 AFR with it, as you can run more and more, if you don't think about physical processes in the engine.

There is so much more to tuning than making such blunt statements...

Plus, with the existence of KFLAMKRL, you would never want to tune off of pedal anyway, except for high RPM gear changes.

I've tried both on the dyno. You lose significant amount of torque down low, and lose both timing and power up top w/o pre-emptive enrichment.

The only solution for 91 is watermeth.

Show me logs to the contrary, and i'll admit i'm wrong :)

Try KFLAMKRL instead of BTS fueling.
Yes, you lose some torque. Hell, you are always losing torque when you are running leaner than 13.3.

It's a tradeoff, economy vs torque. I often go 100% pedal before I make 7 psi. If I enriched every time my fuel economy would suck donkey balls.

The context tune wise here is a 1.8t 20v on an S3... in barks instance its a K04 hybrid and in mine a GT3071r

@professor...
my TABGBTS value is 849 from factory of 819... main reason is I wanted to get fuelling sorted initially with as little intervention as poss... I log the egts on most runs anyway but I have more development work to do on the mapping and not had the time yet..

@prj...
I can see where you are going with that but at my power level I would prefer to know I have the fuel there to keep stuff happy rather than rely on egt enrichment models... fuel economy isn't high on my list atm but I may have a play with this especially your comment about LAMFA as part throttle off boost could do with a tidy up economy wise...

I see your point also with the fuelling requirements revolving around egts and cyl fill rather than rpm but I am working on the basis that my hardware config and appropriate fuelling request should keep EGTs at a safe level (preventive rather than reactive) so I am never running too lean as to encourage high EGTs and therefore not have to have the ECU intervene unless it gets to a more dangerous level... for the most part my EGTs have not exceeded 750/800 deg measured in the DP close to the V-Band fitting of the turbine housing... I will also be running WMI shortly

We have 98/99 octane here in the UK so we are a little better off fuelling wise there..

Interesting posts though and something I am going to have a look into... :)

<tuffty/>

You're not making that much power... baby turbos :)
But of course it is very possible to blow things up at every power level if you are not careful.

You don't see my point. Use KFLAMKRL for main enrichment. Load based.
Use EGT enrichment for what it is meant for. Set TABGBTS much lower, not 850C. At 900C+ is where the emergency should kick in. The initial EGT enrichment is not for emergency, it's for controlling cylinder temps.

Set base EGT enrichment to the same at all RPM's, use FBSTABGM to enrich *based* on the EGT. If you enrich via RPM's in the KFLBTS table like you had now, you will run too lean for higher gears and too rich in lower gears.

By just running a flat AFR via LAMFA you are having crap economy, and your car will run inconsistently, because you will be well into knock in higher gears if you don't leave margin and don't enrich based on EGT. Tuning is not only 3rd gear pulls or steady state. It is best to model the cylinder situation as best as possible in the ECU, and set as few fixed values as possible.

Bah, never mind I guess, do what you want :)

Thanks for all the data and information, the more infomation that is provided the more information is required, and so on..... sound like there are a few differnet ways to skin a cat

Thanks again

Mines a GT3071r (400+hp)... not exactly a baby turbo (for a 20v at least lol)...


I do see your point... and I have taken this on board... don't assume I am just doing third gear pulls though... I have access to a dyno and do road logging too... I have been trying an approach to keep most of the ME7 strategies as stock as is possible/safe while using a fuelling strategy under load/boost that I would consider safe as I find my way around the ME7...

I fully appreciate (and understand) what you are saying here though and will look into applying this to a file in the future as your approach is quite interesting...


Isn't that the point though? I am all up for experimenting with ways to do this... :)

TBH... I have seen far too many files in cars that the fuelling is left untouched... under boost running 14's and awaiting EGT protection to kick in which on a 210/225 K04 car is typically when the EGT probe has seen 920 + deg and gone into save the world by using all the fuel mode...

My approach is a legacy of my experience with standalone ECUs but will give yours a go as there is a certain appeal to this for stage1/2 levels of tune...

