NefMoto

Hi, I haven't been here for a while but I modified my TT BAM after dropping a 5mm allen bit in the inlet #doh
Forged Rods, new rings and bearings, Inconel SuperTech exhaust valves, skimmed head, FMIC, B5 TIP, Cosworth S2000 Filter, 3Inch DP + Sports Cat and a Hybrid K04 CHRA in a stock / ported hotside and 15psi actuator, ported chinafold manifold, Bosch 550 EM14 Injectors, 3 Bar FPR

I have been trying to set up a safe map and think I have the fuelling right after resizing and setting the voltage offset from the s4wiki, and have a basic understanding of the BTS, LAMFA, KFMIRL, LDRXN, KFMIOP - these I played with on my Stock K04 and stage 2 modifications. Pre-rebuild I struggled to advance the timing so I am still on stock timing (but post rebuild the timing has held up quite well).
Whilst trying to understand the best route forward regarding the Wastegate Duty Cycle,  with some pointers I have set the Maximum Waste Gate duty cycle TVLDMX, which I understand is a safety net - I have tried a few settings and 27% appears to be a stable starting point.

I know the tables involved are KFLDIMX and KFLDRL and clearly a stock BAM map is much higher than the current Maximum I have set.

I have attached my currently installed BIN and a log from this morning with a 3rd gear pull to 6400.

I have seen some discussions regarding LDRAPP <-- is this a tool?  but most of the threads appear to be regarding S4 with twin turbos.

many thanks for any pointers or directing to the correct area :)

(I updated the CSV as it didn't work in ECU Plot after I changed the column order but this version does)

LDRAPP is not a tool. Read Datasheed for MED9 aka Funktionsrahmen or look at S4 Wiki and you will find more information about LDRAPP ;)

Thanks I will take a look,  I will say I am struggling with this a little - my biggest problem is I dont like red-lining my car after spending so much time and money building it - and a lot of these tests involve change a setting floor it to get a log and hope it doesn't break  :-\

I even woke up in the night discussing an inlet value breaking and for some reason re-using the same head gasket :D :D

Lol @ your dream ;D. If you want a starting point maybe do a few runs with fixed WGDC using CWMDAPP = 8 then setting KFLDRAPP (revs vs pedal position). This way you can safely see how your turbo boosts with safe limitations. Plus you will be gathering logs for the linearization tables ->> http://nefariousmotorsports.com/forum/index.php?topic=517.0

arrh thanks - I will check that out,  I was just popping back on this forum as I seem to be hitting intervention if I am just driving normally when I have a max WGDC > 27% making it a nightmare to try and improve things,  I am planning to save up a bit and get over to Badger5 in a few months - but nethertheless I still want to learn all this (we all know driving a project car is just the test for what you want to change next) - it really is so much more difficult than simply changing a stock map when you only update the air flow and leave the stock turbo, injectors and actuator eg increase a bit of fuel and load and try some minor timing advance if the engine is happy

I am pretty happy with the fuelling side - the log follows the map even though the desired torque actually is delivery a much higher actual torque. I just need a nudge in the right direction from my logs I am about a stage 1-2 map with a stock turbo but from 5K instead of 6.5K - but it really does feel like I am changing far too much without a full understanding of them, as an example I am trying to see if overboost is causing intervention by disabling the intervention by adjusting KFDLULS (pressure change for overboost protection diagnosis) and TLDOBAN (Time for LDR overboost active) which just feels dangerous.

I will look at the fixed WGDC and roll back to my original post injector settings and post FKKVS_FIXER MAP

I was thinking about getting a 3inch cat back exhaust but the stock TT cat back is not causing any of these issues.

Thanks for your help  ;D I really appreciate it.

This thread you have provided is exactly what I needed - I hope that the car manages to continue without overboost intervention in order to create the logs (I am not sure I will go as far as 80% as that is apparently ~2bar - 30% is near the MAP Sensor limit at 2400mBar

Thanks for the "nudge" that I needed :)

http://nefariousmotorsports.com/forum/index.php?topic=517.30  <-- it really is a ground up approach to creating the starting position for the PID

to be fair I am surprised the resellers of the hybrid turbos / CHRAs etc dont provide a starting point PID for them - as these should not be dependant on the rest of the hardware should it?

