NefMoto

I've been running into a wall here trying to understand how these things fall in place. I've been having issues with my 1.8T GLI that I just cannot explain.  It's a stock AWP, with a K04.  I'm using the stock RN bin file, fueling is LAMF > KR > BTS (activated @ 800 for protection only).  Fueling is stock inj w/ 4 BAR regulator, stock pump.  I seem to be leaning out above 5400ish RPM which is another thing I don't get because i'm only ever reaching 1 BAR max boost, so maybe my pump is weak/going, etc.  I'll have to look into that.  under 5k fueling meets and holds req lambda though.

On my latest file, I'm requesting via LDRXN 180 load from 3k RPM on.  I see this in the load as specified load.  Requested load though for whatever reason is lower, say 160 which seems to be around 1 BAR requested boost.  I'm not quite meeting req. boost (say 1900mbar vs req. 2000mbar) so actual load is like 150.

I have no knock, pedal is 100%, why is it that it's not following LDRXN and requesting more boost and load (180?)

Attached is my latest log from from my file.  Some thoughts would be appreciated :)

The more and more I look at the logs, it seems the KFLDRL/N75 control has a direct relation to requested boost. If that is the case, how would it know that say 75% of the N75 = 1 BAR (eg. only).

I've cleaned up my DRL and IMX on my next revision and obviously the car is now much happier.

But still I'm not quite understanding why the ECU is basically choosing the req. boost setpoint as it is.  It doesn't follow req. or spec load at all IMO.

isnt KFLDIMX  about how much n75 for say 1000mbar ? At least thats how I use it and it works.I rescale the axis .
Then KFLDRL for fine tuning .As I understand it KFLDIMX is 80% N75 DC for 1 bar ,  KFLDRL is 78% for 80% N75 DC.

But for 1.35-1.4 bar pressure I never had to play with KFLDIMX and KFLDRL  .

IMX is the I-term of the N75 valve at different points in boost.  It's basically the table that defines where you want to be with your N75 DC.

Having DRL less than IMX makes no sense.

log the duty cycle of the n75 and make changes to the table . Is the requested load what you want it to be ?

http://nefariousmotorsports.com/forum/index.php?topic=4234.0title=

No, DRL, is a linearization table.... you can have it be whatever you need it to be.

I know what FR says about I-limit etc. I actually use it like I said. You can input 95% all over the IMX or 10%  it doesnt matter much
if you tune correctly LDRL.I am not saying its right , far from it ,but from my exp. only LDRL matters in the end :)

That's all well and nice, but does no one know the answer to my actual question here?

why is the ECU correcting spec load to (eg.) 160 when I'm req. 180?

Nye, re. DRL I'm pretty sure I read somewhere from PRJ that having DRL lower then IMX would cause issues.

KFLDHBN , look at my last reply

HBN is not limiting me in any way from what I can see, unless I'm mistaken that HBN isn't turbo PR (2.5 = 1.5 BAR)?

(https://lh6.googleusercontent.com/-QEKtN2V-LnA/UnZ6BPnASnI/AAAAAAAARiA/NQ4JPUrYAD4/w1007-h521-no/gli_file.png)

The biggest correction to "spec load" to "corrected spec load" is IAT. There are several others though. Check the FR and do some more logging.


That statement has no meaning... since they aren't "compared" in a way that matters. DRL can be as non-linear or linear as you need it to be..

I want to add just 2 observations from my limited expirience  :)

Your KFMIRL is not as it should there is a gap , did you fine tuned it in 3D ?
Also KFMIOP is not corresponding to KFMIRL .

also did you change anything to KFPED ? Yout wped is strange compared to throttle angle.
In my loggings they are very similar.

Just my observations , maybe the experts can explain better :)

001 or 02x?

Not sure I didn't put the turbo on this car.


Yes it's very linear in 3D.


I used the ME7 tuner wizard excel document to make IOP for this file (normally I make it myself), but I have no noticed any odd things happening to be honest.  Everything is very smooth.


Yes, I changed the axis to slightly be less aggressive.  It was something I experimented with a while back and never bother to put it back.  It makes the car much more mellow the first part of travel which I like.  Otherwise I felt like the car wanted to always leap off if I just breathed on the pedal :P

Thanks for the insight and looking at my logs!

Note: I didn't hijack this thread --- somebody merged my thread into this one, i assume, because they are similar topics (although different circumstances)

hey all,
I'm finally really digging in and trying to make a nice tune for my AWW 1.8T.  it's a stock motor with only a FMIC.  The tune i've attached is very crazy and likely has loads of issues.  I initially had KFMIRL and LDRXN max out somehwere around 200 and i experienced the same issues i will describe, so i upped them very high to around 250 to test things.   i probaly am incorrect in this, but i was thinking the fact that the last row in KFMIOP being ~82% meant maybe i'd only see 82% of LDRXN and KFMIRL -- i realize this is probably totally wrong.  I've also added a lot of fuel with LAMFA, although i haven't actually seen the effects of these changes in logging.  

