NefMoto

I have a few questions in regards to my logs:

Looking at my MAF, I am not sure what to do about it.  I think it is reading a bit high.  I have been contemplating scaling it down -10% but I am unsure if its really needed or a good idea.

The CF shooting up after I let off strikes me as weird.  Any thoughts on that?  My timing map is a bit aggresive and I have been dialing it back.

Also the knock voltages from bank to bank are all sorts of WTF...

That log is almost entirely useless.

Stay in one gear, go wot from 2.5k to redline.

I did a second gear pull and saw some things I definately didnt like.  I dont know if it was bad gas or what, but I nver logged CF that high before.  I will flash to a much more tame set of timign maps and repost.

 From the log I posted you dont see anything wrong with at least the MAF readings?  A bit high?

What I would think of as a better log.  I had to adjust the timing maps, some.  I also put the N75 on the external wastegate, and I have a MBC on the VGT actuator.  Boost a bit low on this pull, but it is safe.  I have been slowly adding to the KFLDRL.

I am still trying to get things sorted here.  I cant seem to get my fuel back down in spec.  Narrowband and the in cockpit widepand both show the car going stupid rich at WOT.  I just revised my tune: lowered min injection, lowered MAX load to be more in line with what I am runnign now.  I also am now running a MBC on the wastegate.  Max boost is set to around 24psi.

I have the VGT locked open to slow the spool down and to help with troubleshooting, so the LDRXN will have to be changed again once I can figure out how to fix the fueling. 

Help please.

Have you pressure tested your car? Are you 100% sure there are no leaks and it holds pressure?

If yes and it is going rich on high loads, then your krkte and tvub are wrong or your MAF scaling is wrong.
Good idea to scale one at a time.

Actually disregard that.

Your requested lambda is 0.69 - 10 AFR.
Log more and see where that is coming from.

Your CF when you are lifting can be because of fuel exploding and shooting through the exhaust etc...

I dont see where the requested went that low.  I checked it with xPLOT and I have been reviewing the CSV.

I did another log this morning.  This one is with the tweaks I mentioned earlier.  It did pull a lot better through the revs, but still was pretty rich.  The pull was in second gear.  Idle was a bit choppy at times, and I was seeing -10% corrections.  I am fine with that right now, since I am sure I am going to have to mess with the TVUB/FKKVS/KFKHFM once the fueling is all figured out anyhow.

so pull 10% fuel out of krkte.

Also, it looks like you are maxing out your MAF.

How do you suggest fixing the MAF?  Scale MLHFM back a few percentage that all the numbers are smaller?

Depends if you are maxing the voltage or not.

Right now I am using the Scaling from Tony's stg 3 file.  In what circumstances would I need to move the scaling around?

add...Last night while I was lgging I did see the AFR start to go lean 12.xx-ish towards the end of a few of my pulls.  Brainstorming on what you said and what I saw it kind of adds up.  What do you think? 

I think you might need a bigger MAF housing :/ but that is a good thing. it means you're making a lot of power ;)

What's this then?

I didnt say that the requested did not go as low as you said, I said I didnt see where it went that low.  Maybe my wording was off....REPHRASE:  Where did you go to see the requested going that low?  I normally look at the xPLOt logs and, and I look at the actual log data itself and I havent seen anything like what you said.

NVM...I found it.  Its towards the middle of the log.  Could that have been a reaction to the CF? 

Regardless, he's that rich when he's NOT requesting it, AND he's got trim issues.

Get KRKTE/MAF set up right first....

one thing at a time (IMO)

The log I did this morning has minimal CF and the requested looked more reasonable.

I pulled 10% from my KRKTE.  STFT at idle and cruise was better, but my WOT fueling is still the same. 

Do you have wideband logs? I don't see anything obviously wrong with that log, it actually looks like a good start.

I dont have any wideband logs.  I was on the group buy, so I will be installing the Zeitronix setup after work today.  I know you guys say dont pay attention to the narrowband.....but all the logs I have seen showing the narrowband voltages show them more in line than what I have beebn seeing.  I have moved my KRKTE all over the place and it hasnt changed much at all.  Right now I am at .04207 on my 630cc injectors, I havwe been all the way up to .053ish and the other than cruise and IDle being shot to hell, WOT was pretty much as seen in the current logs.

Is there anything else that could be affecting fueling like this?  I will pressure test the car later on today also just to be sure.

If you don't have a wideband, how do you know what your AFR is?

I havent been able to log it yet, but when I do have a standalone WBO2 setup in the cockpit that I use for reference while I am logging.  By reference I mean I glance at it a few times while doing a pull.  Thats how I noticed it going leaner on yesterdays pulls. 

But I agree, I need to get the new setup all wired in so I can overlay the logs into xPlot.  You think the narowband is really that inaccurate?  Evn though its actually a WBO2 signal being converted to a narrowband signal?  (Innovate LC-1 with two outputs emulating the narrowband, Innovate LM-2 standalone in cockpit for monitoring)

If you are using an innovate for your narrowband simulated output, why hook up a computer to it and log the WB values too?

Is that information useful without RPM ?

I'm not sure what you mean. What do you think the narrowband values are telling you?

Hint: it isn't lambda.

Better than nothing for sure, and you should be able to pick out where you did your pulls.

I just installed the ZT-2 in the car and now I am excited to get going.  I started the car and was freaked out because it took a couple of seconds to get a reading, and when it did the rpms were off.  Easy fix though, you just need to switch it from 4 cyl to 6 cyl in the settings.  I am buttoning things back up and will run a log or two in a few minutes.

Even though I havent figured out how to get the logs to display in xPlot, the data is priceless.  Damn!  I am runing stupid lean.  Its my own damn fault though, I had been adjusting the AFR based on the narrowband.  No wonder the car was feeling like I was hitting a brick wall around 6k under full boost. I will attach both logs in case someone can help me figure out how to graph them together.  ECUx CSV, Zt-2 CSV, and the original ZT-2 log attached.

For the ECUxPlot aspect, see this post here:
http://www.audizine.com/forum/showthread.php/479416-****-Complete-Guide-on-How-to-Take-and-Graph-Logs-****?p=7462266&viewfull=1#post7462266 (http://www.audizine.com/forum/showthread.php/479416-****-Complete-Guide-on-How-to-Take-and-Graph-Logs-****?p=7462266&viewfull=1#post7462266)


For the Zeitronix aspect, see this post here:
http://www.audizine.com/forum/showthread.php/493904-Huge-Post-on-Wideband-Info-and-Heads-Up-on-GB?p=7824356&viewfull=1#post7824356


In ECUxPlot, I generally first open my ECU logged file, with the filter set up.  Then I go to File, add file, and select the Zeitronix .csv that I exported.

What a mess.  I have been dialing in the fuel getting the WOT fuel to match the requested AFR at the end of the run.  I now have way too much fuel everywhere else, and my injector duty cycle is getting pretty high up there.  I really dont like messing with the KfKhfm map to dial in the fueling.  I would much rather deal with the FKKVS map since from what I have read it deals directly with the injectors themselves.  I have been staring at th FKKVS map for days trying to figure out what to do to it when the time comes.  (staring=searching and reading)

i generally do all fueling with krkte, tvub, and MAF tables. i have yet to need to adjust fkkvs. ymmv.

I was originally using the KFKHFM map before, I just didnt like how it started moving load around....second thought, Maybe I need to move load around consodering my plot tapered off pretty steep after it hit the peak load.  What map do you think I should use along with the KFKHFM to get the fueling lined up better?

Getting the part and idle right is going to be a pain since I dont use LTFT.  I havent had much luck tweaking TVUB ogg of STFT.

I opened the VGT some more so the spool would be really long.  I am having a hard time seeeing why the AFT goes so low.  I am adding more fuel to match WOT at the end of the run, but that massive dip in AFR is really bugging me.  The AFR at the end is just the VGT being wide open now, and I think it is flowing a lot more air in its new position.

If you are adjusting more than 3-4% in KFKFM you have something very wrong with your maf or fuel system.

In the above log kfkhfm was all 1.0 IIRC.  I just made changes to it in the middle rpm/upper load areas that corresponded to the rich area in the log.  I think my lowest is .94xx-ish.   Can you take a look and tell me what you think?  I just made these changes to the file after the above log.

I just recentered kfkhfm, and took my krkte and tvub back down.  I think I am heading in the wrong direction, and need to look at why it is going lean at the end of the runs vs.  why it is going rich.  I keep looking at the fuel curve, and it is following the requested curve (just a whole lot richer) until the middle of the run, and then the AFR goes lean.  With that observation I think I would be better off (easier) getting the part that is actually following in line with requested, and use those KRKTE and TVUB #'s for the rest of the curve.

Once I get it to line up, what should I do to get the lean area at the end of the run dialed in?  It seems as if the MAF is loosing track of how much air is being pulled in, but to tweak the KFKHFM to get those areas dialed seems a bit excessive.   

My rationale is that the most common reason for actual not meeting requested AFR (assuming the car is mechanically 100%) is that the MAF isn't metering the air correctly, as you said.. So it makes sense to me to tweak KFKHFM until it is right. If your fuel system is 100% linear (or at least close enough), once you have actual AFR matching requested AFR, you know you have your MAF right... which is a good thing, since MAF readings are used for all kinds of things in ME7.

So really, what I'm saying is, getting actual to match req AFR is kind of a secondary thing; what you are really doing is calibrating your MAF :)

I mean, if all you cared about was actual AFR, you could just mess around with req AFR until actual is what you want. But that would mean that something isn't right; either your fuel system isn't linear, or your MAF values are completely wrong.

I see a lot of noob tunes where the req AFR is totally completely screwed up, but they don't care because the actual AFR is more or less what they want. I think that is a mistake. Again, that is just my opinion.

Agreed.  Thanks! 

Something is really bugging me about this setup right now.  I didnt get around to pressure testing the system over the weekend.  I am going to try and get it done after work to day for my own sanity.

I did a pressure test today and found two leaks. One from the Map sensor.  I guess I forgot/lost the o-ring from whe I got the flange welded to my charge piping.  The other was from a small crack in my intake tube.  I sealed both of them up and the car was noticeably smoother. 

Before and after log.  Note the max MAF readings.  I think it was sucking in unmetered air hence the lean condition.  I am going to recheck it tomorrow.

I don't see a lean condition in that first attached graph...

A few graphs up.  I have been tweaking the fueling to counter it, instead of actually fixing the problem. 

Did some more boost leak testing and I think it is all sealed up now.  I added a little bit more to the KRKTE and then I think all the rest will be done with the MAF maps....once I make sure idle and STFT are ok.

I had the filter configured screwy...... :o 

Still something going on at the end of the run.

Out of injector?

Post injector duty and MAF...

I have plenty injector left.  I am running less boost since the VGT is wide open and spool is so slow and late. 

The lean condition happens consistently at 150-ish% load.  My KFKHFM map only goes up to 159.750.  Does the axis on this map or any other map need to be changed to go past the load that i am going lean at?

Hitting the load limit generally isn't a problem. Hitting MAF or injector duty limit definitely is.

from what I have seen the injectors arent maxed, and at the boost I am running now ~15psi max with a very very slow spool up, I dont think the MAF is either.  The logs a few pages back, when you said the MAF is porobably maxed I was running 23-26psi.  The injector duty cycle back then was around 60%.

Pressure test again; the only reason to go that lean with that duty cycle is a post MAF intake leak..

YESSIR!

Cross reference your fuel maps and your actual load at 15 psi. You might not be getting into the tuned parts of your maps.

Fuel maps like AFR versus LOAD? 

Which maps? His req AFR looks fine all the way to redline...

Which leaves MAF tables and FKKVS..

That, and I think all the charts I posted show the AFR versus the LOAD.  The FKK and KFK maps are still all 1.00.

Three more leaks found.  One silicone coupler from a slightly out of round pipe, brake booster line was rubbing against something and worn through, and loose hose on the fuel pressure regulator.   :o  Pressure tested to about 40psi and held, so hopefully that fixes the issue.  

how do you deal with the air that rushes out the oil fill hole in the valve cover? I hooked my compressor to the FPR vac line and air flows freely out the valve cover oil filler hole..so frustrating because I can't hear anything else but that. I've done the soapy water bottle trick also but don't know where to spray...cant hear to spray...so I just spray everything!! and then soapy mess on everything..
 I know you clamp the hose that connects to the pressure regulating valve (hocky puck lookin thing connected to the TIP). My pcv system is 034motorsports silicone everything including their billet PVC valve and full catch can kit. sprayed all of it and no leaks there. checked the pcv valve before i installed, it it's a check ball design good for high psi and shouldn't fail anytime soon..

http://www.034motorsport.com/engine-components-18t-check-valve-18t-27t-aan-pcv-billet-p-21834.htm (http://www.034motorsport.com/engine-components-18t-check-valve-18t-27t-aan-pcv-billet-p-21834.htm)


I thought about timing and if some of the valves are open because the cam lobes are pushing them..allowing pressurized air into the cylinders..and blowby.. i just thought about this as I'm typing...maybe if i turn the engine over while I listen to the air rushing out I can determine when all the intake valves are closed or as close to that as possible. maybe this will do something to quiet the noisiness so i can be a little more strategic in spraying soapy water for boost leaks.

It sounds like you are pressurizing the crank case.  I dont think it matters where the valves are at/cams are at.  The upper head are where the cams and valcovers are at should be completely sealed from pressure from the valve piston/combustion events.  When I had a pcv system in my car I would do a check with the PCV line pulled from the intake tract and the both ends capped.  I am kind of foggy, it has been a looong time since my car had the pancake/pcv valve setup in it.  Maybe pull the PCV lines and cap both ends. 

Looks a lot better.  I think fixing the leak at the fuel pressure regulator was the key to fixing the lean under boost conditions.  :o.

 I thought if there was a leak post MAF the car would run rich under boost and lean under vac.  Either way, I feel a lot better now, thanks fro your help thus far! ;)

Forgot to re-adjust the filter.  See second log.

One more showing the MAF.  Nothing in the file changed.  

AWESOME! now just add a bit of KFKHFM at those lean load/rpm points and you're good to go.

OR it might be a better bet to add 5% krkte first... having your trims pull a bit of fuel part throttle is never a bad thing.

Add?  I just made a map where I pulled from kfkhfm.  I thought I was reading rich, and by pulling I was making it lean out the major dips.  I was trying to focus on the what I thought was overly rich areas, making the line smoother.  I left the higher rpm area alone since it was already starting to head towards desired AFR.

Ideally, you want actual to match requested...

I must not be reading the chart right.  The blue line is actual and red line is the requested.  From what I saw on the chart the actual was quite a bit richer in  most areas.  Am I reading it wrong?