Rest of the hardware will have a big effect on the spoolup and overall behaviour of your turbo.
Its not possible to provide a pid starting point with a turbo.

read this http://nefariousmotorsports.com/forum/index.php?topic=12352.0

Would start with a stock file. Do it yourself and it will be better than "the english way" @ badger.

Thanks for the pointers - how times have changed from trying to fit a twin-webber carb to an escort mark 2 1.3 - to essentially becoming a maths expert :D

wow that is an amazing tool in VS2015 :D :D

will adopt this feed-forward mindset will re-read the thread again whilst gathering logs :)

Hi,
I wonder if someone can help me please?

I am trying to do some logs for the feed-forward mapping approach.

I am setting the calibrate item to 8 and LDRAPP and have run 0, 10, 20, 30 (I previously set the max limit to 28 so assumed these were a safe set to start with)

when looking in the log files the WasteGate is 0 as expected in the 0 file, but then 3.9, 7.8, 11.7 instead of 10, 20, 30?

I have attached the 30WGDC Map and the accompanying log file.  Is the XDF Parameter incorrect or something?  there is no equation in the XDF I am wondering if it should be something like X / 2.56 - or are there other offsets coming into play here?

I am not going to do any other logs for the moment.

no log attached
will have a look into your file later today

Right this sounds like your .xdf has the wrong equation setup for KFLDRAPP:

10 wgdc = 3.9, yes?....     therefore     3.9*2 = 7.8 (20 wgdc)     and    3.9*3 = 11.7 (30 wgdc)

You are out by a factor of 0.39

The solution I believe is this:

1. In the parameter tree left click on KFLDRAPP and then go to 'Edit XDF Parameter info'
2. Go to the conversion tab, click 'Edit Global Table Equation' and then click 'Edit'
3. Whatever the equation says, do not change it, just add this to it; *0.39

    for example, if it says: X*0.123456     then make it      X*0.123456*0.39      if it just says X          then make it X*0.39

4. Click, ok, ok, ok.

edit: use 0.390625

This should fix your issue

great answer and totally correct!

Indeed.

thats cool thanks - it will work better than my x / 2.56  which had a similar effect.

I will make the correct changes to the XDF

I have done a few logs using the calibration mode - does it purposely reduce the loading?  I seem to only get to 114% actual load even though the boost pressure is going up the map sensor?


also I am getting the ESP light lighting up at 6K RPM on my 30WGDC and 5K RPM on the 40WGDC.


17963/P1555 - Boost Pressure Control: Upper Limit Exceeded
16487/P0103/000259 - Mass Air Flow (MAF) (G70): Signal too High


Also when I use the SB GLI C# KFDRL program and upload my logs the figures go crazy high like 5000 Mbar?  even though none of my logs exceed 2450mbar?  I know that the tool was originally used on a 1.8T so it cant be that?  I have even trimmed out all the data that isn't 100 throttle plate and faked the 100% throttle plate on a low 4WGDC as it could only get to 88% throttle plate on 100 PED? I am wondering if I can manually fill out the table fro the logs.

I have a hybrid turbo and wonder whether it is giving out strange readings?  if I leave the default range to 75 / 0.1 seconds I doesn't create any figures for 20 or 40?

I have tried to zip together the logs - to see if anything obvious becomes apparent and will re-read the pre-control / feed-forward threads :)

What about your load cap?

Check out HBN and/or LDRXN and increase it to have a boost request near map limit.
This will result in a non-closing throttle because actual boost will not be significant higher than req boost.
You will have to set a reasonable boost profile with LDRXN after you did all the logs.

relog 40wgdc and check if the values in the ldrpid tool will make sense.

While logging 50wgdc and higher you should reduce timing (check retard in 40%log) and what about your lambdareq? its very rich without being close to 950degree egt.

check out my last post here. i ran into similar problems during logging.
http://nefariousmotorsports.com/forum/index.php?topic=12352.msg122780#msg122780

Hi,
Thanks for your reply I will have a scan of your journey :)

This load restriction only happens with 8 in that codeword for calibration.