My Goals:
- ~19-21psi (maybe more -- i don't mind pushing this motor as the car/engine were very cheap)


Problem:
There are a couple...
--I only ever see about 15-16psi requested boost.  The maps should make far more.  
--AFR ratio stays very lean and doesn't match request at all.
-- The car feels like there's intervention or something.  Boost surges like crazy.  

I'm not sure if i am misinterpreting the log and maybe it is switching to LDRXNZK because of knock?  I haven't tried lowering LDRXNZK very low to see if it changes boost.  Is there any other way to identify if that map is being used?

when i first started loggin, i was seeing about 4-5 degrees pulled for knock in cyl 3. Removing timing and adding fuel (although im not sure my fuel changes had an effect) seemed to correct this as i don't have any knock retard.

I'm stumped -- I know this is all probably very basic for you guys with more experience.  I really appreciated any help or advice you can offer!

one more thing -- when this log was taken, i had KFLDHBN very high -- somehwere overwhere 3 pretty much across the board.  I realize this is not necessary, but i was just experimenting.

Post your CSV log file because that graph is a mess :o

naw the graph is fine, it just needs RPM on X axis and to be a lot smaller

also put data with different ranges on different y axis.

sorry -- meant to post actual log file ..... disregard most of it.  I was looking specifically at ~61 seconds to ~65 seconds.

I can't get my filter to work on your file, what gear did you do this log in ?  Are you not logging throttle position or something?  Usual log run is in 3rd gear with pedal to the metal from 2500 rpm to redline.

edit: I don't think you logged gear position is why.

I'm 99% sure I logged gear. It was 3rd gear. I realize it's usually to redline, but I really just wanted to narrow in on this range to narrow down the problem.

KFTARX

ddilinger -- thanks for the response!  I did look at this table, but didn't understand how it could be causing this.  The IATs are under 20deg which would mean KFTARX would be 1.02 or 1.04, right?  Maybe i don't understand how KFTARX works?  I would think 1.02 or 1.04 would give me 102% or 104% of what was requested, no?

thanks again!

There are a few things that will limit requested load.

Namely:

KFTARX
KFTARXZK (if you're getting heavy knock)
KFLDHBN
LDRXN
LDORXN (if you have overboost codes)
LDPBN (if you're overheating, or in later files all the time)

Make sure all of these are set above the load you're requesting. KFTARX should be set to all 1's anyway, the same way you set LDIATA to all 0's. It keeps things consistent.

someone look at my logs since we are on the same engine platform, AWP, but with Chinese k04-022 with like a 22,000PSI WG actuator, tubular SS equal length turbo mani, 3" exhaust all the way back. fmic, stock MAF sensor with 3inch housing, 7.5lb flywheel, fluidampr, IntEng lightweight OEM size pullies, EV14 550's, Genesis fuel pump. There's alot more mods but I forget or don't think they are that relevant to the tune like an 034 motorsports catch can and shit like that. All emissions are a go, even equipped with a high flow cat. Right now I have a small EVAP leak and a small boost leak I haven't found yet but you can see it has no issues making boost, only controlling it. I imagine the leak is what is making the boost sporatic except for the spikes which is from the stupid jacked up WG actuator and my lack of adjusting for it thus far. The tune is a work in progress, I'm just now thinking the MAF is calibrated OK, but I'm at a stopping point with the tuning until I find the leaks.

2543 mbar @ 6234 rpm - not that bad for Chinese turbo. How is your impression about this turbo, is it worthy of these money? I didn't believe it could make more boost than K03s until I saw your log...

I understand you since I mess up and expiriment with my car a lot regarding throttle response, turbo spool etc.  :)
My first file was made with ME7Wizard and I remember I had problems with part throttle,not WOT though
I remember when my KFMIOP was way off KFMIRL the car didnt spool as it should and req =/= actual .

Since you are expirimenting to find if it is your Remap or hardware related I would do this.
KFPED= stock
KFMIOP=manually as 1:1 as you can with KFMIRL
KFMIRL=dont go from 160-200 I would do it more linear.

A k04 should hold that much boost fine, it just becomes more and more inefficient as the rpms go up.
So far, I say it was worth the $237US I paid for it off Ebay. It has about 2000miles or so and still feels like it did when I installed it new. I haven't checked shaft play yet.