Appreciate you taking the time to reply... will keep you posted

<tuffty/>

Question re your approach on this and what you have in your head, spec wise as to what you are thinking this is appropriate for..

what size/spec power output and turbo are you doing this with...

you mention very specific rpm for example, >4krpm, but this is going to be very dependant on what hardware is on the car is'nt it...
Also, the controls and factory figures for lamfa and kflbts are very different between egt wideband cars and non egt wideband cars.... Significantly diffeent approach used, and non egt ones use kflbts in the factory settings..

your thoughts?

It is. But I guess we are used to different things :)
Yes, there is lots of bad tuning out there, but thats not any news.
My entire point is, that RPM does not factor into fueling requirements.

The amount of req fuel depends on the heat in the cylinder basically. You can measure that with LOAD, IAT and EGT on this ECU and map accordingly. Try to model the combustion chamber in the ECU from a performance standpoint as close as possible.
Most standalone ECU's are 10 years or more behind ME7.
Don't map ME7 like a standalone which only has a MAP vs RPM table for VE. Makes no sense at all.

I mention a specific RPM, because my main fueling is done off of KFLAMKRL which is actual LOAD based.
The only reason I use LAMFA is for enrichment when going through the gears.
4000 RPM is a pretty good bet on a 01E box with a rev limit of 7200 RPM. But the idea to determine this is just to see your RPM drop at your shift point in 1st gear, as that's where the biggest RPM drop usually occurs.

I also don't care about factory figures. From factory the car is tuned purely for emissions.
If you want to make power, ignore almost everything the factory does fueling wise.

You can get pretty good fueling on this ECU with like 8 cells of data. Or you can do mediocre fueling with 80+ cells.
The reason I say not to map based on RPM, is because the fueling need at the end of 3rd gear and at the end of 4th or 5th gear is going to be very different at 6000 rpm for example.

to put into context.. 1.8t ecu I am relating to, as is tuffty
Std Octavia/Golf wideband golf fueling... which is clear to see kflbts is not set to lambda 1's like egt sensor equipped cars are.
tabgbts on this is 400'c std

I ask whether the approach on these is as much ecu dependant as it is hardware dependant..  I cant find KFLAMKRL on any non-egt wideband damos/mappacks I have. Not saying its not going to be there, but I cant find it to comment on the 1.8t non egt cars. It is on egt car map pack/damos.

Coud you post up your lamfa, kflbts maps up to see from where yours is operating from?  many thanks.
(http://www.badger-5.com/bin/mapping/032HJ-lambda-std.JPG)

I think, if I understand this correctly, that this is the key bit here.. (for me at least) Measure... IAT, EGT, yep I get.. Measure Load tho? (from ecu? a calculated value)

non egt cars cannot be tuned in this way is my suggestion.. No way would you want to rely on downstream lambda heater temps for tuning against egt's.

Where a GT3071 is "small" to you, what level of power are you tuning via egt's etc for?  Tune to egt's I get, and understand is whats done on alchy cars, muchos bhp engines etc, but in context of 1.8t (me and tuffty) I am trying to see if the context and methodology still applies.

thx

This is basically what I am doing, but I start the enrichment at slightly lower loads than most people suggest for 91oct.

Also, I am not currently using LAMFA fueling.

My mpg is around 18 city, 25 highway

Just because someone doesn't agree with you DOESN'T mean they don't understand your point.



Now this I don't see your point. KFLBTS has a Act Load axis. So, it doesn't just enrich via RPM only. Factor in the load axis and the map can handle low verse high gears.

When my LAMFA had a 0 col like that my LAMFA AFR request went back to 1 @ 100% req torque. That was my experience. I had to change my first col to > 0 to get LAMFA to work right @ 100% req torque.

cause your lamfa was not difinded correctly .
->http://nefariousmotorsports.com/forum/index.php?topic=1272.msg11844#msg11844
 (http://nefariousmotorsports.com/forum/index.php?topic=1272.msg11844#msg11844)

http://nefariousmotorsports.com/forum/index.php?topic=1272.msg11967#msg11967 (http://nefariousmotorsports.com/forum/index.php?topic=1272.msg11967#msg11967)

here you'll find some thaughts and tests I donne on an 1.8T 210 hp S3 AMK .