I am happy with the fuelling,  it is not running a BTS Table because the EGTs are nice and low this time of year so not sure where 950 degrees centrigrade comes into it - that is BTS intervention.

The log files and fuelling are following my "desired load" <-- which we all know is a nonsense driver on an ME7 it should have been actual? I set in Lamfa table - I have set this to as low as 0.85 to 0.82 at high desired load high RPM as that is ~12.5 - 12 AFR designed for power and torque and obviously keeping the components cooler.

http://www.endtuning.com/afr.html

I have run the 50% WGDC and have a small increase in boost to the map sensor limit, and actually managed to get the LDRTOOL to work if I trim out the other columns.  although there is a drop in mbar between 3K to 3.5K RPM which I think I have seen somewhere else too. and the LDRTool put in PID WGDC to 50,60 even though I had not made any runs before there maybe to compensate?

 also had a bit of a problem using the FKKVS Tool I used it and in the mornings the car coughed and spluttered for a few seconds? I think it had too much fuel as the issue was worse if I double primed the fuel rail turning to 2 then off then 2 again - since rolling back to stock FKKVS it han't done this?

Some my tips

950 degree C is for me very hot. Some tuners maybe setting 12.5 AFR at top but is no good. This solution is used for stock fueling system but not good for engine. You overheat engine very quickly. If You not have custom high flow manifloid, You should add more fuel at top. On big boost i trying not be too close 1000 degree.

I calibrated some hybrid turbos. I made logs with fixed duty cycle of WGDC and created table of duty by hand. You must remember about compressor limit HBN map and DMX map for maf because You hit limits.

If You hit 1.55bar You must change map sensor to 3bar and make some offset/scaling of maps. Goes over 3bar map is harder to do because You should implement 5120 hack for boost.

You can work with duty, reqest boost map, ignition to prevent overboosts if You want stock map sensor.

turbo needs temperature because its energy and 950max is okay for borgwarner cast
this ecu has active egt probe so its no problem at all
when done with tuning you should run the highway with 4th, 5th and 6 gear to check

regards

Hi so I have 2 directions currently with my maps - one I am tweaking the PID after using the tool and the other one I am using stock PID and raising the TVLDMX

Attached is a log with TVLDMX set to 29% max wastegate duty cycle.

I am happy with my fueling and happy with my EGTs.

Lamfa is generous 0.82 up top,  my EGTs are currently maxing out at 850 (although cold weather here currently)

as a reminder I have
2003 TT 1.8T BAM

with modifications

K04 with hybrid K04 CHRA and Ported hotside
1 bar Actuator
Ported Chinafold Manifold
Bosch 550 injectors  (no uprated fuel pump yet, although I do have a stock HAAS pump if necessary)
B5 TIP
RAMPRO Air Filter
3inch DP with Sports Cat (going to switch the decat)
R8 Coil packs
WellyCooler FMIC with 64mm pipework
PEC rifle drilled forged Rods with racing big end bearings and rings gapped at 0.4 and 0.45
SuperTech Inconel Exhaust Valves
additional earthing

attached screenshot from tunerpro including lamfa and Max WGDC and CSV from a 3rd gear pull yesterday.


what I a keen to learn is whether a 5120 hack and map sensor is a better option than simply going to "open loop" - I am struggling to know whether the car simply goes into "open loop" if the LAMFA is different to 1 and outside the limits of the sensor?  isn't it supposed to fuel the car based on MAF g/s and wideband o2 sensor, there seems so many Me7 arguments regarding this subject on this forum as s4wiki says one thing and people say that's not the case on 2002 wideband 1.8's? but I have noticed the MAFG g/s are limited by the WGDC limits as an example a limit of 25% is 200 g/s - 30% is 218g/s

you can see from ECU PLOT that the boost desired PSI simply hits limit now and stays there, whilst the actual PSI is dancing under the desired whilst hitting the TVLDMX, hence my urgency to control the boost using the PID tables.


