NefMoto

Hi everybody,

I have started a complete (not yet tho lol) guide for understanding the M3.8.X POS as a lot of information is spread everywhere, and for someone like me that a few months ago had absolutely no clue how this worked, it's been a bit of a nightmare to get started with this ecu. Link below:

http://nefariousmotorsports.com/forum/index.php?topic=20919.0 (http://nefariousmotorsports.com/forum/index.php?topic=20919.0) - As you can see in the post, the ecu and software number i pretend to start from are 8D0907558E - 0261204805

I've been gathering information about how to start with a simple Stage 1 tune, and from several sources have found different lists of maps that should be modified. I'm looking for a good list to have a startpoint in order to compare 150 with 180 oem files, other stage 1 files that fly through the internet, and start learning which changes should be made and the correct methods and order to execute them.

So here are the lists and I'm asking if people with experience can validate the correct maps that should be considered to study for the most correct Stage 1 tune. I just need the best startpoint possible and I'll keep on from there.

Thank you very much in advance.

http://nefariousmotorsports.com/forum/index.php?topic=2013.0

KFLF
KFTLWS
Drehzahlbegrenzer
LDSMXN.0
KFLDS
KFZW.0 (.1, .2)
KFLUL
KFLDSAK.0
KFMLDMX
KFLDTV

https://www.youtube.com/watch?v=raAmPkPr_8w

KFTLWS
KFLF
TLRAN.0 (.1)
KFZW.0 (.1, .2)
KFMLDMX
KFLDS
KFLDSAK.0
KFLDTV
KFTVLDRE.0
KFLUL
LDSMXN.0

Map Pack, attached, found on another website

KFLF
KFTLWS.0 (.1)
LDSMXN.0
KFLDS.0 (.1)
KFZW.0 (.1, .2)
KFLUL.0 (2.0)
KFLDSAK.0
KFMLDMX
KFLDTV.0 (.1)
TLRAN.0(.1)
TLAN
KFFA
KFFAVL
NOAWS
NUAWS
KFLDP.0
GLDSKNN.0
GLDSKPN.0
KFSWLSA.0
KFLDTVAK.0
Drehzahlbegrenzung
KFWLM
KFZWS.0
KFZWNA.0
DWVLN.0

Hello, I've kept searching in my quest to tune myself this POS. While comparing AJL (180hp) to my AEB (150hp), both stock softwares, I have seen KFLDS has higher values across the map than LDSMXN allows.

Can we confirm that besides having 8,50 ms/umdr in KFLDS for example, if LDSMXN has only 8,00 ms/umdr, the computer won't inject more than 8ms¿? Still trying to understand how manufacturer tuned them and how they worked.

Thanks in advance!

(https://i.ibb.co/Xt90NMf/KFLDS-LDSMXN-comparison.jpg)

It's all already chewed up! there is no point ... search the forum!

Hi! Thanks for your help. Do you consider this enough searching?

(https://i.ibb.co/QF17LFq/sc1.jpg)

And this enough info gathering, processing and concentrating in one only usfeul place? http://nefariousmotorsports.com/forum/index.php?topic=20919.0title=

(https://i.ibb.co/wSHVjDV/sc2.jpg)

I am authorised to have now more than "search in the forum" as an answer? When is effort enough to be allowed to start asking what I think are SERIOUS questions?

To whom may interest, i have good news, these maps work:

Copy pasting these maps from an 180 oem file to my 150 file did produce buttersmooth results, bumped flawlessly the car's performance. I still want to go a step further and make my own perfect tune and  I am docuemnting the process in ONLY ONE PLACE for future people searching for this information. Just asking if they are still more maps to tune, or better ways to do so, and make it more perfect. I don't have to pay nobody to do the tasks, I can rescale ALL load maps if the time arrives and if it produces the best results, it's not costing me more money and it'll be worth it if its the best that can be done. Just have to ask to learn to do it if I doubt and I'm trying to give something back in exchange.

Yes, I know information is spread all over the place. Again: I am just trying to unify everything in one place and in the most perfect way possible, but seems to be not interesting... sad.

Still thanking anybody who is willing to help this get somewhere and the people who asked and answered already in the forums.