Interesting... changed LAMFA on an Leon Cupra R project I have.. set it to the following from page 3 of the thread you linked too(1ML906032A  -> LAMFA: 0x81C98A, X_AXIS: 0x81C97E, Y_AXIS: 0x81C96E)

Swapped the X axis from 16bit HiLo to 16bit LoHi and get a load axis from 50% to 100%... you are saying this is now correct and the 0% is wrong?

Anyone else care to shed some light on this? I also checked a G box file (8D0907551G  -> LAMFA: 0x81C390, X_AXIS: 0x81C384, Y_AXIS: 0x81C374) and did similar to pull LAMFA in to this revised layout....

Must say I am a little confused about this now... fuelling has always been right on mine before this revelation... :(

<tuffty/>

gbox damos LAMFA is completely wrong.

1) the damos address off by a byte (the 16bit axis values are not aligned to a 16bit boundary!)
2) the binary axis data itself is scaled wrong... it should be % (50-100) not .5-1.0

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

http://nefariousmotorsports.com/forum/index.php/topic,141.msg9068.html#msg9068

There was a member on here (Argdub?) who posted a long list of lamfa maps with corresponding x and y axis for many files. I believe it was pretty accurate, but I can't for the life of me remember where it went...

Edit: http://nefariousmotorsports.com/forum/index.php?topic=1272.30

TheDSI actually already linked the thread above, but here's the direct page for lamfa addresses...

Just something quick to add to this...I've heard said several times in this thread that fueling is not RPM dependent. This is generally not true. In most occasions for a given load value, your lambda value should be getting richer as RPM increases. I think maybe the misconception comes from thinking of the combustion in the cylinder as "engine domain/engine synchronous" (perhaps this may be partially true at low speeds where the turbo is operating in the "pulse" regime, but not at higher speeds when it is considered to be in "pressure manifold" regime.)  In addition to considering combustion events in cylinder, you must remember that as RPM is increasing the turbo remains the same device (obvious, but important to consider in this case). In turbo applications, the lambda value for maximum output is dictated by the maximum allowable turbine temperature. As RPM rises, even if you are holding cylinder filling constant, the time averaged mass flow across the turbine is increasing. (In simple terms: the engine is refilling its charge roughly the same amount on each intake stroke, but the turbine size is of course not growing) This will generally lead to a higher pre-turbine pressure and as they are linked by gas laws, the temperature as well. In summary: for a given filling where timing is operated at approximately MBT, mixture must be enriched to decrease the pre-turbine temperature to below its operating maximum as RPM increases.

A lot of people spend time saying ".85 for max power" or whatever, but you have to realize the driving force of that. It is largely not in-cylinder combustion (for this case, it is generally .93-.95, or 1 if you have perfect mixture and flame propogation). The .85 becomes necessary at full load mostly because of turbine temperature. Increasing the fueling allows you to increase the load, and the increased load that is allowed by .85 or richer mixtures is what allows you to make a torque increase over what it would be at .93-.95.

LAMFA the first row is not 50 its 0 and is used on idle. It dont matter what its reads in the damos .KP or ols, 0 or 50 as long as your not changing the axis scale. If your going to rescale it then you need to correct the axis address to its reads correct.

Nick

Load is cylinder filling. Not understanding this, is not understanding what Motronic is about.
ME7 has an EGT model. If you don't have a properly calibrated EGT model, then you can't use EGT to fuel off. When the EGT model is properly calibrated, it is usually pretty damn accurate.
Let's just say, GT30R+, but this is irrelevant, I was saying it tongue in cheek. You'd have to screw up pretty bad to blow up an engine on ME7 anyway. 200hp/litre is baby power for it.

Nope. My point is you can have 700 EGT at 6000 rpm in 3rd, or 900 in 6th at same load. Is the fueling need equal? Please do tell.
If you map via RPM and LOAD as done on all old ECU's, you must run not enough timing and too rich in low gears, so that high gears don't knock. That's why I said tuning is not steady state, and tuning is not 3rd gear pulls.