I was thinking of putting this wheel on it and 5120 hack with a k04-02x turbo.
http://www.ebay.com/itm/290993307008?ssPageName=STRK:MEWAX:IT&_trksid=p3984.m1438.l2649

It is the stock wheel size, bolt on installation and prebalanced. 11+0 blade config, looks super efficient. Might be a good choice to experiment with. says it rated to 70 psi???a k04 boosting to 70psi? lololol maybe possible. I guess anything is possible..

I did get a bad turbo before the current one though from Chinaland, or it was my fault it failed lubricationwise and/or 30+psi boost spikes. Either way, It only lasted about 1000 miles.

Talk to Dan (littco) before you order anything off ebay. Chances are he can set you up.

You're right that should be my next step. :)

If rlsol follows rlmax, then you can stop looking at the torque model.

It should be easy to track down where the problem is by logging the right variables (%LDRLMX and %MDFUE).

rlsol_w doesn't ever meet rlmax_w, rl_w meets rlsol_w more or less (to the point where I don't think that is causing any major issues).

But that is my question.. why doesn't rlsol_w meet rlmax_w, I cannot seem to find any reason why it wouldn't in my file.  I will add the suggested variables to my log for next time :) 

Thanks Phila

Dude can you post a freaking graph and not a screen shot of a spreadsheet ;P

I've tried everything you suggested (almost) to no avail....
KFTARX set to all 1s
--no change

KFTARXZK (i dont think this is getting used, as i dont have knock) -- raised
--no change

KFLDHBN -- set over 3 in all values
--no change

LDORXN - raised
--no change

LDPBN -- have not located, but i'm not overheating.  Coolant temperature logged to a normal 180deg F

I tried a couple things just to see what happened...

Increased last row of KFLDRL to 95% (removed dip in mid-range RPM).  Created higher spikes (which is what i expected), but boost just surged right back down to 15/16psi and kept surging.

Increased last row of KFMIOP from ~82% to 98%.  This didn't appear to affect anything -- my request in KFMIRL and LDRXN is VERY high, so i didn't expect this to have an impact, although it's still unclear to me if the 82% in this last row would basically take my request, and reduce it to 82% (even if it did, with my very high requests, it shouldn't matter).

Thanks again for any help (and sorry to hi-jack the thread -- My thread was merged into this one)

I'm used to looking at it that way.  Besides, the logs are attached to my 2nd post in full. ;)

Sounds like you're seeing the exact same issue as I am. Very strange. 

i feel like there's some secret to this or something lol.  This can't be this difficult, there are TONS of 1.8T tunes out there!  -- Hopefully one of the experts will be able to help.

I didn't have this issue with my LP box that's for sure

I inquired via PM to littco like 2 months ago-ish on a hybrid set up and have yet to get a responce. The last I read, he responded to a request I asked for a ball bearing conversion for the k04, to which he replied he should have a prototype within a few weeks. this was a month or so atleast. I forget exactly how long its been but there isn't update in the thread on that either. I guess he's too busy right now.

Looks like you'll have to log rlsol -> plsol path... i assume you have ME7L working?

I've taken the IOP/IRL out of my working LP box, because when I checked the logs on that one (LP) rl_sol does meet rl_max just as it is supposed to, so maybe it will end up just being a funky IRL/IOP situation here and if that is the case I will know not to use masterj's ME7 wizard for that purpose again.

I will update either way.

IRL/IOP was my problem back then too.Your situation reminded mine a lot.

 I give more  emphasis now to KFMIRL/IOP than I did in the past.Also your KFPED some times is 100% at 50% throttle angle !! :)

Actually rechecking my RN file, I never did mess with KFPED or the axis, so the weirdness must be from IRL/IOP not being aligned I suppose.

With my issue i don't understand how IRL/IOP could be the problem.  I have last row of KFMIRL at 250 across the board.  I have tried last row of IOP at 95% (stock is about 82%) and really no change.  THe only thing i could think is you MUST change the last axis value (i believe it is 160 stock, but im not looking at it at teh moment) for KFMIOP, but from all the reading on here, that doesn't appear to be the case.

As far as i know, any IRL/IOP weirdness should show up in rlmax/rlsol.

if your rlmax/rlsol looks right, but plsol doesn't, it isn't IRL/IOP related.

I could be wrong, of course.

That is not an issue from my experience.  and in my LP file I left the stock IOP axis and scaled IOP numbers for my IRL that goes past 200 and everything was peachy.


Define right :P  since my rlsol is seemingly capped lower than spec.  Although I don't see any irregularities other than it seems to have an invisible cap.

Have you tried tracing MDFUE for what is setting rlsol too low?

I would start by logging the following additional variables:
milsol_w
etazwbm
etalab
misopl1_w

Based on the FR these seem to be the only ones that could be causing a cap on rlsol_w as rlmax_w is high enough not to limit it and rlmin_w is only a lower limit.