Going more than the stock 180 file you need to fit a 4 bar FPR and sort out the fuel linearisation.
Otherwise you will hit 100% DC and lean out.

Hi prj, thanks for your answer. My bad I did not introduce the car, it's a A4 B5 1.8t AEB, so already have a 4bar FPR and 210cc @ 3bar injectors.

I have one doubt about what you comment on duty cycles. I've searched for it too but haven't found how to calculate DC in this ECU.

Have found a formula on other forums that says (injection time in ms*RPM)/1200 but that results in 12.75ms @ 6000rpm like 31% duty cycle lol.

Do you think you can help me to know when my injectors have maxed out?

Thank you very much. Soon I'll be receiveing a wideband with controller in the mailbox to be able to start looking at AFRs and perform my own changes, so more info coming.

Kind regards,

RPM / 60 = revs per second RPM / 120 = cycles per second (by “cycle” I mean four strokes) 120 / RPM = seconds per cycle = maximum time available for injectors to squirt

“Fuel injector duty cycle” is the actual time the injector is open, divided by the total time available. So:

IDC = IPW / (120 / RPM)

That can be rewritten as:

IDC = IPW * (RPM / 120)

…and to get an result of “90 percent” instead of “.90” you add a couple zeros:

IDC = RPM * IPW / 1200

If you look in RomRaider’s logger.xml you’ll find that formula.

So, for example:

7000 RPM * 17ms / 1200 = 99 percent IDC

What you see 12.75 is not injection

Hi aleks19411, thanks for the explanation! I will upload tomorrow your explanation to the file with your name in it ;) It'll be useful to calculate maxed out injectors for novices like me.

I still have my question related to LDSMXN, if somebody can help with it:


- How does KFLDS manage and LDSMXN limit boost? How fo they interact with N75 cycles? Isn't KFLDTV supposed to control boost by N75 duty cycles? I still don't get how or why does ms/umdr (load) translate to boost.


Thank you very much in advance!

12,75ms is the theorical maximal LOAD (time).

injection time can be logged in channel 2 and is limited by protocol in 16,32ms (wich seens to be 255/100 x load)

Load can be correlated to torque (as its function of theorical fuel need to achieve lambda 1) - you can see it through MDIST.

You ask load (torque) with KFLDS, correct it by Air temperature and elevation... LDSMX é maximum load.

KFLDTV are for overboost section (LDOB).

try those 2 things and you will undestand how it works

Zero all KFLDTV maps and see what happens, then restore then and max LDSMX, see again, then zero LDSMX and see what happens.

It will be cristal clear after that.

by the way, KFTLWS is not "only used when MAF is defective", it´s the predective load value

Hi overspeed, really appreciated about your explanation. I am working on installing an A3 flashable ecu to my A4 to be able to flash in the car, and your recomendations will be the first thing I'll try.

There is just this single thing I still don't understand, and that I am starting to think it's what somebody said on another M3.8.3 post: "still waiting for the moment that older Motronics start making sense":

- HOW does Boost control relate to Load, or why does the ecu use Load (time) to control the boost? I just cannot figure it out and get to used to using this unit to understand the boost control maps of the ecu. In KFLDTV i have N75 DC percentage which makes sense to me, but afterwards I have maps in Load controlling the same stuff. Why?  ??? :'(

Sorry if I seem dumb but this thing is starting to make me feel that way.

Very appreciated for your help. Kind regards,

LOAD = Theorical injection time to achieve lambda =1, so it´s direct related (by 14,7:1) to air mass...

This ECU don´t care about pressure (as it does not have a MAP sensor), its only base on how much air it would like to consume to produce some torque (load) wich is specified by TPS position

principle is the same...   Driver step on gas, wich TPS translate in a spec Load to be achieved by pulsing N75 (ok, advance, TBI angle, moisture are in the equation)... on DBW BEFORE ECU pulse N75 it "knows" how much boost, EGT, advance, etc thar will be necessary.

M3.8.3 just open throtle and reacts to MAF readings... if MAFreading is greater than what load is requested then it lower N75 and vice-versaf
 

I got it now! Thank you so much for taking the time to explain. Will make my tests and report back.

Kind regards!

Hi!