You basically just said exactly what I have been saying. Except RPM does not factor into the equation.
The correct measurement is combustion chamber temperature and exhaust gas temperature. The higher the combustion chamber temperature gets, the lower the knock threshold. The lower the threshold, the more timing is retarded. The more timing is retarded, the more EGT rises, as more energy is converted into heat. The fact that it occurs with higher RPM is logical, but when you have the ability to fuel off of combustion chamber temperature directly, it is much better than mapping based off of RPM, which is just a side effect of things.
The reason older ECU's are mapped via RPM, is because they don't have an EGT model, they don't have EGT measurements, there is no way to model the combustion chamber.

But the actual req fuel for best power constantly drifts, and if timing is advanced sufficiently, then it is almost directly proportional to EGT.

It's good we have discussions like this though. Perhaps it will remind people, that the ability to change a few numbers in the ECU and the ability to understand the physical processes inside an engine, and understanding *why* every change is made are completely different things.

The best way I have found to tune the 1.8T me7.5 is to use LAMFA as main fuel map and use BTS as an additive when under extream conditions.   Its take a little time to ajust BTS and egt threshold as its only a simlated EGTs. I have tuned this way on stage 1 all the way upto 500+bhp and not had one issue or unhappy customer.

Remember there is not right or wrong way to tune and engine/ecu.... just a method thats works and is all safe.

I dont like the idea of lambda via knock, I fuel to optimise combustion and keep temps in check to stop knock and elevated EGTS..



Nick

These are retail customer cars so the tuning is aimed more to be safe under all conditions.



I can see what your saying about cylinder filling/load boost v rpm, its like tuning OMEX ecu via MAP and TPS.

But when me7.5 tuned via LAMFA and BTS v rpm, for some reason its works 100% fine at all stages of tune.  The main thing is getting the simulated EGTs corrected and really fully understanding the hardware.


Nick

Combustion chamber temperature and EGT are heavily driven by engine speed. For a given load, as engine speed increases the time between combustion events decreases and so does the time for the cylinder to transfer heat to the head/block/etc. This is kind of the idea i was talking about before where we must not separate our considerations for "engine time" (crank angle) and the "physical time" at which chemical reactions and heat transfer occur. Because of this heavy relationship between RPM and EGT/combustion chamber temp, it makes sense that a model and mapping strategy would be based on both load and RPM where EGT is the driving factor (ie you are using these things: speed, load, thermal mass, lambda, efficiency, etc to predict EGT because it is heavily dependent on them).

You are correct though that if you can map directly off of EGT, that is the way to go as that is what all these load/speed/etc based strategies are trying to predict. It sounded like you were saying there was no physical relationship between the two, but if I understand you correctly now you are actually saying that the influence doesn't matter because you have direct access to the EGT measurement which is the critical measurement here. With that I can certainly agree...and also wish that all 1.8T's came with an EGT sensor.  :-[

Exactly! This is what I have been trying to say from the start :)
ME7 has a simulated EGT model. This simulated EGT model when calibrated correctly off of a real EGT sensor can be made to work very well without any actual sensor.

There are a million ways to tune things - I just like to model the engine as close as possible.
And the only reason I say that fixing AFR based on LOAD and RPM is incorrect, is because there are better ways available in ME7, and because it is impossible to work out the correct fueling based on LOAD and RPM alone for every situation and every gear.
You can choose a "golden middle" - and that's how almost all the older Motronic cars are mapped as well, but it will never be as good as it could be.

Based on your statements about LAMBTS here I think you didn't fully understand how this is ment to work.

Also there are a lot of situations were you want pre-emptive enrichment to stabilize the combustion just like nyet uses it.

And I think the only point of this statement, as you said it was to cause personal offense.
You might as well say "I think you are an idiot", well, that would be just your opinion ;)

If you think I misunderstand something - then do tell exactly why, and exactly how you picture it. That would be called "constructive criticism". What you are doing now is called "mud slinging".

I agree, and I do this as well.
FYI I don't fuel solely off of LAMBTS, I use it for additive enrichment when things start to get hot.

Don't know why you take it this way but I have no such intentions whatsoever.



So in other words you say that the enrichment to reach a certain EGT is the same on all RPMs. This is obviously not the case like already described a few times here.