I little tweaking of the KFKHFM.  Still more to come.

No matter what I do, I can't get the car to stop leaning out at the top end.   ??? Last ditch effort is to finally add in the fuel pump setup I made.  I really thought the fuel injector duty cycle would show signs of me reaching the limits of the current pump.  For now I will turn the boost way down and run safe.


(http://i360.photobucket.com/albums/oo43/SpeedWorksAutoTuning/IMAG0080.jpg)

I don't get it either. It looks like you are running richer than requested.

Oh, sorry I didnt upload the latest logs.  All the previous logs were in 2nd gear to keep boost, load, speed modest with the VGT openned up.  I did a couple of pulls in third gear over the weekend, and it was pretty bad.  Around 5k the trims would just lean out really bad in third gear with the turbo at a 1.85-ish AR on the hotside and probably geting close to 23psi about 210 in load at end of run.  IIRC all the previous logs since only got up to 160-170 in load. 

Wow. You're right. There is something VERY wrong with your hardware.

You sure you aren't sucking in unmetered air?

I pressure tested it over and over again before I left for the weekend and it was sealed up tight.  I will test it again after work today just to be 100+% sure.  

The only other thing I could think of is the fuel pump.  I always watch the injector duty cycle to see when it would be time to add in the new pump since I am not looking forward to adding it in just yet.    

Under boost, if there was a leak wouldnt I go rich?

IDC will not give you any insight on your pumps performance. It is only percentage of injector on time during combustion cycle.

Damn.  I dunno why I thought the duty cycle will skyrocket indicating its time to swap out the pump.  Its looking like along with putting in the new pump, I should get a boost and fuel pressure sensor to go along with my Zeitronix setup.

True, but I've seen strange ME7 side effects where if you are near the limit of load, even if your MAF *AND* req AFR are good, the ECU inexplicably *still* pulls IDC. I have yet to track down the source.

I doubt this is happening here, but keeping an eye on IDC is always a good sanity check nevertheless

Huh?

IDC is calculated from the calculated required injector on time. This will only show you if your injectors will be able to deliver the required fuel mass in the amount of time available.

Agreed. On time/IDC aren't distinct.

If you like replace all instances of IDC with on time (given that you know you are below 100%) in my post.

I'll dig around my logs if you like. Maybe you can tell my why i get a droop in IDC/on-time even though req AFR is flat, and MAF looks fine :)

By droop you mean shorter on time?

If so, shorter on time while target AFR remains consistent or even enrichens?

yup.

I would assume that it would have to be somewhere in the injector calibration or possibly LTFT.

Start a thread with some logs. We'll figure it out.

Is there something in the file that will make the ecm read only from one primary o2 and trim both banks?  I still cant seem to get the STFT dialed in just right.

From my logs, I think I see the banks compensating for each other.  One bank will read -13% the other will go +17% flip flop one adding and the other pulling fueling on part and cruise.  I am assuming it is becasue the reading is coming from one sensor in the one downpipe.  There isnt a real left bank - right bank reading anymore.  I dont really know how to explain it, but I think the ecu needs to know there is only one sensor in one central location and to trim both banks similarly to achieve its target AFR.

Timing:  This is another one of my areas where I feel I am not informed enough to make a good map(s).  I have played with the two maps and have had "good" results.  I managed to get the cf's as low as -3, but in the high load high rpm are somehow I would always get a major amount (-8 cf) of timing pull.  Maybe it was from my previous lean condition, I dont know.  I reverted to a totally safe map, and kept digging around for info on how to handle the timing.  I dont really like guessing and dont think I was headed the right way.  

I "know" the timing maps are fro the different cam states, but from the tunes I have seen most of the tweaks are dont on the KFZW map, with a few on the bottom right hand corner of the /2 map.  I cant seem to really follow which map gets tweaked when.  I have absolutely no timing pull whatsoever since I am running such low timing right now.

Slightly modded stock mbox timing right as of right now: 

Timing will always be a trial and error thing:

Add timing till the dyno says you're not gaining any more power, or your CFs get too high :)

Its trial and error as to which map gets edited?

Sure. Or disassemble to find out which is used under what conditions.

Typically, the areas you will be modifying (wot) will only EVER see one cam state (at least this is true for 2.7ts), so it is even safe to have both maps match where you are modifying them.

Cool!  So, I can make all the changes in KFZW and copy them in KFZW2?  Do you do any changes to the optimal timing maps?

Only if i'm doing axis changes.

This is not optimal. Cam state does change during WOT and ignition angles can be safely advanced further in KFZW2.

Log fnwue.

When you scale the timing load axis, that axis is used in KFZW 1/2, KFZWOP 1/2, KFMIOP?  It looks like the max value is also used in the KFMIRL map.  Does the LDRXN map need to have the highest load refelective of the timing load axis too?  

I plugged Berttos stg 3 IOP and IRL and load axis info into my map and it made a really difference as far as driveability was concerned.  I changed the timing load axis to match, IOP, IRL to match his file. It actually felt like a higher powered car should. The flipside was I didnt really understand what the changes he did were for, and wasnt really comfortable using the maps not knowing if I needed to change anything else so I reverted back to "safe" again.

I dont understand why there is such a large descrepency in the highest load values between all the files I look at.  I have files that range from 219 load to 248 load.  

Guess that means I need to get on board with that other logger and quit ECUx....

but at a given load/rpm pair during WOT, is it possible to see both cam states under different conditions?

marcellus - have you moved to the k-box or considered it?  i'm convinced the RS4 timing values and change-over would work perfectly for your setup.  this would eliminate so much variability in your tuning process.  audi knew what they were doing for sure... i've spent so much time look at the s4 versus rs4 timing, and i'm convinced rs4 is the way to start.

I have considered it, but I am waiting until you guys do a little bit more ground work.  I am having a hard enough time dealing with this setup as it is, and don't wanna introduce a bunch of extra variables that I can't get at least a pointer as to which direction I should go to fix it.

I do like what u have read about the timing and fueling maps.  The larger maps, for lack of a better word, will surely help me fine tune this thing past where the m box tables end.

I'm not sure how conditional it is, but mine has always been predictable so far. High load middle of the RPM band.

Kill cam change over until everything else is dialed.

I have the VGT working now and the results as far as the AFR are concerned really have me bummed out.  I have been trying for days to get more fuel in the higher load areas.  I have been adding to the KRKTE and KFKHFM in the high load high RPM areas and the best AFR I get is just under 12.7x.  


Three logs of 2nd and three of 3rd multigear pull.  Ignore the repeat AFR on the 3rd gear run, I couldnt get it removed with the filter settings.

I can think of two things to try (since i've had problems when actual load is too high with both torque intervention and fueling problems)

Try underscaling your MAF so load doesn't go over 210 (you will have to add krkte to compensate)

Take a another look at your IOP and log the appropriate torque intervention variables

Subtract a % from the whole maf volt linearization map?  That's what u mean by underscale?

My iop and irl and pretty much copied from berttos stg 3 as of Sunday.  I think I got a bit too aggression with the timing, and I had the vgt base opening set to small.  I think the combination of the fast spool, the high timing, and the iop/irl I was playing with pissed the ecu off.  I had crazy torque monitoring codes that wouldn't clear until I pulled the power to.the ecu.  It kind of had me freaked out since I was practically disabled in traffic a few times.  

Unrelated, but what are you guys doing to xPLOT to get the calculated whp, wtq to be more realistic?  I went from about 280ish whp with vgt wide open to over 600 whp with it semi functional.    ;D  I wish! 

Fix your hardware first THEN tune.

Your fuel pressure is most likely tapering. Attach a fuel pressure gauge, stick it out from under the hood and floor it.
The pressure should be nominal + boost. So with a 4 bar FPR and 1 bar boost, it would be 5 bar.

I will go double check my Fuel pressure.   Last time I checked it was fine.

How exactly did you check it? If it's fine at idle, does not mean it's fine under boost.
You must be able to check the gauge while driving the car.

With a gage while driving.  But, I did it when I first "completed" the engine swap.  Its possible I am losing pressure, or not getting enough flow now that I am pushing it a bit.  I dont doubt anything right now.

Use the right rpm/mph for your gear and make sure you are logging level.

If it doesn't give you consistent results, you are definitely doing it very wrong.

Yeah so check again at the level you are running it at now.

Oh, maybe its the fact that I havent figured out that part of my tune yet.  I didnt get around to doing the AIMVM or whatver the map needed are to get propper MPH logged. Also I cant log gear past third.  I havent figured that part out either, and havent really seen a need to yet.  

I am not running the 6 speed on my swap, I am running the 5 speed from the 2.8.

logged mph is NOT used for hp estimation. it is assumed you'll get the rpm/mph right (in ecuxplot vehicle profiles) for the gear you are logging in.

OH, the whp/wtq calculator needs to be done in third gear and the rpm per mph needs to be set up properly for my third gear?

Yep. ECUxPlot has no real way of knowing your gear ratio, and it can't assume the MPH reading is right because 99% of the time people aren't running stock tire sizes etc.

I "scaled" MLHFM a tiny bit.  I think the results are pretty good.  I didnt really know what to do so I took around .5%.  TunerPro was being a biatch so I couldnt get it exact, but relatively close.  I didnt scale the whole map just from the .3027 area, where I thought it was past idle readings. 

It was all just a really big guess, I have never rescaled a MAF before. 

Is load still maxed?

sorry, didnt post that one.  I am going over teh scaling one more time and I think I should have just multiplied the damn thing by the percentage of the value I wanted to run.  I actually confused the piss outta myself trying to get decent numbers.  I should scale the MAF back until I am about where in the Load?

 I am thinking I should recenter my KFKHFM, and dial the mlhfm until I get close to the WOT AFR. 

Naw. I dont think that is a problem (from looking at your last log)


Definitely a good idea.

I'm wondering if your MLHFM is screwed up somehow. Can you go back to stock and then make sure you scale the whole thing by the same percentage?

The only remaining possibility is that your fuel hardware is messed up.

I started off with Tony's scaling.  Its probalby wrong for my setup since in his tune since I think he said it was for a stock airbox.


I still need to double check my fueling, but the small scaling I did to the maf last night does look promising.

My borrowed IRL and IOP maps have load values past 248.  Is that an issue? 

Also, I am running quite a bit less boost than I plan to.  Isnt my load going to go higher once boost starts getting added in?  I am asking becasue you said looking at the last chart load was fine. 

No, actually, what I was looking for was if keeping the load low would get rid of the AFR problems. It did not, which leads me to believe it is hardware related.

You dont think the scaling helped?


NVM...I will get the gage hooked up.

I have requested 280 load without issues. There is not much reason to underscale load apart from the ECU's, which do not have the timing axes to accommodate it, as you are running the last 1 bar of boost off of the last timing map cell.

I rescaled the Maf from a bit from where Tony's file was set and it helped alot in drivability.  The car doesnt stall AS MUCH when rpms drop fast.  I still have yet to figure that out.  I blame the LWFW.  The trims are are a lot more stable. Idle settles nicely with a bit of misfiring after a while of running from the fuel trims fighting each other.


ADD:  second gear

That fueling curve is drastically different. Was your original MAF curve plain wrong?

Car stalling when RPM's drop is in 99% cases wrong TVUB.

I really dont know.  I am finding out stuff as I go along.  I do know my MAF setup is quite a bit differnt than a stock airbox and 85mm housing.  I am still learning right now, so I try and be very cautious as to what changes I make to a base fiel, and log what the change does for a long period of time.  So far the MAF scaling was by far one of the best changes (asides from the being able to log actual AFR).

OH, I made two changes to the tune to get the above logs.  I lowered the KRKTE some to match the mid rpm fueling up better also.  

I tried messing with the TVUB before using the calculator, but I think I was trying to do too many things at once and was chasing my tail.  I dont use LTFT anymore so from all I have seen, I dont know how to get a good TVUB.  I mean, all the calculations I have read about are based off of the LTFT.  If I start to use LTFT again, what I see happening on the STFT gets amplified so bad, that by the end of a trip the car wil barely idle because of the -25% one bank and +25% on the other. 

I revision later and I have this:  Still second gear. 

Well done.

Now it's time to quit running baby boost and crank it up.

I have a quick question.  I have been palying with the MAF scaling and have been "underscaling" it to get the fueling better.  I have noticed that each time I scale it, the car progressively gets slower and slower.  I flashed back to a MLHFM that I used a while ago, and it was night and day as far as performance...(logged,recorded, FATS)
 
http://nefariousmotorsports.com/forum/index.php?topic=2444.msg23615#msg23615

I have been following this thread along with a few others.  Apparently there are other maps that need to be messed with to go along with the underscaled MAF.  Thinking about it, I came up with the timing maps kfzw/2.  What else needs to be gone through to compliment the MAF scaling?  

In the thread above :  "It works more or less OK, but of course spoolup could be improved by properly using the PID and underscaling LOAD takes a while to dial in, even if you only do it up top."- PRJ.   At what area would you start underscaling the MAF so you dont mess with the idle/cruise?

 Right now I think I am overcompensating in the KFKHM map trying to pull fuel out of the 4000-5500 rpm range.  If I pull from KRKTE I start going really lean again.  I am seeing that my best fueling is when the load was around 190ish. 

You can start at about 120-130 load and progressively go higher.
Do the same with KFLF to offset it.

Set the KFKHFM map to just linearly taper down. Do all mixture corrections via KFLF.
Rescale KFLF axes for more RPM, etc.

Slower, how?

There are only 3 things that affect power

boost, AFR, timing

Which is changing when you rescale, and why?

I can see the timing getting lower and lower as I rescale.  Still doing most of my logging in second gear, i have gone from a 1.4x ish Fats to a 2.4ish.  Its a difference I can feel every time I rescale.  Does the IRL and IOP need to be moved some since they deal with load also.  Maybe I am not requesting enough torque at the new lower load values?  Requested torque does have to do with timing and boost correct?

Unless you are getting torque intervention, all of your timing is from the timing maps (KFZW)

Get your fueling/load where you want it, then fix up KFZW. Stop worrying about your HP and get everything where you want it first.

You are over complicating things by worrying about power output.

I am not worried about the power output right now, I am just noticing a trend.  The more I scale to get the fueliong where I want it, the more it's starting to seem as if I am going to have to do some adjusting to other maps.  I went the other way with just the MAF scaling just to verify, and each time I scaled back up, more and more timing was restored, five degrees in some areas. 

Right now I am in the high .8 range in the mid to highest load areas of KFKHFM 3500-5000rpm.  Its making me thing I need to make adjustments somewhere else since its such a big deviation.   