I did some logs this weekend to verify the car was running ok and to see the information about load and injection time, but as you can see on the log, VCDS shows injection times higher that LDSMXN allows... See the picture attached too. It is my 150hp tune converted to 180 from oem AJL maps. In LDSMXN I see up to 8ms/rpm and I did log higher values, exactly as KFLDS says.

(https://i.ibb.co/ZcnLfQ9/Captura.jpg) (https://ibb.co/vL7wcN9)

Also, the corrected injection time was over 16,32 as you can see in the log, going up to 21.68. Am i doing something wrong?

Thanks for your help

Hi everybody,

Just to inform I added v0.3 file to the Community Project thread http://nefariousmotorsports.com/forum/index.php?topic=20919.0 (http://nefariousmotorsports.com/forum/index.php?topic=20919.0) with updated and revised information. Go have a look! I still need help with stuff you will see to be completed in the file and with questions here in this thread. ¿Somebody willing to keep helping?

Thank you very much in advance!

Kind regards

Hello everybody, I am already testing several tries of stage 1 files I've made myself but since the first try I've experienced the same weird thing: boost surge only at WOT.

I've done several runs to test possible fixes: raised a bit KFTVLDRE, raised a bit according limiters like LDSMXN and KFMLDMX for example, but boost still surges at WOT. Boost won't surge if I back off a bit the gas pedal and will run great.

As you can see in the attached log, air readings from the maf will stop increasing for about 400-500rpms and then keep going up to redline. Also load does surge a bit accordingly, where i have set it up to 9,5ms from about 2500 rpm up to 4500rpms more or less, and the N75 duty cuycles do decrease way too much too: Values go down to 63% and straigt up to 88% while I have it set it up to about 75% at about 2500-4500rpm to maintain boost.

I've done changes, playing a bit with KFTVLDRE but I can only move the point where the air flow hump is, one test I did was at 120g/s for 500rpm and then kept going, this one is at 100g/s...

Air intake temps are ok at about 45ºC at most and there is no more than 1 o 2º of ignition retardation at the highest rpm only, that's why I have not included them in this log.

Do you guys need any more info to help me diagnose it? I am a but stumped, I don't know if I don't have the experience to read the data or if I'm missing something, or if maybe N75 is bad... I do not suspect leaks, and also have replaced with new hoses the N75 lines and DV holds good vacuum. Maybe I'm wrong assuming about leaks? Also, "actual RPM speed reads only up to 2550rpm which is weird, other fields read the whole rpm range...   ???

Hope somebody can give me some light with this. Thank you very much in advance.

(https://i.ibb.co/ZTMcpMR/Captura-de-pantalla-2023-05-27-a-las-15-57-51.png)

Nobody knows about this? Further research I've done has demonstrated it's definetly related to N75 duty cycles, KFLDTV map.

Does somebody have any insight on how to read the info in the log properly and tune accordingly?

Kind regards,

Hi marantzvieta,
thank you very much for your excellent effort. You inspired me to try a first step and convert a 150hp AGU engine (m3.8.3) to 180hp AJL maps. However I have a slight problem with the KFZW map, which is 16x12 in AJL and 16x16 in AGU. Did you do a simple resample? If the axes in the map packs are correct, they seem to have slightly different values and limits, which would mean a proper inter/extrapolation would be required from perfectionists point of view.

Thanks

Hi doktor,

Yes, you are correct I performed a simple interpolation and also used a bit of common sense. I used the perfectionist point of view you mention, even helping myself with excel charts to do calculations and observe different values that may work, and it proved to be a good investment in time. My car runs awesome.

Last load lines in AJL map are 8,25 and 9,25 (from memory). When I converted to M3.8.3 ecu, I used 9,25 load line values for 9 and 10 load lines in M3.8.3. I used them in load 10 line too because as 9,25 is it's last line, it uses those values for anything from 9,25 and above, so I considered it safe to be done this way.

Also used 8,25 values for load 8 line as I considered a difference of 0,25 in load to be almost insignificant, and it did actually prove good results in my case.

Always double check the changes you make, do logs and use common sense. Better be safe than sorry.

I am happy to hear my work has been useful to you. Please let me know if you need further help.

Best regards.

If anyone is interested I can dump the M3.8.x stuff I have also on my site.