KFLBTS (together with dlambts) is the enrichment you need in steady state to limit the EGTs to a certain goal. All the dynamic stuff you described is handled with the enable conditions and the EGT modelling. This is how the function is intended to work when you use it only for component protection.

Bosch component protection is not a controller like other ECU systems that control lambda to reach a certain limit, it's all based on precontrolled lambda 

Because of the way you stated it, without providing any explanation on where you drew that conclusion. Because you are playing the player instead of playing the ball.
I never said this. I don't use BTS as main fueling, I don't use EGT as the only fueling either.
A lot easier to achieve nearly the same behavior with FBSTABGM, where you add a certain amount of enrichment based on current calculated EGT.
Of course the theoretical best way would be to have linearized KFLBTS and FBSTABGM together.

Btw, in stock tunes on RS4 FBSTABGM is not used.
Are you going to declare it useless, and "not meant to be used" as well? ;)

My point is to fuel constantly to preemptively decrease EGT. Not tune based on emissions, and then dump fuel when you reach a certain EGT setpoint, and then set lambda rich enough to hold a certain EGT limit.
If you tune for performance, not for emissions, it's just not the way to do things IMO, so obviously you end up using some things "not as intended".

Not to derail the discussion, but I dont have the FBSTABGM map defined in my m box XDF   ???  .  Could someone please post the info so I can have the map and follow along with you guys.  Sorry...

0x150C2, factor 0.000031, 16 bit, LoHi, 4x1
The axis is right before it at 0x150BA, it is EGT, 16 bit, LoHi, 4x1. Factor 0.019531, offset -50

Make sure you look exactly at FR how it is used. It is a multiplier for enrichment. That means for (1-KFLBTS+LAMBTSZW).

Thanks. I'll add it to my next mappack release

Another question if I may..
EGT model.. and hardware effects on it... eg: higher flowing exhaust manifold, bigger turbos K04 vs GT35, very different scenarios.  do you say the vag egt model is equally valid between very different hardware setups?

I am trying to understand more about the me7 here.. please bare with me. thx

This is where understanding the hardware plays a huge part, If you know the turbo setup is going to be prone to raised EGTs and you want BTS to play a big part to keep the engine safe, then you will need to recalibrate the simulated egt maps buy using and external EGT system and lots of logging and graphing data.   

Or you can just offset TABGBTS to a temp you know and have recorded that works.  This is the way I work, its handy when you have a braked dyno and you can load the turbo upto the max and monitor the elevated egts under the most extream conditions.

With Me7.5 tuning you can try to go too much into detail thinking you have to do it right or a way eveyone says.  IMO just find your own method of tuning me7 which works and is safe.

Sure you did. You said
which results in excatly what I have said.


What you are trying to create here is a P controler. This will not work properly.

Yes because that's what it is.
That's one of the reasons why it is not in later systems like MED9.
You would not believe how much useless stuff is in a current software, that was used to fix a specific problem ages ago but is not used any more.

I'm not sure why you think that emissions have to do anything with that.
Even on a FTP75 vehicles like a RS4 will not exceed 3500 RPM or more that half load.
LAMBTS is for component protection were it is needed nothing more. That's what it should be used for. For other fueling stuff LAMFAW has enough options to suit all needs

If you want a accurate EGT model you have to calibrate it for you specific hardware. Backpressure, ignition calibration s.o. all have a big influence.

No, it does not result in what you have said.
I quote:
I don't know where you are reading this out from, but it's not the case at all. You assume I want to reach a certain EGT with LAMBTS, which I don't. Also, there is ATR.
And why exactly? I've logged different gears on the local airfield with this approach, and it behaves exactly the way I want to.
So tell us how exactly you are going to use LAMFAW to run main fueling on a performance oriented car.
And I mean tuning an existing software when you can't compile what is needed from source.

Btw, I fully agree with this.
In the end, as long as the car is fast, good on fuel, has good performance, and things do not overheat on prolonged WOT usage, you have been successful.

cheers nik.
I am happy with how I choose to tune, lamfa, bts etc... It works.. and as you say, loaded eddy brake dyno and logging shows whats required.
This thread direction and prj's approach did make me ask the question.. re another approach/methodology etc..