You will have to tweak timing REGARDLESS of MAF scaling.

So pick a scaling, and fix your timing. Done.

Three logs from today.  I started with the "standard" MAF scaling from Tony's tune, then underscaled it, and underscaled again.  I am runing quite a bit leaner now that i turned up the boost a bit.  Fats from each run are 1.79, 1.99, 2.05.  As I scaled down, the point at which the car went lean shifted further towards the right....happenned later in the rev band.  Also, the more I underscaled the more timing was sustained throught he mid range.

So, I guess all I am asking really is if this is normal, or is this something I should be digging in to?  

Definitely not normal. Do you have a fuel pressure gauge?

I do, but its the mechanical one thats a PITA to set up (duct tape to the windshield while running type of crap).  I just used it to verify FP when I was last having fueling issues.

Your MAF is maxed out. Have fun mapping KFLF to last cell.

Yay. We are officially getting nowhere in this thread

http://nefariousmotorsports.com/forum/index.php?topic=2112.msg21794#msg21794

I agreed with you long ago that something was maxed.  We also werent sure if it was hardware or tune related.  I checked over everything you suggested in order so I can narrow down the cause.  I just like to make sure everything is in order before I make drastic changes to the file.  I make the changes small and over a long period, datalogging the changes for days before I post more results or ask questions.   

I scaled the maf down some more and added KRKTE, and the fueling is fallling in place again in the end.  Fueling is retarded everywhere else.  With the prvious scaling no matter how much I added to the krkte the higher rpm/load areas wouldnt enrichen past 13.5xx.  I guess that was my indicator that the something was indeed maxed.  I couldnt be 100% sure until I checked all my hardware. 

As Nyet stated before, maybe I should go to a larger housing.  I could have swore you guys are running more power than I am on the 85mm MAF housing.  I am only running 22.5psi right now.  Up from the 16-18 I was running from the logs a page or so back.  How does )#$ get their setups to run 600+awhp on a 85mm on stock ecu?

I dont even know what to do with KFLF.  Nahalem said he set his flat on his thread when he was troubleshooting his starting issues.  When I was using the Maestro I could have swore we were using this map to edit fueling.  But, that was also 1.8t, mafless, wideband ecu....I dont know if it really translates over.     

I don't think you need a larger housing. If, when underscaling the MAF, you don't go lean, you aren't maxing the MAF voltage wise.

You could still have an FPR problem etc.... or possibly sucking in air post MAF.

I pressure check the thing almost twice a week just to be sure.  Sucking in air post maf would make the car run lean/correct more in idle and in boost run rich since the air would be blowing out the leak?  Or, maybe not getting enough boost to the reference line on the FPR so, going lean?

Trying to wrap my head around this timing thing too.  Since I am shifting load down by scaling the MAF I am now using more of the load/rpm sites to the left of the KFZW/2?  Thats what it seems like is happening.   

That would be a post compressor leak. If you are running fine idle, but not under boost, you could be sucking in air pre-compressor


Yup. but if underscaling your MAF fixes the lean condition... thats probably not it either


Yes. Don't worry too much about the part throttle timing. Worry more about the load/rpm path during WOT pulls.

If you look at his graph it is pegged at 400g/s. But this could of course be because there is a problem with fuel pressure, and he just raised the MAF to compensate (maybe time to get that damn gauge out).
No underscaling is going to fix this I think, because mshfm_w and mshfms_w do not really cap out. rlroh_w does not either.

Easiest way to find out if you need to underscale is log ps_w, rlroh_w and rl. If ps_w is close to 2550 or you are already at the point where rl caps out and stops following rlroh_w, then you need to underscale.

Two issues I have now fixed...or revealed.  First was my collant temp sensor was jacked.  It was readng -45 deg, a lot of the time causing really hard starting issues, and some issues with fueling along the way.  Secondly, I never really graphed the spec and corrected spec load.  I was riding right at max spec load (ldrxn) and have been gradually adding to the limit in the areas that I exceed the limit.  Now the maf isnt flat lining around 400g/sec.

Last, I will hopefully get to install the new fuel pump and gauge setup this weekend.

I decided to eliminate a variable.  Since my turbo is a variable AR turbo I figured it might be best if I put it in a fixed location for now.  The last few logs the VGT has been able to open and close on its own using just the pressure from the exhaust to force it open and the actuator I have on there was disconnected from any boost reference. 

The logs look quite a bit different to me.  These next two log are from when the VGT actuator was refernce fee and acted as minimal restriction against the vanes opening up.

Spool is slowed due to the fact that the vanes are slowly opening.  from 3500 to 4k is pretty quick and then th rest is really gradual. 

Now with the VGT kind of locked in minimal a postion as I can get before being closed down enough to acually act as a exhause brake.  Each pull I got it closer and closer to the edge of choking the exhaust off:  Two pulls, all third gear. 

The "reworked" logs are a modded file fom the other runs.  The MAF scaling is a bit less, I took off some of the wacky numbers I had in KFKHFM mostly at the high RPM high load areas, and I made a slight addition to KRKTE.  NO real big adjustments were made to the tune since I do everything in really small increments.  I was just testing the waters to see if it was a good direction to be headed.  Looking at the difference in teh MAF numbers, I think the VGT has a really large impact on how the hell I am going to get this thing tuned.  Way over my head, even more than before.  

I think I can get it to act ok in a stationary postion, but what do I do once I start using the VGT again?  Right now the turbo is falling flat on its face in the upper end since it is choked down so small.  With it wide open the thing spools like a GT35 on a 1.8t.  In between.....look back a page or so at the logs.  

I think I can add back to the KFKHFM to sqaure the MAF, maybe add some more to the KRKTE.  I like the MAF readings now, but it is only because the turbo was really restricted.  WHat really has me worried is the timing.  It is worse on this last run than it has ever been.  Pulling more at the richer AFR than it was when it was running in the 13's for AFR.  I am going to o back to my "safe timing maps"  to be sure timing is messing with fueling for right now.

This Worked GREAT!   I was screwig myself over messing with the KFKHFM map trying to add fuel in the top end.  From the logs all it seemed to be doing was skewing th MAf readings not really adding fuel.  

I started from all 1.00's again in the KFLF,KFKHFM,FKKVS, and I also took off the MBC and am trying to dial in the N75 to my TIal WG again.  Last time I tried t use the N75 I failed miserably.  


This time I took a different approach and re-analyzed what Nyet said in quite a few threads.  Something like " get your actual AFR to ROUGHLY match your requested on a full throttle run"  SO, I go the shape of the profile to match just by doing my own scaling MAK MLHFM from tring from the stock M-box file.  Th AFR's were not in line, but the overall shape was ok...well until I got to the trouble spot after 5k and the car was still running lean.  (fuel pressure checked out OK)  

One thing that was really messing me up was I messed with the KFLF to where it wasreading in AFR.  I dont think that map should be read as AFR, but as a correction factor.  Maybe thats why it threw me off so much the more studied it.  With it set back to how it was originally I can see how the stock KFLF was actually dialed in pretty good.  All the areas I have been trying to lean out using KFKHFM (part throttle, cruise, idle)  were already dealt with using this map stock.  I had some rediculous corrections in the wrong KFKHFM trying to pull back fueling from a too high a a KRKTE.  Again, KFLF also had the highe end taken care of since it had a "additive"  correction to the high load high rpm area.

I guess what I am trying to say is, I have been chasing my tail all this time, and should have just used the KFLF map to add the fuel needed.....for now.

Thank you guys !

I am looking into how to change the axis values for the KFLF map to cover more areas, and use this map for all my fueling corrections.
  

N75:  I would really like to get the N75 to work properly and use the MBC on the VGT actuator.  RIght now I have a massv spike and then boost drops to 17 psi.  I think my am running a 13psi spring in the TIAL.  

I know I can get it to hold more boost after the spike by adding to the KFLDRL on the last few rows/columns where I have the duty cycle capped off pretty low.  I just cant seem to pinpoint where to get the spike to not be so high.  At least not right now.  I wouldnt mind running 26psi, just not while I am still figuring stuff out.  Around 20psi will be fine.

SO, I am thinking bring down some of the values in the 4000rpm range to help with the spike and raise the values in the rows below it to bring the boost back up to spec.  Add a bit to the 3500 row towards the right end since I appear to be above my cracking pressure at that point and want to keep the the WG to keep it shut longer to keep ot spooling fast. 

I got fueling pretty much where I want it.  I need to get bigger injectors for sure since I am well over 100% IDC after 5k at anything over 20psi right now.  My timing is really puzzling me right now.   I see around 5 degrees timing pull vs. actual.  Next file I pull 5 deg. more timing and the ecu pulls yet five more degrees vs. actual.  At the moment I am down to about 2 degrees timing at a requested 7.5 degrees.  This is coming from a file that was requesting like 12-13 deg and pulling to 8.5 deg.  What am I missing here?  I even lowered boost again to get it in the "reasonable" range  ~23 psi.  At WOT my fueling is damn near flat at 11.8 after a slight lean area at 5k that I am gradually reeling in.  My IAT are in the 70 deg range. Attached files are subsequent revisions. 

got a wideband? maybe you're leaner than you think?

Also, might want to log the various knock control variables to see where it is coming from.

If, opened to the largest aperture it goes, the VGT is enough to keep boost creep in check and not overboost, why don't you use just the VGT to control the boost (and get rid of the wastegate)?  It seems redundant and really complicated to try to control boost via both a wastegate AND the VGT.

Is the ignition retard accounted for in dwkrz?

If not it could be dzwwl or wkrdy.

In my experience, dwkrz is usually about double the correction required. For example, with 6* retard, pulling 1.5* will usually bring that down to ~3* retard.

Yes, all logs now have the wideband and desired. 

I couldn't get the VGT under control.  But, that was a while ago when I last tried it.  The actuator opens too fast and boost fell off.  That was also before I discovered that back pressure was forcing it to open and was slowing spool.  This is something I might have to revisit, but I still need some sort of safety since the vgt is known to get stuck

As safety you can use the N249 for example.

And you could use a pulsed solenoid to control the VGT as well.

Now I have to go Google "pulsed solenoid".  I do have a restrictor coming from one of my vendors here at work.  It is supposed to slow the pressure getting to the actuator have unrestricted flow in the opposite direction, fully adjustable.

I dont have the N249 in my car.  It probably wouldnt take much to hook it up though.  I dont remeber for sure, doesnt the n249 use a vacuum res.?  I will have to find one of those and stuff it in the engine bay somewhere.   

I don't know if this was discussed before, but can't you hook up the vanes to a wastegate actuator? I think someone did this before. Worked rather well.

Yessir, that's how mine is set up.  Pics from a quite a while ago. 

(http://i360.photobucket.com/albums/oo43/SpeedWorksAutoTuning/IMAG0434.jpg)
(http://i360.photobucket.com/albums/oo43/SpeedWorksAutoTuning/IMAG0435.jpg)

It works OK with just the actuator, but from the logs it looks like it needs a more precise means of control.  From what I have seen the wastegate actuator opens too fast and will kill the spool.  On my setup, the ideal opening is at peak is a close to peak boost as possible.  I am going to try and revisit the N75 on the actuator tomorrow.  I gave up on it last3 time because I didnt realize the actuator didnt have enough preload to keep the VGT close, so I thought the N75 wasnt capable of doing what I wanted.  Boost was really slow, and I blamed the n75.  Recently I have been palying with the n75 on the wastegate and I machined a shaft that allows me to add a bunch of tension to the acutator without pulling the VGT shut.  It has been working really awesome.  Spool is really nice, and I have the n75 working really good keeping boost in check.  I can see in the logs where it starts to open up.  The boost will skyrocket up to about 20psi and then take another few hundred rpms to build to around 25psi.

As I sit and think about it, the N75 might be what I have been looking for.  It can allow fro small amounts of boost to get to the actuator and hopefully let me control how it makes the VGT open.  I def like the idea of using the BOV to help regulate overboost via. N249.  I will dig deeper into that too.    

I tried to finally swith over to the ME7 logger tonight. I couldnt get it to work.  It said there was a "SWnumber mismatch need '1037354837', ecu is '1037360857'..."

I am kinda excited to get this going though.  It has a lot more things to log in it than Ecux, and I think it would really help me out.

I would love helping you tune this beast. Do you think you can send me an email?
I think I can help you with the N75 ;)

Pm me your email.

Maybe this has been said before, but if you have a WG also you could hook it up to a MBC and cap the pressure at a few psi higher than you want to tune the vane regulation to hold, just as a safety while tuning the vane/n75.

Really cool idea to regulate boost with the vanes btw.

Thank you sir!  I do have a mbc in there hooked to the actuator to keep it from opening a too early of a boost level.  Switching this all over will be as easy as swapping the hoses. 

I have gotten fed up and am about to order the Pro-M MAF setup.

http://promracing.com/mass-air-meters/3-pro-tube.html

I will definately pay for the 30 point transfer function, and run this as a blow through just before the throttle body.  

I am also upgrading the fuel system a little bit.  I have some new rails coming, I will run larger lines and another pump in line.  


So, with the 30 point transfer input into my MLHFM, I shouldnt need to mess with the KFKHFM map anymore and just flatline it all out at 1.0?  Then use KFLF to dial in the actual fuel?  

EDIT:  after more digging I am torn between the Pro m and the HPX. 

Still working with the 85mm MAF setup:

Basic question.  I hear you guys talking about how your MAF is fully corrected, or dialed in 100%.  How do you know when your MAF is dialed in properly?   

When I move the scaling around for my MAF there seems to be a lot of other things that change.  On one scaling I can get pretty close logged load from 3-5th gears.  With this I can set up my LDRXN  pretty close without having to worry about it varrying from gear to gear.  Fueling is a lot easire, the PID is easier, but I max out my MAF.

Another scaling I have to log 5th gear from about 2500 rpms and floor it, and use that to set up my LDRXN since I now get much higher load in that gear.  I have to compensate my fueling all over the place...and generally is a PITA to set up.

For the moment I am trying a new scaling.  I am still trying a few more thigs to get it right, I just upped the boost some more and am maxing my new scaling out again.  So, at what point do you know your MAF is setup properly, and other parts of the tune need to be focused on? 

Use known injectors, scale the MAF, then use unknown injectors and scale them...

If you change both things at the same time, you basically don't know.

That's kind of what it seems like.  Is there a way to know when you at least get it close?  I can see that when I change the scaling it just makes it skew the readings lower overall.  It feels like I am just doing something wrong, like I am overlooking something.  You guys make it seem so easy  "just pick a scaling and..."   