I do have a file with specific injectors (Bosch green EV6) + MAF that is re-calibrated and all problems are fixed (cold start, acceleration enrichment, load cap, excessive fuel dump on WOT etc), and uses HFM5, so the MAF is not NLA.
It can take you to ~360-370hp.

I think anything more is better to switch to ME7 or standalone. There is no good datalogger for this old ECU and it just becomes a pain with the load limit and injection time limit. I spent a crazy amount of time 10+ years ago calibrating that stuff.

I would be very, very interested, I am sure your work can be super useful. I am still working on these ecu's and recently have been working on the VR6 maf conversion (you can see it in the helpers and how to's subforum). I know it has it's limitations but still haven't found them myself, and the power figures you talk about are even above my expectations/needs. I have complemented this ecu with a wideband and boost sensor logged with arduino, and narrowband output to keep the ecu happy and functional.

I will also update my guide as there has been quite a bit of development since then. My car is running up to 167-169g/s on the stock turbo very steadily, with a good power delivery, and with reliability in mind. I would love to keep improving my work and I belive your contribution will be very helpful.

Thank you for turning up in this topic and offering to help.

Best regards.

I did a car with a GT28RS over 10 years ago on this ECU. It made 350hp on the dyno.
I won't dump the file for free... but probably fairly low priced.
So don't really thank me for anything besides the info.

Knowledge is power, I understand your point. PM me with a price so I know what you are considering. Maybe I won't go for it right now but surely the time will come in the following months.

I appreciate you offering to provide your work.

Best regards

I have been reading a bit on the functioning of the m3.8 system and something is still not clear.

KFTLWS is starting point (?) for fuel injection based on TPS position. I found that "Fuelling calculated by (KFTLWS x KFLFLAV) x KFLF". The values resulting from the calculation are lower than those practically logged. How does boost come into play here? KFLDS is desired boost, based on target fuel (as this ecu does not have MAP). How does KFLDS interfere with KFTLWS? Or is KFTLWS a "starting point" for injection, while the turbo spools up and then targets (is regulated to) the values in KFLDS?

My point is that, for a simple Stage 1 tune, is there any issue with raising values only in KFLDS and corresponding load cap in LDSMXN (and possibly maximum MAF in KFMLDMX, but there seems to be a considerable margin)? Of course all this considering that 100% injector duty cycle is not reached. For medium rpm this should not be an issue and nice torque gains might be possible...

Thanks

See in the excel file guide, KFTLWS is a base map for prediction and used when maf is defective. Can't confirm how much is used in standard driving but i haven't touched them myself for stages 1 or 1+ kind of tunes.


Key point is to massage KFLDS, KFZW, KFLDTV and other maps like limiters and others in the list in the excel (I made a specific list of maps for stage 1), to acheive the goals you want. Most torque and midrage rpm driveability come from ignition maps.

Understand and forget from now on to not talk about boost with this ecu. It does have absolutely no notion of boost, only load, load request, and cycles n75 accordingly to obtain the needed air reading to reach the desired load. There is an "eureka" moment with this ecu when the load stuff makes sense in your brain haha.

Also there is a ton of people, not to say almost everybody I've known, that throws the keyboard to the wall to make a stage 1. I've seen tunes from professional shops with values as high as 11,70 on KFLDS that only made 1 bar of boost. I myself have acheived 1,1bar of boost with only 9,60 of load in KFLDS, so learn to optimise the values too. The study you made of converting from 150 to 180hp is a very good path.

Best regards

Any updates on this information? I am very interested in having a look at it.

Best regards.

Marantz, thanks a lot for the discussion.

"overspeed" in this thread wrote that "KFTLWS is not "only used when MAF is defective", it´s the predective load value". Not sure, how it would interfere with KFLDS then...

Sure timing can make quite a difference, but looking at the maps of the AGU and AJL, the timing looks somewhat similar globally. There are areas with slightly more advance in AJL, but also in AGU. I do not see a substantial difference. Also I do not trust the axes data, sometimes I see a weird anomally in them (like as skipped value), so I am not keen for interpolating the timing maps - as these are "high risk".