I also need to know if I should be pulling back the timing at peak power? Timing is still stock but I am getting ignition retard voltage so will probably reduce the timing to keep this nice and low, rather than reducing the energy?

Any thoughts or guidance I am about to do a run with the PID data built from the LDRAPP tools, I did try the FKKVS Fixer tool but until I think I have proper control of the turbo via PID (instead of max WGDC - I shouldn't worry about these Fuel Trims - as the car coughs and spits first thing in the morning for about 4 or 5 seconds.


This learning is great fun but I really am pushing the limits of my understanding, the steps involved in getting all this extra hardware configured are pretty extreme compared to a few updates and copied map contents for a stage 1/2 of just increasign fuel, load and timing.


Any thoughts or pointers to reading will be gratefully recieved.

way too much different things involved here

you shoud do it like s4tuning wiki: first do fueling, then boost, then timing

in your case simply change lamfa and thats it for the moment. egt sensor will safe your ass. fkkvs to all 1's for the moment. keep an eye on your injector duty with higher boost. you need a pump.
then i would change ldrxn for a more k04 like profile. your req boost should look like: spool up with 95% duty as fast as possible until your wanted boost, then taper down to your wanted boost at redline.
In my car its like attached which results in 1.6bar to 1.1bar
your ldrxn is wrong in my eyes because your max value is at too high rpm.

this is why you bought forged rods. max boost as early as possible. you should log from below 2000rpm to allow the car to spool the turbo. so at 1700 hit the pedal hard!

your preload is strange! so much boost with just 29% duty.
attached my ldrxn, my lamfa, a log with 1.6 to 1.1bar (but with too less 95% on spoolup)
maybe its best to stay within the mapsensor limit to have a working pid and a understanding for how things work.

excellent, thanks for your guidance / advice.
 
I am setting a max WGDC to make sure it doesn't go crazy whist trying to be properly in control of the PID.

I actually ran a post LDRAPP tool log yesterday still with a limited WGDC and it did behave better in mid range.  I think I know why my fixed WGDC logs were low load - I had misunderstood the overboost table regarding setting them all to zeros from s4wiki - I assumed it meant that it would ignore the content - but I think it meant it will not allow any overboost.

I will try and create some logs from 1700 rpm, with a more aggressive onset.

thanks :)

Hi,

This comment

"your preload is strange! so much boost with just 29% duty. "


You have to remember I (a total amateur) made all the engine modifications, from gapping the rings, rebuilding the repaired and skimmed head, constructing the turbo (from a hybrid CHRA, stock K04 compressor housing and Chinese hotside - ported penny wastegate and turbo/mani to fit match chinafold stainless manifold)

I bought a 1 bar actuator and fitted it after porting the wastegate hole and when the door was closed gave 3 turns of preload.

so I may well have added too much pre-load to the actuator?

I was advised to put a 30% max WGDC after a strict running in period was completed, as the hybrid and actuator would be delivering about 2 bar at 80% WGDC - so I need to ensure the engine build etc is safe.

what does it mean if I have too much pre-load - does this become a challenge to map neatly?

So if you used a "upgrade" chra but a stock compressor housing what is upgraded powerwise?
You cant combine a upgrade compressor wheel with a stock compressor housing.

Close the WG by hand and add 1,5 turns on preload. This is enougth and will allow the wg to work.
Then do logs in application mode starting from 0wgdc, 10wgdc and so on.

it's got a bigger compressor wheel and upgraded internals.

https://www.ebay.co.uk/itm/Audi-S3-TT-Leon-Cupra-210-225BHP-Hybrid-Turbo-Billet-CHRA-Cartridge-K04-023-BAM/321811933542

I don't know where to start.

First off, you are confusing open loop fueling with open loop boost control They are completely separate topics.

Secondly, you will be adjusting timing based on knock correction, not voltage.

I'd stick with requesting 22-23 psi for now. Don't worry about 5120. Stay below the hard max boost ceiling for now, and set DSLOFS to zero to make sure you don't screw up.

thanks nyet ;)

keep calm and carry on :D :D

yeah not the voltage the highlighted section of "ignition retard" I know that is bad and also s4wiki suggests to reduce some timing advance over peak power to try and keep from continuous knock intervention.