I wanted to try another MAF, but I put it on hold.  Reason, if I am having a hard time getting my current setup dialed in, I couldnt imagine the hassle I would be going through by introducing a new variable.  I mean the Pro-m only had 30points, we need 512 IIRC.  I am sure there is software to fill in the missing points, but then will I be in the same boat now? 

I just read somewhere on this forum where someone said a proper tune will have the KFKHFM table all set to 1's.  On many other platform forums I have read that they actaully log MAF votage against corrections and actual dial in the MAF voltage to the get the trims proper.  That is kind of what I want to achieve.

It just makes me uneasy not really knowing if it is right, or if I am just making it all work somehow.   

That is a load of crap :)

No intake setup is that perfect.

The problem is, you have no real way of knowing what is the cause of your fueling being off. If you dial in the MAF (in KFKHFM) assuming fueling is perfectly linear, and KRKTE (and TVUB) is magically exactly right, and you manage to magically get req to always match actual AFR, it means your MAF model is probably close... but it STILL doesn't mean your MAF readings are accurate and your scaling is right.

Oh, well thats nice to know.  That kind of information stresses me out since I am trying to learn how to do things the "right" way. 

So, at this point all I can do is dial everything in so that it all just works?  I have had that several times now.  I had the actual fuel follow requested prefect.  I hacked the crap out of KHKHFM, KRKTE, TVUB, KFLF.  I also borrowed a few maps from Berttos' stg 3 tune.  I had that retarded notch from 3800-4k and it affected actual fuel, MAF, and LOAD.  Without the mods to I think it was three maps, no matter what I did the car would dump fuel at 4k and then overshoot towards almost too lean, and settle back out around 5.3-5.5k.  Now, notch is gone in all maps as far as I have seen.

Because it is...

I just think this thing is a little over your head. You have some very different hardware going on compared to stock, but little experience mapping.
You are slowly learning (as we all are/did), but I am not sure that you will get this dialed in anytime soon.

You also seem frustrated - if you are frustrated maybe it's a good idea to let someone who has more experience help you.
Unfortunately people who have more experience are usually busy, and want money for their services :p

As for dialing it in fully, you have to realize why you are doing things. Your setup will easily hit the ps_w limit. Are you logging ps_w?
You don't need any more fancy hardware, you just need to tune.

I completely agree with you.  As sad as that makes me.........
I did have it running really well for a while, but I am not satisfied with it just running well.  I want to know why it all just works, and if I did it the right way.  Since I dont fully understand what is right or what is wrong, I wouldnt know the difference if i did it or paid somebody to do it for me.   

The biggest help was when you and Nyet kept telling me that my fueling was hardware related and keep digging around.  Since then I added the fuel pump and yeah, you guys were absolutely right.  I didnt have enough pump and I was adding fuel to compensate for it.  As soon as I added the pump and tried a WOT pull the AFR was so rich it maxed out the gage.  Newb lesson learned. 

Since then, it has been pretty easy actually.  I have an idea of what maps do what and I log this thing so much it isnt even funny...."dont guess log".    :)    Right now, the car runs just fine.  Other than the STFT being retarded, but I can attribute that to the LC-1....the one sensor feeding the two narrowband inputs.  I still havent been able to succesfully get the banks to trim off of one sensor, and will more than likely solve the problem soon by adding the two stock sensors back in.  Max @ 034 says he hasnt had a problem with sensors lasting pre turbo, and that was my only concern when I started the build. 
 

they do last pre turbo, just not very long if you boost a lot.  Heatsink spacers help.  I added them back to mine just because I had it all apart.

I am following suit here shortly.  I can get WOT pretty much nailed but this partial and idle is not going to work with one sensor until the tuning gets figured out.  At this point I have worn myself ragged trying to figure it out.  No matter what, I still end up down the same road, the ecu adding a ton of fuel to one and starving the other....but it is content since the AFR at the one sensor in the one downpipe is reading stoich.  No matter how dialed in I get the trims, given enough drive time the banks will start to fight each other.

What are you doing for heatsinks?  I thought it was the pressure not so much the heat that would possibly kill the sensors.

Okay, so why not fit a single narrowband sensor and run both banks off of it?
Can you post your binary? I'll do the necessary modification.

Here is the BIN I am currently running.  I am still messing with a few maps, but if you can figure out the single sensor thing, that would be great.

Right now I am I have the single WBO2 connected to the LC-1, and the LC-1 sending two narrowband sensors to the ECU.  If it were to just read from one, which sensor would input would it be reading from? 

Are you still planning on simulating the narrowband input from WB?

It doesn't matter which bank, whichever one you want to plug into.

It's a simple solution really, where the ECU first gets input from the sensor, you just replace the variable for one bank with the other. That way it is reading both banks from one sensor.

To go one step further, identify the ESKONF bit pair for the unused bank to diasable power stage diagnosis and the sensor can be removed completely.

No, I dont want the lc-1 in there anymore.  Its too much of a hassle, and I have a zeitronix in there now. 

If it cant be done in tunerpro, I cant do it yet.  I am not that advanced.  That is pretty much why I figured it would be easier (for me) just to replace sensors a bit more instead of beating myself up over simple driveability right now. 

prj offered to do it and if I get the time before he gets to it I could do it for you.

Which bank do you prefer to operate?

I haven't looked at your binary yet, what platform/software are you on?

It doesnt matter which bank.   I am running M-box 2.7 software.

I changed the file you provided to operate both banks primary O2 sensor readings from the sensor connected to bank one.

I had to change three references to uusvk2_w.  

==============================================================================
me7check v1.11 (c) mki, 06/2004-07/2011
Checking file marcellus_single_b1.bin (size=1048576)
Reading Version Strings...
-> Bootrom Version = embedded in CPU, asume 05.12/05.32
-> EPK = 40/1/ME7.1/5/6005.01//22m/DstC2o/011200//
-> Contents of ECUID data table:
   - '0261207143'         (SSECUHN)
   - '1037360857'         (SSECUSN)
   - '8D0907551M  '       (VAG part number)
   - '0002'               (VAG sw number)
   - '2.7l V6/5VT     '   (engine id)
-> Contents of ECUID data table:
   - 'HW_MAN004'

-> No errors found. File is OK.   *********************

Thank you so much!  I will give it a try first thing in the morning on my way to work. 

new approach and i hope it works, but i thought we covered this and its due to stereo lambda trying to have the two banks slightly diff?  like the same as two sensors in one down pipe, it expects to see different values and when it doesnt it gets pissed off.

I do vaguely remember this being mentioned, but I haven't seen anywhere in the FR or code that this would cause an issue. Granted, I haven't looked very hard either. Logically, I don't see any reason for a that to cause an issue either.

Front and rear sensors are evaluated against each other for catalyst checking and response time, but not B1 vs B2 AFAIK.

Thank you!  Major + for you!

From what I have seen so far the tweaks worked.  I was only able to watch VCDS for a while when driving.  The lambda values did hover around from each other slightly, but overall stayed pretty close to each other.  That was just a short drive to work.

I still had the LC-1 going to bank two.  I wasnt sure if I was supposed to pull it, or if it mattered.  After I got to work I pulled the plug for sensor 2 and let the car idle for 10 minutes.  Usually during a long idle like this the trims will go crazy and car would misfiring bad.  This time the trims stayed close to each other and idle stayed rock solid.  That is a very good sign.  Hell, after about two minutes.....idling at a stop light, I can feel the trims going their seperate ways.  

I am excited to see what happens on a lengthy drive.  If its anything like I am hoping, I finally will be able to plug back in the evap solenoid and dial in my LTFT.  

Sounds good. Keep us updated.

Bank two should be completely ignored now for all purposes except maybe power stage diagnosis (O2 sensor heater circuit).

Nice, I will give it a shot soon and report back.

Philla... how the heck did you even find those variables, i cant even find reference to them in the FR, cant find them in any definition for any box ever, good work!

can you give locations for all three so I can test it on mine?

Define these as 16 bit (do not designate LoHi or LSB), output type will be Hex digits.

To delete B1:
0x3C8BE
0x3C8D0
0x3E832

Original value will be 1A91, to replace with B2 input change to 1891.

To delete B2:
0x3C8F0
0x3C902
0x3EA90

Original value will be 1891, to replace with B1 input change to 1A91.

I can explain more later if you would like, but no time now.

i think i have my head wrapped around it

Thanks. Added to wiki. You guys rock.

After a bit more idling and driving it seems as if it is a lot closer than before, but the trims are still doing their own thing.  Definately not as bad as before.  I watched the voltages on the two banks via VCDS and with only one sensor input being used, both sensors are showing the exact same reading all the time.  Something is still making them trim different.

Maybe I have the wrong idea of what should happen.  In my head I pictured them being pretty much batch fire and not so much sequential.  Thats the best way I can put it.  The reading from the one sensor would translate into one trim across the board.  

I do realize I need to adjust my fueling now that things are a lot more consistent.  Maybe once I get that aspect better, the trims will settle in, and worrying about them being exactly the same will be a non issue.

Any other areas I should look?  

This may not apply because I think I remember you saying you disabled LTFT's, but have you reset trims?

I disabled LTFT by unplugging evap.  You think I should plug it back in and let the LTFT do its thing?  

I reset the ecu a few times just to be sure.  

Yes.

What is the situation with your post cat O2's?

Are you using ME7L?

The post cat o2's I believe are deleted.  I dont them plugged in and have switched the codes for them to be off.  I also changed CWONKLS to 17, which I understand to be just the front o2s turned on. 

I am not using ME7L yet.  Every time I get around to setting it up, there is some other issue that makes it not work on my car.  I havent figured them all out yet.

i bet they will still swing, r32 tuners had these issues for years.  it wants individual bank control and readings.  when it reads the same exact voltage it tries to make sure it has control and it trims the banks away from each other, keeps doing this till it gets results, with the same reading still it never gets those results and eventually swings to +25% on one and -25% on another.

system constant for stereo lambda is our issue. 

i tried to fix it by offsetting my simulated narrow band data by .001 volts from bank to bank, it helped but didnt cure it.

phila_dot - cool, thanks.

I would have done it tonight, had to travel due to work so was not able to earlier.
Good you sorted it :)

What if you did the offset by more than .001 v? If it really is that there is some threshold that the difference between the 2 banks needs to be higher than, then you should be able to breach it some time. They may have a check for 'too close to identical', but I can't believe they would have a check for 'make sure the values aren't just offset from each other'. Wouldn't this stuff be visible somewhere in the disassembled code (I don't even know how to disassemble, but if there is a stereo check routine it has to be visible somewhere down the logic stream past the O2 voltage.)

If it is that bad then don't turn long term trims back on.

We should be able to track it down and figure out a solution. I'll get some variables if y'all want to do some logging.
There are offsets in the code that we can use as a last resort.

Please back up your claims with some data from the FR :)

There is a subfunction, LR-SYNC, in LR that may be related. I also see a codeword, CWLRSYNC, that apparently allows one to sychronize fuel trims and set which bank is master.

I will sort through the code when I get the chance.

I turned off the trims because with them working I became slave to whenever I drove the car, I HAD to have the laptop and VCDS running.  Maybe a quick down the street and back I would be fine, but any thing over 10-15 minutes and the LTFT would have adapted to the point where the car seemed like it would barely run.  I was hoping that it was a fueling issue and once I got the fuel closer the trims would settle more, and then I would be able to dial in the trims better.

I drive a good distance to and from work. Half the trip to work the trims would sit at some real nice and low numbers like +/- 3%, it would be close to the same for both banks.  At some point into the drive the trims would start distancing themselves from each other.  I was usually able to feel it coming because the car would feel like it was stumbling/misfiring during cruise.  Since it was the LTFT the only way I was able to get them to clear out was by clearing codes in VCDS while driving.  It wasnt uncommon to see -25 on one bank and +25 on the other by the end of the trip until I figured out what to do to clear them out while driving.  I usually caught them when they were around 12% split since right aorund there it was really noticeable.

Not a claim, just what it does.  Every dual bank me7 system that has tried to go to one downpipe path has had these issues.  The FR doesn't talk about it much at all.

Emulated narrowband signal from Innovate does not work with ME7. Even on non-stereo exhaust systems.
And the FR talks in *great detail* about almost everything, since this is the *specification* for the system.

and if the system is setup with a system constant they go into basically zero detail about setting things up, because you cant with system constants, they just are what they are.  stereo lambda is a system constant.

posts on this very issue go back over a year, trust me I would love to find a true solution but up until now I do not know of anyone, even supposed professionals who have an answer other than to run two sensors pre turbo, bank specific.

I have spent more time going through the various documents available to me on just this issue than most have probably spent reading the fr in general.  the answer is not in there.  if it is in some better version that you or others have then they should probably share that version before arguing about not backing up info that comes from real world experience, over and over and over again, not reading a document once again for info that isnt there.  i have tried two narrows in one downpipe (didnt work, ecu still tried to trim), i have tried both reading from one sensor (which looks like this new approach is repeating the effects of, still tried to trim), I have tried the emulation route to have two unique signal inputs (sorta of works but still eventually trims very slowly)  I have tried two sensors pre turbo/divided banks (only thing that worked) and now I am going to go back to pre turbo sensors and only run them when dialing in new hardware as they get killed after lots of time in boost.

if someone can figure out how to make the banks not try to trim then awesome, i have always said i am foggy with me7 due to lack of access to certain info or logic.  i treat every new approach as possible and with encouragement.  i have always theorized that the individual bank control was bosch just being smart, fuel pressure can greatly vary bank to bank, ve could vary due to exhaust issues, 3 injectors to trim for compared to 6 is more precise, i could go on and on.  it makes sense why they did it on various engines, just makes changing things hard.

emulated data does work with me7.  bosch has a voltage line offset of usually about .007 volt, you have to compensate with any emulator to make it think its a proper sensor and not an emulated signal, also the other factory settings with innovate are wrong.  i trial and error ed over thousands of miles as well as measured sensors to get it right because no one else had or wanted to share the info, not even bosch on what the offsets are for their narrow bands.

Look, if there is an algorithm that compares the banks it will be described in the FR.
Whether you have found it or not is a different story.
The FR /is/ the spec for the implemented logic, and it must conform to said spec. Simple as that.

I don't have time right now to look through it, but all of this can be modified in code, instead of trying to fool the ecu into doing this or that.
There is SY_STERSY and SY_STERVK.
Searching the FR for SY_STERVK should yield the answer, or a clue.

As for what FR I am using - the german one of course. If you use the alfa FR, loads of stuff is missing or translated wrong, even from factory.
The lambda stuff is probably better used from the alfa FR though, as it should be narrowband and it is not translated.