I was comparing the maps for about an hour or two, but it is not a simple "bigger numbers in AJL" thing. It seems they used a somewhat different approach, some maps in certain areas are considerably lower in AJL, while others are higher, they might still work together to produce a similar result - like KFLF with KFTLWS. This is why I started to dig one level deeper, to understand which things are only cosmetic corrections (which might be the case of KFLF and KFTLWS so these might not need any alterations), and which are the real thing. Also, there are quite large differences in KFLDTV, both positive and negative, this might even look they are for different turbos (but they both should be k03), weird. As I understand it, this map might not need a lot of tuning for stage 1 as there is feedback n75 duty regulation anyway, but could produce a larger turbo lag due to lower-than-optimal initial n75 duty.


I know, thats why I wrote "desired boost, based on target fuel", but I understand thats a quite unsuitable formulation.

So, for a not quite optimal but safe first tune, altering only KFLDS (with LDSMXN and possibly other applicable limiting maps) should work okay? Timing and other maps are usually factory-calculated up to 10ms "load", which is still sufficient for a stage 1.

Thanks again.

KFTLWS is part of "BGTLW: A 41. 20  Berechnungsgröße tLw - Lastsignal" wich is the load expected for each TPS position wich is your EXPECTED LOAD


With MAF, RPM, KHFM etc you´ll get  ACTUAL LOAD

There are a bunch of filters


KFLDS is you DESIRED LOAD, based again in TPS x RPM, 


KFLDTV is the N75% output to make ACTUAL LOAD = DESIRED LOAD,  If  ACTUAL LOAD < EXPECTED LOAD  you are in a underboost condition and  special KFTVLDRE is used to cap N75 output, and DTC will be set

Thanks a lot. Makes sense.

Thanks mate, your explanation is brilliant, I will include it in the excel file. :)

Actually, timing makes most of the power change in this case: yes 150 to 180hp is a noticeable change, but 210nm to 235 isn't. The change is basically in air flow, but in Stage 1 style of tunes, these cars can go easily to 290 to 300nm while still being safe. And this is done mostly with timing. Be sure these 1.8t's have room to work with. I you want to touch KFZW I recommend to search and look for tunes all over the internet (they are) which have been run, and develop your own ignition map (most changes are made on the higher load rows). Make small changes and log for ignition retardation, you'll see. Remember to make changes with common sense and in little increments, and then log, log and log to check your work.

Please everybody, if I am mistaken in something or think different, let us know. This is how I did it and got great results on the road, butt dyno, and logs, but I'm sure there is more experience out there than mine.

Best regards

I get it you mean tuned timing maps for ME7.5, after proper load recalculation? As I was not able to find a tune file for M3.8 other than the one floating on Youtube, from which I had to rewrite all the values by hand one-by-one. ;D Some of his edits are Italian-style, like almost disabling protective features, but the timing map looks quite reasonable(?). I think there is no reason why ME7.5 timing maps should not work, but the load recalculation must be done properly, I have the basic idea how to do this but not sure about precision.


Thanks again.

P.S. this is the video https://www.youtube.com/watch?v=raAmPkPr_8w

Edit: After investigating the timing maps further, the youtube tune looks somewhat extreme... +9 maximum advance... But it is not far away from a stock 180hp ME7.5...

Another edit: Does the M3.8.x implement a functional knock timing retard? I was not able to find this info anywhere...

when running a log on vcds for ignition retard it definetly cuts out ignition angle, so yes the ecu has a routine for detecting engine knocking

There was adaptive knock control already in M2.3.2 and knock control in M2.3, so of course M3.8 has adaptive knock control.

Cool, thanks. And sorry for the noob questions  ;D

You can find it on the file site now, if you ever need it. I threw in an XDF as well.
I think if you're doing something with larger turbo, this info will save you at least 10 hours of calibration and chasing all kind of messed up problems like car not starting right hot & cold etc.

Some of my findings:

KFLDTV is not a starting point for N75 duty. It is the actual limitation. I was having problems with a constant underboost and raising the values in KFLDTV fixed the issue.