For alchy burning cars, drag cars, mega power, I believe they tune to egt's.......
Interesting thread tho.. 
thx

yep, Bingo.. thats my point I was eluding to. thx for confirming.

who actually does this, and does it make for a "better" job?  I am sceptical.

I don’t get it. If you don’t want to limit the EGTs to a certain threshold why do you enrich by EGTs then? On one hand you said that using LAMFA on low RPMs is not good because it hurt’s fuel economy and on the other hand you enrich by EGTs when it is not needed  ???


That was under the assumption that you want to limit the EGTs with it.

What exactly is a performance orientated car? In your opinion a stock RS4 falls not into this classification then?
Stock cars are calibrated to meet their specified power goal most efficiently (use more load and leaner mixture instead of running richer and use less load). That’s why all modern Audi/VW turbo engines only enrich from lambda 1 based on component protection (and where it’s needed for other safety reasons or to stabilize combustion).
Other concepts like the supercharged B8 S4 for example have to use full load enrichment by a functionality similar to LAMFA additionally to component protection, because otherwise they will not reach the aimed power on lower RPMs.
Now we want most power possible out of the engines. Strategy is still the same. Use the protection maps for protection (LAMBTS, LAMKR) and use LAMFA for best power enrichment when high load is demanded. Simple as that.
I also don’t understand why you are against using LAMFA on low RPMs. What do you want to achieve when you smash the pedal on low RPMs? Win the fuel economy award or accelerate as fast as possible?

ATR can limit EGT's to an exact number. LAMBTS mixture output will be similar regardless if EGT is 700C or 900C (unless timing retard significantly goes up).
If I enrich enough via LAMFAW to be at the mixture that is giving me best torque, then enriching too much too early with LAMBTS will cost me performance, where it might not be needed yet for component protection.
Plus, unless I have a dyno and can load the turbo very high at each RPM, then filling it out KFLBTS based on 3rd gear logs is pretty meaningless.
Stock RS4 is detuned quite a bit.
My point is that when you don't leave a huge margin, you can't stabilize EGT's with LAMBTS, as eventually they will still get too high and ATR will kick in.
I perfectly know that remapping my RS4 to run 22 taper 20 psi means I am going to have to reduce boost to meet my EGT goals by the time I'm in 6th gear.
Whereas if I had more margin, I would not have to do this.
I don't like fueling based on requested load. I am not saying it doesn't work, but until the turbos have spooled up, and at lower RPM's in lower gears even if the driver is mashing the pedal, the richer mixture is not going to do much besides reduce fuel efficiency.
Maybe I am too used to the other ECU's, but usually fueling is done based on actual load. And transient enrichment is a different thing.

For load based fueling I use KFLAMKR, KFLAMKRL. I use LAMFA to enrich at higher RPM's at WOT, so that I get better fueling during gear changes, but the requested lambda I have in LAMFA is higher than the combination of KFLAMKR and KFLAMKRL.
I use LAMBTS from 700 with FBSTABGM as additional enrichment. I don't enrich via EGT's when "it's not needed". I only reach in excess of 700 calculated EGT at 6000+ RPM in a 3rd gear pull from low revs.

When logging, everything works as expected.

Like I said, you haven’t fully understood what LAMBTS is about. Have you ever looked up what comes out of KFLAMBTS with stock rlmax values? Almost nothing. That’s why TABGBTS is so low on S4/RS4. It’s only a small precontrol to help ATR not to overshoot too much. The massive EGT sensors are shit for dynamic situations so you need some sort of precontrol. That’s one of the reasons (beside being expensive and prone to failure) they have disappeared. On vehicles with no EGT sensor TABGBTS is significantly higher.


See above


What’s your thing with third gear logs? If someone uses max 3rd or 4th gear in the critical region what’s the problem then? It simply has to be calibrated with the intended use in mind. I live in Germany and my LAMBTS is calibrated till top of 6th. 


Stock RS4 is calibrated to meet the specified power goal. That is not only under norm conditions but also up to a certain altitude at a certain temperature. At high altitude with high temperatures the turbos are very near to be maxed out with stock power level.