P.S.
I think the problem is just due to how the algorithm works, so the trims start creeping away.
The right way to solve this, is to merge the fuel trims in ASM code, so that all fuel trims are used from one bank. I don't think it tries to move the banks away from each other, it just happens, because it has two trims and one sensor.

dude, im not trying to be an ass but i have read all of that 100 times over.  others have too.  no one has found anything that has lead to a solution.

Edited the post, and described the solution.

This really isn't all that hard. We just have to isolate where it is happening and adjust.

I can provide whatever variables are needed. I will take a quick look through the code for anything that might apply.

The trims originate in fr(2)_w correct?

I agree that it's not that hard, just most people can't do anything with assembly :P

To spell it out - I think the problem is because of a Snowball (pun intended :p) effect.
Basically since you have two separate trims, and amplitude adjustment, the trims will eventually start creeping opposite ways.

The right thing is instead of simply copying the lambda voltage, to actually adjust the logic and take the mixture adjustment of bank 1 and write it into the mixture adjustment of bank 2.
This can be either done by copying trims over, or the trims can be left as is, and all mixture adjustments that actually use those trims can be re-written to use bank 1.
The correct way is probably to see what is the easiest to do and requires the least re-writing.

Copying fr_w to fr2_w would be easy, but I think we need to isolate exactly where this is happening first. Then decide the best course of action. I'd prefer to neuter it in the beginning vice at the end.

This is correct.

I think it is enough to open the Alfa FR, locate everywhere SY_STERSY is used and make modifications in code, as if this was set to 1.

I wouldn't rely on the FR that much. Especially when I can look at what's really there on my computer.

Better just to xref all of the B2 variables (that apply) starting from GGLSV.

Every function I have seen processes B1, then the code is duplicated directly underneath for B2.

I guess we could just jump out after B1, as if STERVK was false, but how do we ensure all of the bit states are proper? Then again, it wouldn't matter if we thoroughly go through the entire path.

This might be quite a bit of work though.

Jeez guys, I didnt think it would be this much work.  I really appreciate your help though.  I wish I had something to add other than sitting at the edge of me seat hoping it gets resolved.

Well, I have been trying to get ME7L to work.  I cant get past the software version error.  I am still researching it.  Hopefully I will have it resolved toda and will be able to get better logs for you guys.

All I know is that I wish I was as advanced with assembly as you guys.  I'm a tuner and a welder, not really a software guy.  I got money via paypal going to whoever figures this out.  Least I can do for your guys time

I will add to the pot

I will be wrapping up my map switching routine in the next few days and then I will knock this out for you guys.

Marcellus,
I don't have alot of time these days, but I went through GGLSV and tracked all of the B2 variables. I haven't gotten to go through every reference yet, but I threw something together that should work at least for now.

The attached file is the same one that I posted for you previously except for a few small things.

B_svuerf2 is permanently cleared. I think that this would be set in conjuction with B_svuerf and cause B_ripuv to be otherwise always false. You could check this by logging B_svuerf @ 0xFD38 0x0008 and B_svuerf2 @ 0xFD38 0x0010. If they aren't causing a problem then you can just change 0x83EB64 from 4Eh back to 4Fh.

fr2_w is being created with fr_w. This should force all fuel trims to match.

I also set your post cat ESKONF bits.

ME7CHECK v1.11 (c) mki, 06/2004-07/2011
Checking file marcellus_single_b1.bin (size=1048576)
Reading Version Strings...
-> Bootrom Version = embedded in CPU, asume 05.12/05.32
-> EPK = 40/1/ME7.1/5/6005.01//22m/DstC2o/011200//
-> Contents of ECUID data table:
   - '0261207143'         (SSECUHN)
   - '1037360857'         (SSECUSN)
   - '8D0907551M  '       (VAG part number)
   - '0002'               (VAG sw number)
   - '2.7l V6/5VT     '   (engine id)
-> Contents of ECUID data table:
   - 'HW_MAN004'

-> No errors found. File is OK.   *********************

I really appreciate your work.  I will try this file out in about a week.  I am taking the car apart to replace the motor right now due to some sort of knocking.  New motor should be here mis next week, and I will install it that weekend. 

Thank you so much! 

I will use this bin as my new starting point for my single tune.
Will hopefully have it under power this weekend. 

I would definately be careful with that file.  I am not one of the better tuners on here, and there are some things I have done to it that arent necessarily the right way of doing things. 

Hey man. Yeah I saw your reply on AZ, just hadn't had a chance to reply yet...
I'm going to actually go back to stock with all the tables as a starting point so it's all good :-)

I got the motor in over the weekend and tried this file out.  So far, it is working great.  I havent plugged back in the evap yet to let the car do its fuel adaptations.  From just watching VCDS the trims from bank one and bank two were EXACTLY the same at ALL times.

Excellent!
I'm getting ready to tune with that file this weekend... Good to know it seems to be working out!

Here's my car yesterday after the new fuel system install. Ready to tune!

http://www.youtube.com/watch?v=-Cl46Bw8kE4

Share results please!

Oh I will for sure!
Probably the weekend I'll get a few things on the car finished and start tuning.
I have an LC-1 Wideband setup with Narrowband out going to B1 O2 sensor.

Good to hear that it's working out so far.

Does the trimming appear to be behaving normally?

I think so.  I should have some logs tomorrow morning.

I flashed my ecu just now with the "mono" lambda sensor changes.
I can confirm that both banks show exactly the same O2 voltage, with only the Bank1 hooked up to the wideband.
Good work!

$260 shipped, I just ordered the PMAS, wiring pigtail and aluminum weld on flange.  I am finally going to convert to blow through and hopefully get rid of the no-ideal intake setup.

Blow Through ftw!
(http://i1357.photobucket.com/albums/q755/britishturbo/IMG_20121028_170926_zps355d17da.jpg)

I am kind of nervous about the transfer function I came up with.  Hopefully it will be a good place to start. 

(http://i360.photobucket.com/albums/oo43/SpeedWorksAutoTuning/IMAG0682.jpg)

I can check over it if you like :-)

File I was messing with:  see MLHFM

Using the HPX maf tool, calculated for 3" pipe......30 points transfer

Interpolated the 30 points to 512 for me7.  The Low volt region .01 and a few points above have me kinda concerned.  When I interpolated them I came up with a bunch of negative numbers.  TunerPro made them 0, which is fine, its just something I think I m going to have to keep an eye on.  I really need to start using ME7L so I can log MAF voltages, or better yet use one of the inputs from the zeitronix.

MLOFS exists so that MLFHM can be all positive.

More specifically for return flow detection.

^this

I'm using the HPX sensor too and the MAF curve I ended up with is a good bit different than the interpolated curve I started with. However I'm using a mostly stock air box (just with holes cut in it) and with KFKHFM all set to 1, so I kinda worked the airbox's flow characteristics into my MAF curve. There's still a few rpm/load spots I have to fix up with KFKHFM but overall I think doing it the way I did got me a much more accurate reading with my set-up.
I would definitely suggest trying to get ME7L working, being able to log MAF voltage, g/s, fuel trims + Zeitronix AFR made getting everything dialed in pretty easy.

Would you mind sharing your Mlhfm?

Here's what I'm using now, It's for the HPX sensor in a stock hitachi 73mm housing (except modified for the slot style flange) and stock air box with the front lower triangle cut out. It's scaled as close to real values as possible, using a KFKHFM set to all 1's for now.

Be careful with it, do some part-throttle pulls and watch fuel trims/WB AFR before doing a WOT run. I have only tested this curve up to about 3.25V (320 g/s), and I kept it almost linear after that point due to the flow characteristics/restrictions of the stock air box (based on the shape of the stock KFKHFM) with the intention of fine tuning the rest of MLHFM when I get there.

Yours is quite a but different.  According to yours, I definitely would have been running lean up top.  I am going to take the MLHFM posted, and multiply it by the difference in diameter and use that in one of my first start files.  I am also going to break down and order the LCD display for my Zeitronix so I can see the afr at all times instead of only when I am logging.

You should always always have a display for your wideband.
Its one of the most critical tuning aids. You won't regret it.

I know, I know, I know  :'(

Here is my calculated MLHFM for the HPS in a 3" tube.   The first column is the Erroob's 73mm, and the next is mine calculated off of his.  I multiplied his values by 4.383561643835621, the difference in diameter.

Technically, you want the difference in cross sectional area...

This is one thing that I've never got resolution on, I can't find it now, but I remember a thread where people were talking about the MAF scaling of Tony's stage 3 tune and MAF scaling in general, and it turned out that scaling by a ratio of the maf diameters, rather than by the maf cross-sectional area like one would think, yielded the right results.  If I recall correctly, you were a part of that conversation.  Your current opinion is that MAF scaling should be done using a ratio of the cross-sectional area, and not a ratio of the diameters?

In theory, yea, it should be cross sectional area. In practice, it seemed like diameter had better results, but I have no idea why.

I do remember the thread where you said you got better results using diameter.  I also tried diameter versus area and I got better results.  That could be due to the fact that the scaling came out lower so I wasn't pegging the MAF as bad. 

I shouldn't have to worry about that with the new MAF, so I will try it both ways to see which one is a better start.

Not just the MAF :)

ps_w and rl_w as well.... ps_w being the main culprit for me, now that I have had a chance to play with prj's KFLF hack.

Any of you guys planning on trying the real scaling adjustment of ps_w/rl_w?

I kind of snickered when I saw him post that in the other thread.  I stumbled across that in my hundreds of logs and had been doing it that way with good results.  What I noticed was that no matter what I did the car would lean out around 5k no matter how much fuel I added.  I looked at the MAF readings and noticed the more I added to KFKHFM the worse it got...thinking I was making it better, instead was making the readings higher and higher.  I ended up looking at my first few revisions and noticed the fueling wasn't as bad when KFKHFM and had a thought....maybe reducing KFKHFM is another way of underscaling the MAF.  The more I lowered the map the more in line the fueling got.  I also remembered a thread where somebody said lower the values in MLHFM and add in KFLF to equalize fueling when underscaling the MAF.  I used that method with KFKHFM and things lined up really well. 

With this new MAF I was hoping to not have to lie to the ecm anymore about how much air is coming in.  I am not 100% sure but I think with my hitachi in the 85mm maf, if the readings got over 380g/s the fueling went sour. 

Here's a little video of me dialing in my single with blow through :-)

https://www.youtube.com/watch?v=R-R3QwIsZEg (https://www.youtube.com/watch?v=R-R3QwIsZEg)

I am going to run the stock M-box scaling that maxes at 22psi or whatever it is.  I run past 22psi all the time and fuel just fine.  I don't think my fueling problems were because the ECU cant see how much boost, but properly meter the air coming in.  Maybe I am not as far advanced in my tune yet to have the MAP limit problems, so until I see it as a problem I don't see a need to change it.  Plus I don't really understand what is going on in that thread just yet,

YOU TRICKED ME!  I didn't see one lick of WOT in that pull!   ;D

Haha that was just part throttle on the way back.
I was having too much fun wot to grab the phone Haha.
I need to do another re cal on the Maf though its still a little too rich at high flow... so I'll grab some video on the next run...

Since the boost is based on load (and load is based on how much air comes in), the tweak doesn't just affect the boost, it affects the load cap/MAF limit stuff that you are talking about in your post #251 above.

Well then it is surely something I need to address.  I don't know why I was thinking that you guys were doing all that work just to be able to read past 22psi.  It makes more sense now. 

What I am having problems with is rescaling certain axis.  I would like to rescale the MAF and KFLF load axis so I can tune where the load actually is and not just leave it be past the 159 or whatever load they are at.  I suspect since I wont be underscaling the MAF anymore it would be nice to actually tune all the range I will have.

I have tried it in tunerpro, and it doesn't change anything.  It leads me to believe its like all the other axis in the BIN/XDF and are defined somehow....kind of like the timing axis for load and rpm are.

You will have to underscale (at least up top) or re-scale all the load in the ECU.

So I'm still very much dialing things in on my single turbo setup.
I will post everything I have done so far soon but for now here's a virtual dyno plot from my drive to work today.
Still very much rich in the middle and getting timing dialed in still.
This was at ~22psi.
This is runnign the stock Hitachi sensor in a 3" pipe in blow through configuration. I have the MAF dialed in to the point where it works great, but maxes out around 5000 rpm lol ;-)
I'kmm be upgrading to this guy soon http://vmptuning.com/index.php?p=product&id=1239

(https://sphotos-b.xx.fbcdn.net/hphotos-ash4/487303_10151236211737895_1600170532_n.jpg)

Car is running a consistent 3.0 second FATS time in 3rd gear.
Here's a plot from yesterday showing Boost and Injector duty cycle.

(http://i1357.photobucket.com/albums/q755/britishturbo/drivehome-1152013-3rd_fats_pul-injector_dutyl_zps69cc4da6.png)

What turbo is that?

It's an AGP Borg Warner S256:
http://www.agpturbo.com/borg-warner-agp-s256-rs-turbocharger/

The install is all custom made by myself.
I'm still learning how to tune the ME7 :-)
Here's some install pics:

(http://i1357.photobucket.com/albums/q755/britishturbo/IMG_20121223_133200_zps1e83628e.jpg)

(http://i1357.photobucket.com/albums/q755/britishturbo/IMG_20121110_163145_zps1812e97e.jpg)

Fuel setup:

(http://i1357.photobucket.com/albums/q755/britishturbo/IMG_20121118_122303_zps65e97a60.jpg)

I have to ask, does a b5 really weigh 3700lbs?

My b6 avant with me in(90kg) and half tank weighs 1680kg flat, thats 3700lbs on the spot...

I had entered 3500 for the car plus 200 for me...
Factory curb weight is listed as 3,384 and it had about 13 gallons of fuel in it (I had just put 10 gallons in it 30 miles were shown for range) , that's about 80 lbs, making it around 3465.
I figure all the extra fuel lines and everything probably make it darn close to 3500 :-P

So yeah, I'd guess my car was about 3700 ;-)

Here's a crappy video from a 3rd gear run yesterday.

http://www.youtube.com/watch?v=v_fcFOsofGo

I see, my avant is listed 3556lbs.. +200lb driver and fuel :)

I was under the impression that the b5 chassis was about 200lbs lighter vs. B6's, I guess the V6 and bling interior makes up for that;D

Great numbers, what is the gearing in 3rd and final on that 01E?

I think the first generation of the B5 S4 was about 200lbs lighter, but the 2001 and 2002 year models (mine is a 2001.5) had heavier steel suspension uprights, whereas the earlier ones had aluminum uprights. So I guess you could say mine is the heavy brother haha.

I have 3rd listed as 1.231 and the final drive as 4.11 in my Virtual Dyno setup.