I will stick with the map sensor limits for the moment :) as clearly have a bit to learn regarding open loop boost control.

will check out DSLOFS - I did see this in the S4Wiki regarding working within map sensor limit - it raises the map sensor limit a tiny bit if a recall but does this then also act as a boost hard cut controller?  I will update my testing map for tomorrow with this for my logs (on the way to work)

have a good evening :)

No, it ensures that max measured boost is never under the max possible requested boost, which could cause severe problems. Analogy: say your thermostat could only ever read 40 degrees Fahrenheit but you set your thermostat to 60. Your heater would never turn off.

please reduce preload to 1,5turns and post logs starting from 0% wgdc

1bar spring has no sense here
0.8 would be perfect
29% better than 0 so whats the question? Read wiki @tighter wastegate section
Set kfldrl to sometging between stock and linear
Make run with some setpoints corresponding on kfldmix axis 1.0 1.2 1.4 and make things done
Thats all
Imo no reason to run over 1.5 bar this tiny turbo

information overload :)  all these map changes are in the evening and a run on the way to the office so apart from the weekends I cannot make a load of changes and runs.

I have done a run after applying the LDRAPP Tool changes based on the current setup I have (I have read other people only going to 30-40 WGDC)  now this is not hitting the max limit I set now which is excellent, and I reduced the timing advance in the 130-191 Load 4500-5500 by 5 degrees this previously was on the knock limit but now virtually nothing - which is good I will push it forwards a little bit more.

when you say no point in going over 1.5bar is this because the damage / risk is raised for very little benefit?  there are K04 Hybrids hitting 400BHP nowadays (this one obviously wont) but currently I am only at the same level I was with a stage 2 map + 80mm intake, air filter, downpipe, FMIC, stock turbo, stock rods, stock injectors - if you are suggesting a hybrid, forged rods and £260 of supertech inconel exhaust valves are only good for 275BHP - I will spread the word and save hundreds of people thousands of £'s upgrading ;)

Mostly so you do non-5120 tuning before making the leap. On thing at a time.

There is no k04 which gives you 400bhp on a non tfsi engine
Simply because of a small turbine a/r = high backpressure.  No matter which wheels are used,  a/r always stays the same
Its a 9 deg timing for now, how do you think what happens @1.6 bar for example? 4 deg? Useless with stock CR and non racing fuel
You can try to raise the pressure after 4k (pretty safe for rods)  and watch for airmass / iat / egts etc changes

Your wg is overtightened,  another simply rule here - take a spring which gives you ~90% wgdc @redline with specified pressure. In that case you have a 90% range to operate (not 29% you have)

Its a weird thing also to tune timing with an kfzwop2
Use kfzw for that
Kfzw2/kfzwop2 used when nws active (1800-3000rpm on your engine)

Also take a look at the k04 compressor map. Running over 2.5 P/R is pointless.

And I'm baffled why you took a screen shot of KFZWOP

Thanks for your advice.

Badger5 AET380
https://badger5.co.uk/turbo-exhausts/k04-380-hybrid-k04-package-gen2-chinafold-aet-k04-380-v2-2-tipexchangespecial-order
with dyno's over 400BHP on his AUM 1.8T Lupo

most AET380 fitments on TTForum / ASN are around 360-380


also I think TTE make some turbo's closing in on these figures too, but they have a K04-64 frame instead of a K04-23 so unfriendly to a quattro 8N / 8L especially in the UK

thanks for the extra details,  I actually dropped all those timing maps (in that area) - but returned to stock as I think it was a fluke not to get timing retard in my test run, as I had also reduced the load in that map too.


I need to do some more reading :)  Like I said I am happy with the fuelling it is following my LAMFA Map and I have a safe BTS map to fall back on.  Other than that it is just inching forwards, then backing off again :) - but as advised now quite a bit releasing some actuator pre-load will hopefully give some more control and a recognisable PID - as I did change the axis to 0,4,8,12,20,30,40,50,90 to suit my LDRAPP Tool runs





I will look to reduce the actuator pre-load to give a wider range in control