Not if calibrated accordingly.


Mmh what fuel do you use and what’s your EGT goal? That behavior seems strange.


When I step on the pedal I want to go all in because otherwise I would not have to.
That has nothing to do with transient enrichment which is ESUK.



Beginning enrichment based on ETGs from 700°C on is dumping fuel for me, because it’s absolutely not necessary.

When you are happy with your outcome for your needs then perfect.

So you say that the fueling need in 3rd and 6th gear is equal for a given RPM?
Because if you just map KFLBTS based on RPM, it's output will be the same any time the threshold is reached. Whether you are 1 degree or 200 degrees over the threshold.
Also, it is pretty damn hard to get to ATR (950+) in a properly calibrated RS4. Even in stock form you have to try hard to reach that high EGT. I also don't think going that high is healthy for components.

Yes, exactly, because stock the car is calibrated to always make the same power, and not make the best power possible at a given moment.

There is a difference between going "all-in" and pure waste. In a higher gear at lower RPM you may want as much as 0.85 and below, whereas in a lower gear, there is a high chance nothing is spooled up yet at all. Yet for the driver it is instinctive to give more throttle. I guess this is more a personal difference of how people modulate throttle input.
It does not dump fuel at 700C. It starts very slightly enriching to be able to maintain ignition angle. It starts dumping fuel at 900, due to my FBSTABGM curve.
And after that ATR takes over...

My point is basically, that if you run the same lambda, then ignition timing at the same RPM is dependant on EGT, so you would get more correction in higher gears at the same RPM.
You can either reduce timing based on knock which has already occurred or you can pre-emptively enrich the mixture to not get to knock.
In both cases you lose some torque.

Take the FR and read the application hints for LAMBTS again.


No problem to do that here.
S4/RS4 are 980°C concepts, but I agree that running cooler is better.


But what does that have to do with downtuning?


If I request full load I want all the engine can give on the current operation point.
That has nothing to do with gear or turbo spool.


With the basically not existing valve overlap on S4/RS4 engines you more likely see an IAT influence going through the gears than EGTs.
At least I have never seen anything like this.   

So here, you get my point. Running cooler is better.
I don't want to reach the measuring point of the narrow range EGT sensors, I want to enrich more earlier, to run cooler, to run more timing, etc.
And that is why I use KFLBTS not as intended. If you have a better idea how to accomplish my goal, do let me know.
I understand that KFLBTS is meant for steady state to pre-control the temperatures, but unless you have a steady-state dyno to dial this in, you will not get it perfectly accurate.
Of course living in a world, where you do this for living, a such thing might not occur to you. Last I checked this site was still aimed at both the professional and the enthusiast :)
The fact that most cars will not reach those conditions. The fact that there is more to extract in 90% of conditions.
Interesting. So you think going richer than 12.5 afr is always a bad idea? Do you think 12.5 afr is always needed at 100% pedal?
My RS4's engine makes the most torque around 0.8 lambda on 98 RON, when the turbos are fully spooled up. However, going that rich at start of spoolup is dumping too much fuel.
Also, mapping the engine based on pedal, creates an on-off type behavior, where at 70% pedal you are running stoich, and then at 100% pedal suddenly quite rich.
This also creates a problem with the timing maps. Because timing is done based on actual load. If you dial in the timing advance based on WOT spoolup with the richer mixture, then when you are running 70% pedal, you will be well into knock at stoich mixture.
Granted, I have not tuned many S4/RS4 engines. But from my experience with other engines (and you could say I have tuned a few...), this statement is incorrect.
What I find is given the same ignition advance and the same lambda, 100-150C differences in EGT can affect the spark advance 1.5-2 degrees. In some cases more. Even when the IAT's do not have a significant rise.
And this has been true so far for many different engines. Different cam setups. etc.
How a RS4/S4 engine would magically break this observation is really weird. And from my point of view - increased EGT means that everything in the cylinder gets hotter, and it's pretty normal that this has a detrimental effect on spark advance.
So I would say there is a pretty significant difference between for example 800 and 950C EGT.

prj,

the biggest problem you have is you are trying to tune based on an EGT model.  This model is developed primarily for cat heating purposes.  The steady state, wot is simply taken from a map.  This map isn't particularly accurate, even if recalibrated, in all states.  Fuelling on req load is actually a god send.  The inrush of fuel cools and prevents knock as actual load builds. 