And thanks! I'm still very much learning about the ME7, I probably make 3 or 4 re tunes each day at this point lol

It's been a month of work, and now I am back at it.  What has been done can be seen in the projects section of this forum.  There have been a lot of upgrades and I am back to "dialing in" my single. 

Major change:
HPX MAF in blow thru

My first flash with the MAF working was a bit off and I was running too rich.  I just dropped the KRKTE to compensate and that revealed a new issue.  Almost as soon as I started building boost, I went hard limp...."torque 2 exceeded..." or whatever that was.  Messed with the KFMIOP and KFMIRL  and changed the Load axis a bit and all was good.

Now I am overboosting like crazy, and need to pay attention to the PID.  It was too slick out to get any good logs, anything aggressive resulted in major wheelspin, and the overboosting wasn't helping any. 

Awesome, my HPX Blow Through MAF arrives tomorrow so I'll be retuning then :-)

It's definately not as bad as i thought it was going to be.  It still runs pig rich, but it statrs and drives fine.  Idle is excellent, and glancing at the WBO2 readings, even with the overboost situation, I have plenty of fuel. 

I did do a log this morning, but for some reason my laptop crashed before I was able to save it.  From the looks of the fuel curve, I can probably scale the MAF back a bit.  Right now the AFR's are rich across the board, and scaling tyhe MAF back will help a bit.  I dont really want to go to much further with the Fuel maps since they weren't this rich until the recent changes. 

continuing convo from  Audizine:

 Originally Posted by SpeedWorksAuto 
"Have you had any "torque exceeded......" issues with that aggressive of a file? " - me

"You shouldn't encounter this problem unless you screw up the torque model " - phila_dot

I have had this torque exceeded issue a few times in tuning the car.  I always thought I was being a bit to aggressive since it would happen when I had been playing with the timing in the mid to lower range, or when I had the car on MBC.

I never have been able to really get over it, I just kind of worked around it.  ...dial back timing and boost.   

If you can post your latest bin file I'll take a look at it.
I never had any torque exceeded problems just timing intervention.

A log would be good as well to see exactly what it's running Load wise.

Its probably a mess right now since I am doing too many things at once.  Right now I need to get the overboost under control so I can get full logs.  I also just messed around with the IOP/IRL maps on the fly last night.

I am about to run to the store to get some new tools for my shop.  I will try and get a log up to the point of overboost if I can.  Shouldnt be too long.

Problem^^^

Log these:
http://nefariousmotorsports.com/forum/index.php?topic=3039.msg30034#msg30034

Problem.

MAF not capping out. 

my guess is your LDRXN is set up wrong. request more load earlier.

BOOST:

I just dropped the KFLDRL in the 3k range and apparently need to bring it back up, and adjust the 4k+ range some.

I am in the process of adjusting it now.

Kinda seems like I am starting from scratch again.   Tune for ~20psi and get fueling proper, and then add boost.  It is going pig rich and then suddenly leaning out.  I think it is related to the overboosting though.  This chart is a lot more tame than the two logs I took previous the two revisions I just did.

I probably shouldnt have said on the fly, as in just willy nilly moving stuff around.  I should have said that I used some maps from a previous revision I was running that fine.

It's not LDRXN, rlmx_w and rlmax_w look fine.

What does rlsol_w look like?

Definitely doesn't look like the problem is on the desired charge side to me.

Here's the log

Boost is getting closer, I still need to tweak the VGT opening PSI and the N75 some more to dial back the boost spike.

Load and AFR chart shows fueling to still be retarded, but not totally leaning out now that I am not boosting to who knows where.

AND.............now that I have some other people to reference I am thinking my timing is really low.

You are requesting too much timing.

Check your timing retard.

I thought I was running pretty conservative.  I know part of the reason is due to the lean spot. 

yea sorry, didn't see that. DEFINITELY make actual AFR match requested fuel before going further. That should be the FIRST thing you fix, not the last :) Don't BE LAZY HERE!

It should be consistent regardless of boost levels.

I am trying to not mess with the kfkhfm map at all.  The map is set to all 1's right now since I am running the hpx in blow through.  I think I might use one of my user inputs to my ZT-2 to log MAF voltage so I can see where the fueling is going off compared to the voltages at MAF and dial in the MAF better.

Or you could add uhfm_w to the log.

I went to my first MLHFM that I interpolated from the original HPX 30 point transfer function and original KRKTE and it got me a lot closer to where I needed to be.

Going through my last log of the night.  Not too bad for a couple of revisions.  Timing is still really conservative.

I am wondering how my car compares to the s4's given I am running a 2.8 5speed trans and 19" wheel/tire combo.

Well I run 3.4-3.5 fats with K04's however timing and boost is not fully dialed in yet (the car is 1800 kg including me). Of course britishturbo has been running way better fats than this with his single turbo and I guess he is still not ready with his tune.

That lean spot is likely because you are hitting the ps_w cap.

I Kind of figured this was the case considering it was really lean right where the boost goes off the scale.  Until I figure out the hack, I did what I have been doing to fix 22+psi weirdness.  I scaled KFKHFM down in those areas and scaled KFLF up. 

2.8 FATS now at ~22psi of boost :-)
I'm getting my tune a lot closer to where I want it now...
And today my new MAF showed up... so I get to start over with my MAF too lol

(http://i1357.photobucket.com/albums/q755/britishturbo/IMG_20130122_102403_866_zpsb9304026.jpg)

Would you mind sharing your MLHFM to save me the time of making it also? Cheers!

The file I originally made/MLHFM I am using now is on page 16.  It should be easier to snatch the MLHFM since all you would have to do is a copy from compare vs.  copying it from a spreadsheet.

You can see from the last log from the night, it does run rich under high load.  Let me know what you come up with when you are dialing yours in.

lol missing file? haha
I know you posted a bin with it yesterday but I just wanted to make sure I started with a non scaled maf... but I can do it myself lol

Sorry, I changed the post.  I thought it would take the attachments along with the quote.

Now, I got a request for you Mr. Britt.  post up a screen shot of your IOP/IRL maps please sir.

Still didn't see the file... haha but here is what you asked for :-)

(http://i1357.photobucket.com/albums/q755/britishturbo/v22-irl-iop_zpscf8c0147.png)

My current tune.  The MLHFM is the same as the file I posted on page 16.  KFKHFM/KFLF is modded a bit to compensate for the ps_w cap (boost past 22psi?) 

Now, I am really confused.

Thanks!

How so? Share your thoughts lol.
I can upload the maps if you need them...

I am a beginner to ME7, and most of the stuff...debates/threads....whatever, contradict what you posted.  Unless I have read all that crap wrong, and didnt understand what the hell was meant, this map shouldnt work.  (question/statement)  

But, it does work, and you seem to have it working quite well.  So now I am so damn lost.

This is mine.

I'm new with the ME7 as well... still figuring things out as I go along :-)

TIming maps please, and what gas you run?

92 or 93 octane.
I've been playing with my timing side of things a LOT. Specifically where it comes to the knock side of things.
If I let the me7 pull as much timing as it wants to then it doesn't make anywhere near the power it does now... so I don't let it.
Yes, I like to play with fire ;-)

How are you stopping it from pulling all that timing?

Well the first step is getting your AFR right so that you have a consistent burn...
Once the AFR is right then you can start looking at where in the load vs rpm range the knock really starts... then massage that area of the timing tables to try and reduce the amount of timing that it wants to pull...
Then if you are sure you are not knocking at higher rpms you can control the max knock retard using KRMXN. but this is very very dangerous... I  have 22 revisions of my tune over the last 2 weeks dialing things in where it is.

I've also disabled knock adaption. But that is also risky :-)

I was just about to ask this.  Since there are maps dealing with how much timing is pulled at a time, and how fast it recovers.  Memory is failing me right now since I am not even ready for that part yet. 

Once step at a time.

Fuel AFR's first

Yeah, let me know what you come up with for the MAF.

Fun times, big ole turbo spooling pretty damn fast.  Still conservative with the timing.  Fueling was nutz, and then I added more boost.  Now its back to playing games with KFKHFM and KFLF.

That isn't conservative with timing... you are requesting WAY too much timing.

get your CFs lower...

My AFR was off in this run again.  I upped the boost a bit this log.  

CF's are what, around 5-6.  That, from what I have been told is ok.  Is there any real advantage to just lowering it anyways?

I have been finding that <5 is getting me the best timing..

that said, your mileage may vary.

you could maybe start with cf <2 or 3, then work your way up until you stop gaining timing.

Interesting to hear, now that you've tuned a few cars on the dyno with good results. Sounds like I'll give this approach a shot.

Edit: and you're running stock knock sensitivity/CF hysteresis?

**************This is extremely important!!!!************

Anyone running the hack to pull fuel trims off one sensor, you MUST be running evenly fed fuel rails.  This is especially important if you have a potent fuel pump setup.  With an aggressive pump setup and lots of flow/pressure the first bank will be rich and the second lean.  Very lean.

My recommendation is to have a proper fuel Y, flow into each end of an after market or modified stock rail and then out to a regulator.

If you are flowing into one bank, then crossing over to the other like a stock rail does, even if the cross over is an AN line of adequate size you will be lean on the bank that gets fuel second.  No matter what.

This is a new issue that is from the hack, it isnt because the hack but the hack stops the leaner bank from adapting.  Stops both banks from adapting accurately actually as the single the trims do get are an average of the two and one will always bee too rich and one too lean.

The only way to combat it is with the banks getting the exact same pressure and flow volume with a proper, evenly split feed.

I cannot emphasize enough how big of an issue this will be for some, especially at elevated boost levels and with already tight fueling.

You just gave me the sickest feeling down in my stomach.  Thanks for the heads up.

Are you the person that a friend of mine in Pittsburgh said melted a motor because of this?

I don't agree with this BTW.
I'll explain my reasons in depth later. Too busy digging through code right now and putting the kids to bed lol

It seems like, if the above was true, anybody not running a fuel pump feed to both rails would run one bank lean @ WOT.  I say that because I thought the o2 was ignored past a certain load and only MAF and a bunch of maps.  Since the hack was to help with STFT/LTFT, dealing with o2 voltages, do you think the car can go that lean enough to cause damage under cruise? 

All I am saying is that if this is an issue, I think its an issue anybody running fuel rails with one feeding the other would have. 

Im running both banks oem narrow pre turbo, single oem narrow post turbo and single wide band post turbo.  I have had the car back going again for a few weeks and I have been testing these new advancements from PD, the hack works perfectly, but this can be a side effect.  I played around with both fuel rail setups as I have done in the past.  The issue isolates to cross over tube type fuel paths.  Like OEM.

The ecu has no way to understand that there are two fuel rails with different amounts and pressures of fuel in them and it applies the same injector PW to all 6 cyl.  It is inevitable unless you split to have a variation as the first bank removes fuel and pressure yet the second bank fed.

Feel free to dispute this through your own testing, be careful.  Get your injectors perfect and keep egts in place, or just listen to me as Im pretty sure I am batting 1000 on advice i give about ST setups on 2.7t, i even tell you guys how to unfuck your setups via a customer on AZ sometimes... ::)

Could it be due to not enough fuel pump?  I just don't understand how the fuel pressure would be different in what is basically one continuous line being fed on one end and and restricted on the other.  

Are you logging NB voltages at WOT pre turbo?  Have you logged fuel pressure? How does a stock setup know that one bank is leaner than the other at WOT?  

I am not doubting your advice, so don't take it that way.  I really want to know how you figured all this out.

I'm confused too :(

If you are running open loop, i don't see how whether or not the system is stereo (or hacked to be mono) matters..

what am i missing?

again, not questioning, just would like to learn

The situation is present at all states, idle with and without trims, part with and without and most importantly WOT.

As usual it goes back to the system really wanting two inputs.

Marcellus, the situation gets worse the more pump you have because then the first bank to receive fuel is getting bombarded with it.  It ends up with neither bank being spot on.

Well, running a "y" in and out of my lines was on my to do list.  I just never figured I could be starving one rail.  I just can't help but wonder what the straight 6 guys are doing.  Seems they would be up against the same problems.  Feeding from one end and, and progressively starving the following injectors.  They don't run bank specific fueling logically, so how do you deal with that?  It's the same thing I see going on here.  Maybe I just don't realize how fast injectors can deplete the rails, but I thought my rails being -8 would be enough. 

Still pretty confused, but I am going to mod my lines because I don't like feeling unsure about my fueling.  Especially the stuff I assume to be ok. 

I'm with the confused and looking to learn party.

So your post says that it is a new issue that crops up because of the 'single input' hack.  I thought, by definition, that in open loop (WOT) there is no input. If this is the case, than the issue would be there independent of whether it's a big power single or twin turbo, no? Because the inputs at WOT would be the same, 0 inputs.  Or does the stock ECU learn bank specific trims at closed loop operation that it extrapolates into WOT open loop, bank specific usage as well?

It does seem that there is bank specific, fuel rail configuration specific calculation/compensation that the stock setup is utilizing, as this guy replaces the crossover with a bigger line and has bank specific issues because of it:  http://www.audizine.com/forum/showthread.php/513394-PSA-Don-t-blow-your-motor-because-of-my-stupidity?highlight=fuel+rail

Fuel trims are always active

Edit: They only learn under certain conditions

In the case of open loop, only the the long term trims are active..

so the issue with mono vs stereo is only if having a hacked stereo is causing your trims to look identical when they should look drastically different?

This makes no sense, since IIRC there is no load/rpm based long term trim (other than idle/part)

If you are running out of fuel in one bank only at high rpm/load, and your part trims would be the same (if you had a proper stereo setup), how would having a stereo setup save you?

Not criticizing, just confused :(

Yeah, the multiplicative, but if there's a major imbalance at relatively low load (closed loop), it seems logical that it would likely be much worse as it choked off in higher flow (as in not something a linear scalar would fix).  That's just my gut feeling though, and that may not be the reality.

I have every confidence in my fuel setup.
I have -8 AN rails with an -8 AN hose between them.
The FPR is in the end of the chain and I have more than enough fuel flow. Many many race cars run the same setup with no problem.
And my LTFT is like -2% so there is very little correction being done at WOT.
I do not agree with this theory... but feel free to prove otherwise :-)

Exactly

Yes, the trims should be significantly different on a large pump, cross over rail system and they cannot be different with the hack.

You mean the rails with threaded in AN flares that taper down to approx -6 size at the fitting that expand to -8 hose ID and then contract again at the next fitting?  Yeah you will have the problem also, it isnt a theory.

You LTFT is based off an average of one rich and one lean bank at this point.  Hardly optimal.