From what I gathered, for a Stage 1 tune, stock AGU (or equiv) internals:
-KFZW (.0, .1, .2): add a few degrees in the high load area. I used similar values like in 132kW AUM engine and I have an ocassional ignition pull. Dunno why, the AGU should breathe better...
-KFMLDMX: Might not even need altering. I did a +10% in the whole map.
-KFLDS: I left the low to medium load areas intact. For a high load, 9ms between 2k and 4k rpm is safe, taper off a bit to higher speed. Some tuners use up to 10.5ms.
-LDSMXN: Basically use the highest load line from KFLDS.
-KFLUL: Adjust percentage in whole map to be a bit higher than KFLDS for normal altitudes.
-KFLDTV: High torque areas should see something like 80%, or even a bit more. Depends on your actual turbo, n75 and engine health...

Do not alter the following: KFLF, KFTLWS, KFLDSAK, KFTVLDRE, TLRAN.

Regards

Hi doktor, this statement is wrong. N75 is in fact starting point and must be used to obtain a fine airflow delivery from the turbo. PID is the actual limitator of how much it can correct the deviation based on how off the request is from KFLDTV.

Have been there. Raising values in KFLDTV just allows the PID to work on its intented range, but relying in PID is not the best way. Make a log of channel 027 and tune KFLDTV with the results of the log +1 / +2% on the last three rows. See how turbo work is smoothed and so power delivery.


I suggest you tune a bit KFLF on the high load and high RPM zones if you intend to run the car frequently there. I would not drive in the autobahn or a track without the peace of mind that it gets enough fuel there.


Also take in consideration that with more boost and fuel, ignition advance should be less, not more, to avoid detonation. This is because denser air burns faster and reaches peak cylinder pressure earlier too.

Best regards

I appreciate it. Time will come in the following months. I like to see this kind of jobs to understand the work you made and learn about it, so I really appreciate giving the opportunity to access it.

Best regards

Hi marantz,
I get the impression that the table might be a "linearization map" as can be sometimes found on the net. Like it is a correction factor between the PID output and actual duty. In that case, the output duty could never go higher than the last row in KFLDTV, and that was basically my observation. I was getting a constant and predictable underboost with a relatively low N75 duty, the ECU did not want to increase it even for a several second long underboost. On a closer inspection, the N75 duty was close to the last row in KFLDTV. Raising the values fixed this problem without issues.


Well, the axis of the table is LOAD. If you run more boost, you have more LOAD, so you are automatically pulling numbers further from the table, which are naturally lower. So even without altering KFZW and increasing only boost and fuel, you get less timing, ig your timing table is sensible.

by the way, WOW, I just logged timing pull for my other AUM engine (110kW), which has been remapped to factory AUQ maps (132kW). The engines themselves are identical so this shoud be a perfect fit. But, I am getting a steady 6° retardations for a WOT pull, some numbers are even higher. This is much more serious than my little 2s for AGU tune and basically with factory maps. The car is mapped like this for maybe 2 years, almost daily driving and there are absolutely no issues with it. And an interesting finding, there is always 0 pull for cylinder 1. This is the same for AGU and AUM.

Thanks again for the discussion.
Jan

Hi doktor.

I have logged N75 duty cycles higher than those on KFLDTV, almost 10% higher than those on the map. My car was mapped in winter when air is colder and denser, so on summer the turbo had to work almost a 10% extra to move the same air quantity. I did notice the car behaving ever so slightly different when it started the hot days here, and as obssesed I am with this I ran some logs and noticed it. Leveling out KFLDTV has allowed me to keep the same behaviour and performance all year around even though we have like 25-30ºC difference between winter and summer here where I live.

I believe (but cannot confirm) a "perfect tune" would have compensate the PID maps somehow for this kind of cases, as it is not the same having to do compensation work for 0,5bar than for 1 or 1.1bar, or even more in the future (bigger turbo incoming). But I do not understand the PID maps yet, it is some work I still have to do. If someone knows about it, I would appreciate it a lot if you could chime in and give some insight about it! :)


About timing: how are your spark plugs and coils?  When I was on stock tune, I had 3-4º of timing pull with the plugs the car came with, and with a fresh new set (PFR6Q for 1.8t AEB), it went down to zero. With my actual tune I only get occasionally 1-2º of timing pull on 95 octane petrol. With 98 (highest available in my area) it is basically zero.

Let's keep this chat up, I haven't had many opportunities to be able to have a long chat about discoveries with this ecu.