The main thing is you are happy with how your car drives :)

Rick

And how do you correctly calibrate timing then?

I already said that my main fueling is not coming from LAMBTS, multiple times. I just use it for additional enrichment, to drift between about 0.83 and 0.77 lambda. You seem to ignore this though.
I use LAMFA, KFLAMKR, KFLAMKRL for main fueling.

I am talking mostly about the "slow-spool" zone here when I say that fueling based on requested load is not a good idea. Not about fueling over the RPM threshold where your turbos basically insta-spool on WOT, for which LAMFA is perfect.

A question for you knowledgable folks...
Sample below, wideband 1.8t, polo.
Log from 4th gear pull on dyno
(http://www.badger-5.com/bin/logs/polo-wideband-lambda-temps-log.JPG)
Sample ols screen of its settings which those logged numbers came from....
(http://www.badger-5.com/bin/logs/polo-wideband-lambda-temps-b5t2.JPG)

From the logs you can see it fuel dump severely (its so much as to cause a misfire), but lamfa, kflbts are not asking for this intervention...
FBSTABGM is shown, although axis name looks wrong.. clearly not km/h... but I think the data's correct(?)
TABGBTS is set to its factory setting 399'c... no egt sensor on this car, but I am assuming ecu is getting a temp measurement from lambda which I logged to see if it correlated to anything.  Wideband lambda is std on this car.

My question is whats intervening to drop the lambda request so low...?
When logging 001 lambda adjust (in vagcom) the % adjust drops to zero when this intervention kicks in (much like it does when hard limit 920'c on 225 cars do when egt protection kicks in)

Set TABGBTS higher perhaps?
I am not convinced FBSTABGM is showing correctly....
another 150/180bhp 1.8t shown below and different numbers again.
(http://www.badger-5.com/bin/logs/06A906032HJ-FBSTABGM.JPG)

Why is this additional lambda request coming in...

Make proper logs with setzi's tool and you will /know/ instead of guessing.
FBSTABGM looks just wrong to me, KFLBTS in this setup looks almost useless. The 0.5 will do nothing in most conditions because LAMFA is richer.

Why disable LAMBTSZW completely? It is very effective when tuned right.
LAMFA at 2000-4000 RPM is pretty useless, it will not really do anything. Depending on the turbo size, go all out on LAMFA after the spool is instant. in 2nd gear WOT.
If you want to enrich at lower RPM's, enrich based on actual load via KFLAMKR and KFLAMKRL. So that you have adequate enrichment without dumping fuel in the slow spool zone, and use LAMFA in instant spool zone...

Use KFLBTS properly for component protection, don't make it always active...

Before I get flamed to high hell again - this is just how I would do things, it's not the only way to do things, it might not be the same others do, etc.

FBSTAGM should be degrees c not rpm

I see your rpm goes upto to 7600 but the axis are still 6520.  Would rescaling the axis not help and if it doesn't would give a better spread, I assume if these aren't done then the axis for timing etc aren't done?

I've just checked an HN and HJ and they both have tabgtbs set to 400 as well. I know on non EGT it's all simulated EGTs.

So on ME7.5 where axles are different from ME7 should rescale and do equal changes or it works different way?

enrich based on ign retard?

just a quick update.. on polo TL... I raised the TABGBTS up to 650 from 399 and lamfa worked as told with KFLBTS coming in only after 7500rpm
lamfa was'nt set to the value I posted earlier.. as it was obviously not correct..  Raising the temp threshold seems to for whatever reason stopped the fuel dumping.

313bhp/290lbft 19psi boost (its running an IHI on 550's running 4bar)
intercooler sux tho, 10psi pressure drop which I now need to fix. mapping tho went very straightforward thus far.

You can, you can also enrich based on actual load if you map top row, as that is always active.

Hello where can I find kflamkr and kflamkrl on a S3 1.8t AMK/bam map?
Thank you