The straight six guys generally have one long consistent rail.  With one long rail or a V setup and big lines, rails and fittings it isnt as much an issue.  For the way many of us run our rails, if we have all -8 or -10 sizing we would be better off.  All these aftermarket rails like IE and 034 use fittings off the rail that are -6 sized so there are restrictions and then in turn losses from one bank to another.  If we split them there will still be restrictions but at least the banks are seeing the same exact flow and pressure.

The system still applies the LTFTs, even in stock form the s4 in particular will trim each bank a little different due to this issue and it is no big deal.  When remove the ability for it to trim each bank independently this happens.

Those of us who were using emulation never realized this because stuff was still floating around.  We were still feeding different inputs and having the banks get different trim levels.  In our quest to stop that floating around this is what we have gotten.

The line from the last fuel rail to the FPR is -6 AN anyway...
As for the LTFT being an average... indeed it's an average... I could have one at -3% and the other at -1%... at idle and light load, hardly relevant when my WOT "average" from the wideband was used to dial everything in under WOT... I also tune a little richer than most on street tunes on purpose... to make sure there is a little extra fuel always in case of any variations that may occur.

I think perhaps you should show some logs to back your claims... if you have your sensors setup as you say then it should be very easy to get logs to back up your claims right?
"If" you are correct, then data to prove it would help everyone I think...

But... when I put my built engine in, I'm planning on installing another LC-1 unit anyway so that I can run one in each back. I do want to do everything I can to make the built engine higher power setup as safe as possible.
Perhaps I'll install some pressure transducers n each rail sometime and log them externally to prove it one way or another.

I run -6 in to the rail, -8 crossover, and -6 after FPR.  My rails IE rails are -8, and that's why I went with them over the -6 that 034 had.  From what I was told IIRC by IE, was that my setup should be good for way more than what my motor can handle.  

What are things that can be done to monitor the system outside of throwing the narrowbands back in the mix?  The other V style motors I have dealt with only use those trims at idle and cruise, every since the newer motors started running sequential fuel injection.  At WOT or above a certain RPM they all went batch fire.

Drilling and tapping the rails for pressure sensors would be easy enough. 

The more I think about it, the more I feel the fuel rails are not the real issue.  There are people running the stock fuel rails making tons of power and they are smaller than the upgraded stuff, have less capacity, and are pretty much built the same way as we have laid ours out.  They feed from one rail, and have the regulator on the other.  

No matter what, with this hack, if the motor is using the trims all the way through WOT, we are screwed any way it goes. ???  Kind of a question....

add:  I keep thinking about it, the hack makes them trim exactly the same.  So I guess I can see why the hardware needs to be top notch. Imma stop now, because the more I rant the more I confuse myself.

Its not a pressure thing, its a volume thing.  The regulator will keep both rails consistent in pressure but that is only part of the equation.

The fuel rails are not the issue by themselves, its the combination with the hack.

BT, dude i like you but you need to learn some respect, your quips towards me on this topic show a total arrogance on your part yet you are operating off your guesses.  I am a shop and person that builds these things and spends countless hours refining them, I have been building them longer than anyone and I dont have to give advice on what I find through my own R&D that will in turn help everyone else and save other people (even shops that compete with me) trouble and time, but I do.  I give it because I care about the cars and the advancement of them.  Stop thinking you know a lot about the s4 in general because you dont, you have shown this to me numerous times.

I tried to help you on your MAP issue, remember when Jibby on AZ told you it was wrong?  Yeah that was coming from me.  Your response "this is how i always do it on my DSM" or whatever you said, yeah well guess what bud, this isnt your dsm.

I posted this as a warning, take it or dont.  The fact that you dont understand what is happening here alarms me considering who you are supposed to be.  Your thinking that the last line leading to the FPR matters alarms me.  The fact that you want to just come out and disagree and then lead other impressionable people to agree with your lack of knowledge alarms me.  Take a step back and use your brain on this one.  I generally assume that other people are smarter than me, although right now I feel i am talking to a brick wall that doesnt want to accept something.

Tell you what guys, pull your AN fititngs off and look at your fittings going into the rail, what ID are they?

If you want logs on this, go install pre turbo sensors and log it, the issue is very clear.  I 100% guarantee that you would see the first bank trimming negatively and the second trimming positively as it should and like yours IS NOT doing due to the hack.

end rant/

I never once didn't respect your experience and if it came over that way I'm sorry.
My point was simply that I (and others) must be missing something in what you are saying.
The point about logs was also to help everyone understand what you are seeing...

As for the map sensor yeah I was wrong on how me would react to that and I said that... I don't remember anyone saying about it until I asked about it on here.

Jared, any idea why the guy in the AZ link I posted was having issues with a bigger/different crossover tube (on a stock, twin O2 setup), and why the ECU wasn't just compensating with the trims as it should be with that?

Got it, I think.  Although the pressure across the rails could be the same the volume could be different.

Eh, screw it.  Still confused, don't care anymore.  In my mind there is one setup that should work, and one setup that will work.  I will feed both rails.  DONE.  I don't want to chance it over something so simple. 

Thanks for the heads up. 

Rob (Rtl the guy in that post) went to 034 fuel rails and his problem went away...

He probably didnt give it enough time or sensors were not spot on.  Or maybe he had enough pump that it was really magnified on his setup.  I dont know his specifics. 

There is almost to be an expected 3-5% difference if you are running a lot of fuel up to the rail (an its a crossover setup, especially stock) and the bi-turbo setups, regardless of fuel rail setup will trim that out easily.  

We lose that ability with the hack and a single using a single sensor in the downpipe.  You go on non the wiser because the readings at the wideband/narrowband look perfect.  Ecu goes on non the wiser because its sensor input looks perfect.

I know, and the bank-specific-trimming ECU should trim that bad boy out, no?

Is there a limit to the percentage delta between the two banks?  Is it just the same limit as all LTFT's?

LTFT's are all calculated in different ml, rl, and nmot ranges.

Fuel trims are never compared between banks and there is no delta consideration.

Just got done doing my first test drive with the VMP HPX 5000 MAF sensor in blow through setup, replacing the stock MAF.
I dialed the boost down to about 15psi to be safe but provisionally it looks like I got everything pretty much spot on the first time :-)

The car did for sure run smoother with the HPX sensor compared to the Hitachi though, especially at idle where the Hitachi would have a random miss here and there.

How much tweaking did you have to do?

Actually none really.
I used the MLHFM already posted and scaled my KRKTE back to where it should be (MAF was underscaled before)
STFT are right around 0% :-)

I will test at higher boost tomorrow when I test some new ecu stuff.
I'm hoping that I won't have to underscale this MAF at all once the 5 bar MAF support is in place, which I'm hoping to test tomorrow.

Did you use stock KFKHFM or reset it all to 1?

All 1

DAng it.  Mine was off in the top end.  Ignore the lean spike, that was from 22+psi boost. 

What form of enrichment are you using? Knock based? Egt based? Lamfa?

I use knock based.  I have been playing with LAMFA in the high load, higher rpm regions.  Last column, from about 4500 rpms ~12.5 AFR from there down.

As of today, I have this.  It is kind of a pain in the butt playing games with the three maps.   I add more boost, trick the MAF via kfkhfm,  compensate via KFLF.  I really wish I had a 4-5 bar setup.

I would use KFKHFM for the tweaking from now on...
Maybe I don't know what I'm doing but when I underscale the MAF I go down with MLHFM and up with KRKTE the same amount.
So for example I'd multiply MLHFM by 0.8 and then multiply KRKTE by 1.25 to get it back up where it was as far as fuel / g... It seems to work for me *shrugs*

Hows your STFT loking at idle after being warmed up fully?

I'm working on that pretty hard... soon... very soon

Well, yeah.  Shifting MLHFM was "OK".  My Fuel curve FOLLOWED the requested when boost was reasonable, under 22psi. With tweaks to KRKTE I was able to get them to line up pretty good.  After 22psi, the only way I been able to get fueling to go proper again was by underscaling the load in the areas needed.  KFKHFM was the only table I saw that was able to do it this way.  It took me a while to figure it out though.  I would add to kfkhfm thinking it would add fuel, and it was actually making things worse.  I really couldn't  tell until I had a bunch of logs to look at.  I even used KFLF alone to add fuel and it didn't do it as well as just by underscaling the areas needed.   I found that when I underscaled the MAF via MLHFM, I needed to do a lot of other changes to the maps that dealt with load.


I guess we are technically doing the same thing.  Its just that you are doing it to the entire MAF/fueling, and I am doing just the offending areas.

I haven't been watching my STFT at idle much lately.  Last I saw, and I am pretty sure it still holds true, they were around +5-8%.  I just left them alone since My focus has been the WOT/high boost part.  I didn't want to dial in the idle while I was making all those changes elsewhere, which would probably in one way or another re-screw up my trims. 

Add to the last log posted above.  It actually pulled nicely all the way to the top this pull.  Usually it feels like it starts falling off.

Lookin good! How is the fueling up top?

few posts up is the fueling from the same run.

Any idea why the guy's setup in that AZ thread didn't equalize itself?

There's way to many unknowns to really say. He ignored the questions regarding trims, how much was he driving, how was he driving, was he clearing codes...

AFAIK, pump headroom should overcome that, but I'm not going to pretend to be an engineer here.

just remember to keep an eye on ps_w every log...

I am way past PS_W.  I am around 25psi tapering to 23.

Then you should under scale some more.

Rephrase: I am past ps_w, and am compensating for it by underscaling so I can fuel past 22psi.

My best to date.  (patiently waiting on the 4 bar stuff)

I am having problems getting boost to spike sooner without waaaay overshooting.  I am aiming for 27psi.  I have been moving the RPM axis for the PID around and trying to dial it within the rpms ranges that need the most manipulation.

I am pretty sure there is quite a bit left in the setup, boost isn't leveling off until 4750 which is quite a bit longer than I want.  It also doesn't stay flat, I am still working on that part slowly but surely.  Timing is a lot less than what I wish I could run, but that's pump gas for you.  Water/Meth should take care of that.      

Lastly it might be time to drop the spark plug gap down a bit more.

What is your current scheme for controlling the VGT and wastegate (ie. N75 only on VGT, no wastegate?)?

post up the .csv .. i'd like to take a look at your WGDC

What's your AFR look like?
It looks like you're losing a lot of power to timing there...

It is running around 11.4-5 at redline.  A bit richer as I dial in boost/underscale MAF, BLAH BLAH around 4-5k.  I will post up the log later this afternoon and you.guys can give me some tips.

I am using a wastegate actuator to control the VGT.  I think I have it set to start opening around 18-20psi.  The n75 is controlling the external WG with a 11psi spring.

From my pull this morning.  I upped some duty and overshot a bit.  I still cant get the ramp up to stay moving until desired boost.  I lost a bit on this run. 

updated 3.33 fats log with .CSV from that run both logger and zeitronix.

I recommend you check your EGT's.
Too lean, too little timing and actual AFR all over the place.

Yessir

ADD:  what kind of things help lower EGT?

You can think of timing as efficiency.
More timing, more energy spent pushing the piston and less spent as heat going to the exhaust...

Also, more fuel helps cool the cylinder and reduces EGT as well. As long as you use a non-vapour fuel... Don't try it on LPG.

Exactly prj.
You can also melt exhaust valves from running too little timing as well. I've seen people do this.
You always want to run as much timing as possible without exceeding MBT and without causing knock...

So I've been using the KFMHFM table to dial in lean and rich areas of flow from the MAF and strangely enough the more i tweak it the closer it gets to the stock M Box KFMHFM...

So whats ideal timing look like?  I cant get anymore timing into it without major pull.  I was compensating for the lack of timing with more boost. 

With your HPX MAF's MLHFM?  When you threw in the HPX MLHFM, did you 1 out KFKHFM before you started tweaking the MLHFM for blow-thru?  If so, that's really weird that after essentially decoupling the two maps, you would still end up wandering towards the same stock KFKHFM values.

Exactly. I started with KFKHFM full of 1's.
The MLHFM was massaged a little down low... but I wanted to keep it smooth so did corrections then via KFKHFM.
I thought it was odd as well... makes me think it's more the VE and flow of the Intake Manifold and Heads more than the air intake system that is coming into play there.

Best way to blow it up.
Log your EGT's.

Back on my setup...
I'm getting things dialed back in with the HPX MAF sensor and 4 Bar MAP sensor.
This was the drive to work today. 2 runs 30 mins apart. Very consistent.
Still very much on the rich side.

(http://i1357.photobucket.com/albums/q755/britishturbo/51204bar-r5-3rd_wot-boost-afr_zpse65c1464.png)

(http://i1357.photobucket.com/albums/q755/britishturbo/51204bar-r5-3rd_wot-fats_zps84583fbc.png)

(http://i1357.photobucket.com/albums/q755/britishturbo/51204bar-r5-3rd_vd_zps1a4bc2c3.png)

More boost with too little timing is just going to be even closer to melting things right?
Are you referring to calculated EGT's or actual EGT's? In your experience is the calculated EGT still accurate with this kind of turbo setup?

Oh, that makes more sense. I thought you did significant corrections to the MLHFM first, then went to KFKHFM and it was still going towards stock even though you had already corrected for it via MLHFM. Makes more sense.

I like where you are. 11.5 is great.

Your PID needs a tad more work; can you post a log with the various PID variables? (P, I, D, wg before/after linearization etc)

Also, i dont really understand your requested boost...

I've always tuned for between 11.5 and 11.8 on pump gas setups... 11.5 when I want to be on the safe side for sure...

The PID I haven't really touched yet, I have only changed the KFLDRL...

AFR should be mapped based on EGT. At 11.5 you are killing power. Even if you can run more timing, flame speed goes down, so running 0.78 lambda vs 0.82 you are going to make less power at the same timing.

Simulated alone EGT's are worthless, log actual EGT's.
Then you'll know where the simulated ones are.

Yeah I'm tuning for 11.8 to 12.0 on the 2.7... that's why I said it's still on the rich side. I can tell from where I was dialed in around 11.8 on the Hitachi MAF that I'm down on power where I am...
I agree with you there for sure.

As far as actual EGT's go are the narrowband stock ones good enough to tell anything? Or are you referring to wideband RS6 EGTs? I'm assuming you mean stock ones...
I do log my Stock EGT's and have yet to see them move north of the bottom range of 945C.

I'm actually working on a plan to use these wideband EGT's and interface with the ME7 using a 0-5V to frequency converter.
http://www.exhaustgas.com/ProductDetail.asp?ProductID=1890&DepartmentID=13&CategoryID=56&RepID=&BasketID=

How do you log EGTs on car's w/o real EGT sensors on the dyno?

Install an aftermarket EGT probe.