Best regards

Marantz, thx for the info with spark plugs, I have this car only for a while so they might be worth replacing.

As this is meant as a guide, here is my take on boost vs "load": The standard boost level on a 150hp AGU engine is allegedly 0.55bar, which translates to 1.55bar MAP. The requested load for this level seems to be around 7ms. So, running at 9.5ms (as I am now), this should translate to around 2.1bar, or 1.1bar boost. Seems quite a lot. On some tunes, I have seen numbers as high as 10.5ms, this translates to over 1.3bar boost? How does the poor k03 cope with this?

I shoot an equation: MAP = load[ms] * 0.22.  But I believe the turbo efficiency comes into play a bit here...

Btw, the apparent difference when going from 7ms to 9.5ms actual load is BIG. The stock engine is not sluggish, but it is boring. There is no turbo kick. After the remap it is fun to drive.

Hi mate, it's all about code optimisation. I cannot explain myself how a reputable tuner everybody knows across europe and the states only managed to make 187hp from absurdly high values in KFLDS. This is something I have seen personally, no assumptions made or "a friend told me".

My advice is to follow your own path as I did and as you're doing: make your own learning, observations, experience, and do what works for you. I learnt a ton from this and other forums, but I did my own testing and observations and I am using what has been working for me.

Also yes, the stock tuning while obviously super reliable, is obviously super boring too. Since I made my first test to 180hp I haven't understand how Audi/VW ever let the 1.8t go into production with only 150hp. I believe this had to be a financial decision as they also had to sell those V6's and 180hp versions for an extra cost. There is a cool video on youtube of two A4 B5, 180hp AJL and a V6 30v 193hp runs compared, and the 180hp keeps up with the V6 untill 200km/h.

Best regards,

I wonder how anyone could get any more with stock turbo. The 180hp variant runs at about 140g/s MAF at full power. See attachment...  ??? ??? ???

you are showing the wrong compressor map for 1.8T, 1870E is the one for 2.7BiTurbo.
1.8T uses 2072G and later models 2075E compressor.
There is still some room for tuning, noticing the 180PS line (red)

Leon, thanks for the reply. There were two variants of the 1.8t 150hp engine, the earlier like AGU with large port head and smaller k03 turbo (and mostly M3.8 ECU), and later like AUM with small port head and slightly larger "k03s" (though Im not sure if this is the official designation). The compressor map you posted is for the smaller k03? I was not able to find the exact type number and assumed its the 1870.

THIS is why i wanted you to post it here. 140g/s for a 180hp tune is low. A healthy 150hp must flow about 130g/s at redline, a healthy 180hp should flow 150g/s at redline. You might have a vacuum or boost leak, maybe a defective component somwhere... this is why you are having underboost issues and not reaching requested loads.

My car has reached 130g/s with 150hp tune, 152g/s with 180hp tune, and now I am flowing 164g/s at 5700rpm and 169g/s at 6300rpm. Which I believe to be the absolute reliable limit to stock k03 turbo.

To estimate power it is said through the internet to use maf reading/0.8 but I consider this calculation too optimistic as 130g/s would translate to 162hp. I use mafreading*1,2 which translates 130g/s to 156hp (more accurate on my perspective). Same for 180hp, do the maths. 169g/s*1,2 is 203hp, which I belive to be accurate when running against similar power/weight cars.

Smoke test your car and make a thorough check. Untill you don't get those values with stock tunes, do not tune any further.

Best regards.

Sorry there was some misunderstanding, I was just wondering how any real figures above 180hp would be possible with a stock k03, but then Leonhard pointed out that my compressor map was wrong. I believe the car has the K03 with 2072 compressor which looks much better than the 1870. I did one more log and it seems it works acceptably well. The slight underboost happens between 2.5 and 3.5 krpm for 3rd gear pull. For 4th gear its better, and for 5th gear it is almost perfect. This looks like a job for PID tuning, but I don't even know any other maps except ZSOLLDRN. I think I will just leave it as is...

Edit: About ZSOLLDRN... It says "Boost Control Filter Time Constant for Low Pass". I know what is a time constant of a low pass filter, but low pass where? The feedback loop can contain more lowpass filters...

-Jan