Also check what your CWRLAPPL is set to, should be 0

well i solved my boost problem...

so after trying out KFLDHBN change, i set it back to stock and went on testing other stuff... well... as it turns out, it was a limiting factor.  It seems that in order to fix my issue i had to increase last row of KFMIOP to 98% and increase values in KFLDHBN.  This seems to validate my thought that whatever % is in the last row of KFMIOP will, in fact, limit requested load.  I haven't gotten a clear answer on that yet....


Regardless, my boost is perfect now.... but, the car is WAYYYY lean even though requested AFR is around 12.  By way lean i mean over 14....  I'm not sure why.   Maybe hardware?  Any thoughts?  I can post logs in  bit... maybe its better for another topic?

Thanks for the help!

start by checking that you didn't cap ps_w...

It absolutely does not limit load this I know 100%.  My the last axis in my IOP is the stock 160 and I'm able to request 180+ without issue and it follows.  My issue with boost is now also resolved (need to rework IMX though for some reason now it's all screwed up again) but it was just a poor IRL/IOP scaling and combination and I know now to continue to scale them manually.

glad you solved it.  I actually was referring to the values -- in my file they are around 82% at all points in the last row. 

nyet,
i'm not sure how i could have done this.  I added fuel (i guess somewhat barbarically) with LAMFA and KFLBTS (which i now realize is stupid).  I didn't touch anything else.  I'm assuming my def file is wrong for KRKTE.  I checked that value just for shits and giggles and it reads 0.10301.  I would have expected this to be something like 25 or 30.  Maybe i don't understnad the value there.

Also, i did some logging tonight and the desired AFR hovers exaclty where i expect it... very rich -- about 11:1 yet my actual AFR is WAYYY lean like 17:1.  Am i possibly logging the wrong variable?  If my requested AFR is correct, is it still possible that there is an artificial ps_w cap?

Thanks!

^^I'm going to say you're not logging the right variable... check that first before you make any changes.

My file is finally doing what it's supposed to... I can definitely see that I don't have enough fuel @ the end past 5200 (stock inj w/ 4 BAR) so 550's are on the way.

I few more small tweaks to the N75 maps and timing and that's all she wrote.

I'm pretty sure you'll need to change the fuel pump to one that will keep up with the 550s, especially @ 4bar. I'm running them with a Genesis pump @ 2.7 bar(stock idle psi for my GTI) with an adjustable FPR and the set up works very well. I get awesome idle after adaption even though my TVUB and KRKTE aren't dialed in all the way yet.. damned boost leak is killing my trim's accuracy some.

My fueling issues at the moment are if I do a WOT run and let off the gas after instead of letting the engine wind down, my wideband shows 9.5ish/1 and the engine stalls. So I think I need to take some of the fueling out at high rpm in the decel map. I forget the acronym.

I would definitely move back down to the OE 3 BAR FPR, the the stock pump should be able to keep up then.  I'm not going to push the turbo over 20 psi between 3-4.5K... and I will keep the taper (stock DP + cat/ehxhaust).

KFBAKL/VAKL

The value in the KFMIOP table (NOT axis) is the highest torque that can be requested, yes.

Rick

Which ended up being exactly my issue.

On my file with the issue (if you look at my screenshot) IOP maxed out (incorrectly) at 82.9991 for the 190 axis 82.9 was like 163 load which was basically my capped load I was complaining about.

This would be useful info for the wiki IMHO.  I don't think the S4 guys run into this usually because the last row of IOP is around 95% on the s4 files i believe.

Correct. I will add it time permitting, unless somebody else beats me to it.

honestly, if IOP is scaled correctly it would never be an issue.  But I suppose it is "good to know" :)

Just to let you guys know, anyone can edit the wiki. That's sorta the point :P

:P

no waii! :P

I'm curious as to how you rescale IOP and IRL. I've attempted going off stock AWP IRL and IOP and adding 15-20% in IRL to the last three columns. From that I rescale the last 3 axis for columns in IOP. Ill effects from doing it this way so far. Also how does LDRXN factor into this? Until lately I've always thought it was simply a boost or load cap for a soft limp. I've read on here of some of you using LDRXN for load request instead of IRL.

I can't see how anyone is using LDRXN for load request.  LDRXN limits/caps IRL

Just use linear interpolation. Most of tables can be easily interpolated to any new axis. Just do it properly. Read here http://en.wikipedia.org/wiki/Linear_interpolation
It's easy to do it in Excel.

Bumping this topic to add it to my watched topics so I can find this for reference when I get home from work. 

As I have run into a similar cap where my requested load lower than spec/act and I am only seeing about 1 bar of boost pressure when I would expect to be in the 19-20psi range.