If you don't wantto buy new ones you can get non-flight certified setups at a lot of aircraft wreckers for VERY cheap. They also have the added benefit of being vibration proof.

Where? Tailpipe is too late...

pre-turbo is most accurate. I find that pre vs post turbo readings can differ a couple hundred degrees.

Pre turbo is the only thing that counts.

I can't see installing that on EVERY car that goes on the dyno ... especially a 2.7t

Post-turbo is OK, if you have experience with pre vs post on a car and know the approximate difference.
For 2.7T it is easy, you already have the factory sensors. Just install a factory sensor and remove it afterwards.

You mean an RS6 wideband?

You don't need a wideband EGT, although it's beneficial.
It's important to not exceed a certain EGT, not so much know all the time where it's at.
Unless you have an EGT sensor on each cylinder it's pretty useless for mapping.

So I made a screw up today...
I was on the phone while editing my KFZW table...
It seems I typo'd a section of the table. The result was an 8 degree ignition timing spike at around 4000 rpm @20psi of boost...
This lifted the heads and blew out both head gaskets at WOT :/

Time to get cracking with my built engine...
Be careful guys all it can take is a typo!

dafuk? 

:(

That sucks but also shouldn't have happened.  Was it a tired engine, old gaskets?

OEM bolts?  New when installed? Torqued with a torque wrench

Yeah I know...
It was a used engine out of an A6... so used is used... no way of knowing the actual condition of the internals...

Ah well, 3.0 it is!

Since we are all here, do you think it is possible to wire in the 75psi Map sensor that my ZT-2 is using and run the 5120?

If you want crappy resolution...

I don't know what the resolution of the OEM TDi 4-bar is, but Marcellus' 5-bar is the high-spec stainless 5-bar that I've talked about elsewhere (I sold it to you, right Marcellus?), and will likely have better resolution than the OEM 4-bar.

And yes, you can wire it in (you may have to change the offset though).

Yup its the 5 bar one.  I know the stock MAP sensor is pre-throttle body, do you think there will be an issue since the Zeitronix is reading from the nipple on the top of intake manifold?  Will there be any ill effects of splitting the wiring to go to the ZT-2 and the ECU?

Yes. The ECU expects the MAP signal to represent pre-tb pressure. Take another look at the FR..

I figured as much.  Just checking.  Plus, I remember BT having issues with his MAP post TB.

British Turbo had his map sensor post-tb, plumbed into the intake. He had some REAL issues running that way.

Yeah it doesn't work lol.
I learned that the hard way :-)

ME7 EXPECTS the MAP to be pre throttle body. If it detects vacuum it throws a fit :-)

That's the least of your problems.... even if you clamp the sensor to no show vac, the ECU approximates pressure drop across the TB to generate calculated manifold air pressure.... if your sensor is post TB, that number is completely wrong.

Yepp. Just don't do it is the easy answer lol.

Agreed

You can get a plug and throw it into the stock sensor's spot with a vacuum tube (which is what I thought you were originally intending to do), as you won't necessarily need to log with the Zeitronix anymore.  Though that obviously means that it wouldn't be hooked up to the Zeitronix, if you've got an in-cabin boost gauge with your setup.

I can easily make a plug/adapter for the stock sensor location.  I just need to figure our what info I would need to make it work in my file (re-read the 5120 thread). 

I really don't like the ZT-2 with the LCD for boost anyways.  If the numbers don't sit stable for long the display is really hard to read, especially while driving.  I probably need to look into it some more, maybe there is a filter option or some sort of smoothing....I dunno.  If I cant get it a bit better I will more than likely go with a different boost monitoring setup. 

you can add an inline restrictor if you want, it will smooth things out...

Yeah, they generally call it an 'anti-buzz' fitting or something. It's just a small restrictor that helps dampen the pulsations before the sensor.

Do you think and inline restictor/anit-buzz piece will help with the readout?  I will look into it further.  I had to make an adapter to go to a smaller I.D. hose to go to the sensor form the intake mani.  I guess I could remake one with a smaller I.D..

Don't do a line with a smaller ID, just do the buzz restrictor.  Look up the installation PDF of AWE's boost gauge to see a picture of what we're talking about.

A trick I've used many times to smooth out MAP sensors is yo put a small carb fuel filter in the line. The type you get on a lawn tractor.
Works wonders for smoothing out the signal.

I am a bit confused about the load right now.  On the freeway in 5th gear I floored it to pass, and I saw the boost shoot up to over 34 PSI.  In my third gear pulls I get a slight spike to about 28 PSI, and then fall back down to 24 PSI.  It is kind of weird to see the load go this high.  MAF is not underscaled anymore, so I can see why its so high now. 

Overall I really like the hack.  Thanks to all that made it possible.  It makes a novice like me a bit more confident.

Add a boost chart of the same log.  PID is going to be much better easier to do what I want it to know I think.  I also couldn't get the car to boost to 25psi by my target 4250rpm before without seriously overboosting first. 

You should graph wgdc (and the rest of the PID components) if you are trying to diagnose over/underboost issues...

I don't think  I have truly had to worry about load tables since my load never went over ~209-212ish.  One thing I am working on is getting my LDRXN in line with the log.  The more I add to the map the more load the car makes.  I guess I was really being capped by the load limit.  

With that being said, I have a few questions.  Since I am now getting a lot higher load, should the IOP/IRL maps be modified to suit?  I am thinking of changing the axis for torque/timing to so it matches up with what my car is actually doing now.  I don't really want to bother the lower end of the maps too much.  Every time I have messed with IOP/IRL/AXIS I ended up with either a really great WOT and crappy part/cruise, or good part throttle and weak WOT.  Should I even bother with those maps right now?  Any suggestions as to which direction I should head with this?  

 On the screenshot below I have already modified the LDRXN map according to the log from this morning.  

ADDED timing CHART:  The 5120 hack has helped my tunes all across the board.  I haven't changed my timing, and since the change over the car is running a lot more timing.

So my car seems to generate about 270 for load with unmolested MAF so far.  Spool is getting faster and faster.

I am going to attempt to run these settings for the maps in the below screenshot.  I multiplied the lasr few rows by 10% to get to about where my load is on my logs.  Ran trough the maps briefly and it looks like they should be ok the IOP cells kind of correspond to the IRL cells "roughly". 

The reason for all of the IRL changes was to get the desired boost a bit better.  No matter what I did today the boost profile looked almost exactly the same.

You really want LDRXN to drive things anyway at WOT, not IRL... which is what is happening right now.

Please post WGDC. If you have LDRXN come up too *late*, you are leaving spool on the table because the pid is cutting DC to keep the ramp away from req boost. Even worse, the ECU might cut throttle plate angle. I know the wiki says early LDRXN is bad, but late LDRXN is also bad if it is killing spool

I know it is cutting boost around where the actual exceeds desired....I was hoping to add some more desired boost earlier to prevent the valve from dropping duty cycle. 

one log from a few hours ago, the second even earlier in the day.

forget about your IRL. just make sure that it is higher than LDRXN at max ped at all RPM points. Let LDRXN do the driving. If you are cutting DC during spool, bring LDRXN up early enough so that DC is at 95% until just before it reaches req. Then Q2 should cut dc WAY back, until req meets actual. At that point, Q1/KFLDIMX should take over.

Forget about IRL, but not forget about it?  :o  My IRL was set to previous logs and revision when my load was no where near it is now.  Before my load just flatlined at around 209-210.  Now, as I move the LDR up, the load seems to stop around 275. 

The last column of my IRL is 273.00 now...the 99% column.  My LDR max is 272.xx.  I thought it was common practice to somewhat smooth the map so it gradually goes to the new starting starting a few columns out. 

my post #440, a few posts up has the screenshot of where everything was (iop/irl) was set prior to the changes I just made.  Not a good idea?

If you are at WOT, the only column that matters is the 99% one. The rest isn't important until you get around to doing part throttle tuning (unless you are seeing torque intervention).

And if LDRXN is lower than all points on the 99% column, then what is in IRL is irrelevant at wot. Again, LDRXN caps IRL. Correct spec load should follow LDRXN (after IAT correction, of course).

Req boost will then follow cor spec load.

Are you still logging with ECUx? Is so, you really need to get ME7L running.

The torque model is nearly a closed system. IOP and IRL are not directly related at all. Tune IRL for desired load. Tune IOP so that mifa_w gives a high enough milsol_w without triggering any intervention.

If you are rescaling the IOP rl_w axis make sure you also do KFZWOP(2) and KFMDS.

In order for B_ldimxn (condition for negative correction ldimxr) and B_ldimxa (condition for adaptation limiting value I component LDR) to be set in LDRPID, rlmax_w needs to be at least 95% of rlsol_w.

Yeah, I still log with ecuX.  I have been too lazy to dig around in me7L to get it to work.  Every time I sit down and say I am going to fix the issues I am having with the program, I get distracted and put in on the back burner. 

I am going to atempt again today.

I made some changes to IRL/IOP/axis to reflect the new load.

Last minute changes to the map since last night:
bumped up IRL/IOP/AXIS to be higher than LDRXN

From what I saw in the brief log I was able to get, the load and boost are doing exactly what NYET said.  Apparently I had my maps set too low, and not following LDRXN because of it.  That is, if I am understanding it correctly. 

It definitely is noticed.  Before when load/boost was off, I could actually hear the turbo spool down when the ecu wasn't happy with what was going on.  This morning it was a much smoother pull.

This chart from the same log shows that WGDC.  I need to adjust my desired boost so the PID doesn't just fall out trying to limit boost.  Its a problem, but a nice problem...the turbo wants to spool faster and now.

Boost after a correction to the maps.  I am going to adjust LDR to keep requested around 27psi.
I am thinking of also moving the 4500 rpm point in my PID to 4250 since my turbo is spooling before this point now, and I need the duty cycle to drop around there.

My best so far.  This time without pushing 30psi  ;D.   Thanks you guys for the pointer on dialing back the boost.  I still thin the timing is low, and I still have work to do to the PID. 

I am still having trouble getting the boost to get to where its set and stay put.  I like how the car spooled with a with it set to 29-30psi.

Looking through my .bin, I have noticed that my KFLDIMX Mbar axis is 1/2 of what it should read.  i dont know how it happenned.  

I have been trying to properly define the axis I need so I can make the adjustments.  All I do, nothing looks quite right.  

what am I missing?

That mbar axis (stock) is plsolr_w (relative requested boost, basically plsol_w - ambient pressure), 0-1000mBar stock on ME7.1

so with the 5120 hack you should have 0-2000mBar range, assuming you have both plsol and pu_w scaled..

btw not sure if this is coincidence, but we're talking about it here too

http://nefariousmotorsports.com/forum/index.php?topic=3280.msg33048#msg33048

DUH, I totally forgot about the hack cutting things in half.

After seeing my AFR's take a crap across the board again, I went back into trouble shooting hardware issues.  I have been beating myself up lately with the very inconsistent fueling of the car.  One log its decent to ok, next log its pig rich, next log rich-lean-rich.   Come to find out my new inline fuel pump was taking a crap.  Only way I found out was by replacing it and with no other changes to the tune:

My fueling never really followed requested either it was too rich or too lean and did its own thing the whole way. 

What are you guys doing to get more timing into the motors without going methanol injection?  I still cant get the timing to stay decent around 4200rpms when my turbo is fully spooled.  

Should I get to a richer AFR sooner?

What fuel are you running?

I would target richer mixture earlier, say 0.9 upon WOT taper to like 0.80 at full boost.

maybe. Its gonna be tough w/o meth or lowering your CR

I am running WA state 92 OCT.

I just changed my KFLAMKRL to add more fuel faster according to load:

With the richer KRL I got this:

Reviewing my logs, I am still in the 12's for AFR when I am near full boost.  Is that something that should raise a big ole red flag?


ADD:

I just added some fueling via LAMFA.  I don't know How much this will help, but I will log it and see.  I have had to slow down the spool since if I let the thing spool too fast, the ecu pulls all timing out.

Yea. Are you out of IDC? Maxing ps_w? Post MAF intake leak? Pump on the edge?

No, its following requested.  I should have been more specific, I meant during ramp up/spool.

And the log that blew the motor:

It wasn't the run that killed it.  The car pulled just fine, it was when I let off.  There was a massive backfire, and a large hole in the oil pan. 

Fueling was the best I ever had it.  I was trying to get the load areas dialed in.

Sorry to hear it! First BT, now you. You STK guys are dropping like flys.

I can't figure why your motor would have given out.

Can you tell what failed?

Wow, sucks to hear.  At least it gives you an excuse to come back bigger and stronger (trying to look at silver linings):)

With all of these blown motors lately, we may have a statistically significant number to try to draw some conclusions from (except for the fact that the majority of them were on non-ironed-out setups and questionable failure causes).

I am really tired of this car, but I am too far in to stop now. 

I will be doing a very small build on my spare motor, rods and a refresh.  I dont plan on breaking records, I just want the motor nice and tight and able to handle a few of my tuning mishaps better.  If anyone knows of some rods for cheap readily avail....LMK.  I am trying to not let this thing drag out.  If the car sits for too long, I will probably loose interest.  My first instinct was to ditch the project and part it out, almost as soon as I saw the oil spraying from the bottom of the motor.

I know it was a tuning error that caused the failure, I just dont know what I did so I can fix it.

I think there is something wrong with my datalogging.  I have been getting very inconsistent runs FATS/AFR/Timing wise on the smae stretch of road.  I can do back to back pulls with some time in between to review logs on the same strtech of road, going the same direction and I can log FATS times of 5.xx to my major WTF pull of 2.44.  I know I wont really be apples to apples with the s4 guys since my gearing is different, so I really use the numbers as overall look of how the car is performing from revision to revision.  

I don't even know where I left off.  I have been trying to keep current while building the new motor.

Been logging for a few days ironing out some things.  Something feels strange.  It will build about 10-12 psi slowly, and then just ramp to whatever max boost is all of a sudden.  I have noticed the fueling is crap all over again, so I will deal with that while I have the low timing map in.  Maybe that is what's causing the bogs down low.  I totally forgot what I was working on with the high bar mapping.

The new setup has rods and 2.8 heads, a bit looser on the ring gaps.

I cant figure out why Spec load never goes beyond 280%.  Any pointers?

DHBN

I have been messing around with the DHBN map and still everything caps at ~280.  I have the map set to 4 in some areas and no change.  Irl and IOP are over 300 where I want it to be. 

I don't know if this means much but I accomplished last motors goal today!  

My first 500 + awhp log.....LOL

I used all the same constants I always have, so it's only relative to me.....shows improvement.  

Bad news is my 630cc injectors are @ 95% duty cycle.....time to upgrade