NefMoto

Good day all,

I really can not figure this out. I'm trying to get this tune all straight but i'm lost on where to look next. Under WOT the actual load suddenly drops once MAF readings exceed 420g/s which drops IDC and thus i run super lean. Now that specific point may be a coincidence but im not sure. And before someone says it i'll address the super rich a/f actual not meeting requested later.

Here's the log: https://drive.google.com/file/d/0B4w-boswX0J3ZVpGYUVhSEVEU28/view?usp=sharing

3rd gear portion of the pull starts at row 438

Spooky.......


Dillinger might know I assume you are not exceeding KFMLDMX?

Edit, I think you are?

Wouldn't I throw a MAF too high code for that?

Depends on the code, dual maf??  idk

Single MAF  b7 s4

The values in MLDMX need to be adjusted so that they are ~30 percent higher than anything you will see in normal operation. As it is, it appears as though you are exceeding this, at which point it changes to calculated values. As you are dumb (like me) and haven't bothered to properly calibrate the rest of your airflow maps, it's fucked.

:)

sure enough...that was it! Thanks guys. So much learning to do. Time to tackle the lambda issue.

No probs.

It is time to tackle KFMSNWDK and KFWDKMSN, so that in the event of a MAF failure you don't run lean and blow the car up.

The only reason this was obvious to either of us is because I just went through this the other day.

Oh wow...guess i got lucky with my post today then! That's very sound advise. THANK YOU.

That's why I said spooky!

when you max sensor voltage it will throw a too high fault.

In ealier code it also set bit in DLDR iirc this car wasn't doing that.  Well Daz's wasn't which gives me an idea on his file lol

ok so limit has been raised and now i'm throwing the damn implausible maf signal but not dropping load/ running super lean. It does however change the drivability of the car.

Below is what i THINK the culprit is:

(http://img.photobucket.com/albums/v223/qballc/S4%20Build%202013/load%20issue_zpsmvsogdb4.jpg) (http://smg.photobucket.com/user/qballc/media/S4%20Build%202013/load%20issue_zpsmvsogdb4.jpg.html)

There's normally a word for byte for max voltage, have you adjusted that? 

I'm not sure I understand what you mean... "Word for byte"? My volt max is 4.9.

I hit this error regardless of boost level even at 5psi.

Meant to say or not "for" sorry :)

There are ML limits too haven't explored that side much though.

What is maf voltage and ml when you hit the code?  compare against limits in file?

GGHFM here  : 

http://www.nefariousmotorsports.com/wiki/index.php/GGHFM_57.60_(MAF_Meter_System_Pulsations) (http://www.nefariousmotorsports.com/wiki/index.php/GGHFM_57.60_(MAF_Meter_System_Pulsations))

I'm using me7logger and I can't find MAF voltage so I can log it...

You can back calc it via the curve then or use IDA to find it.

KFMLDMX is signed. What hex values populate the map?

No more midrange errors.
Also, MAF max (mldmx) has plenty of headroom as seen in this snapshot but is still it cutting load. However it is NOT throwing an EPC/CEL light.

What's knock?  Injection time seems to drop from 14.5 to 10.5 pretty quickly, why?  Also, why is lam 0.66.  Surely that's black smoke time?

knock is around .75~1.5

Injection time drops with load suddenly, that's what I'm here to figure out.

Lambda is rich however O2s read about half a point richer than actual (rear turbo car) due to exhaust pressure

Also injectors aren't fully dialed in according to my fuel trims.

I'll gladly send someone my bin file for advice and another set of eyes.

Could be ATR, logging time!

Could be a lot of things!  What is your lower lambda limit?  Why is it going that rich, etc..

seems like something is capping RL_Sol...whats responsible for that cap on an N/A ecu?

Send me the bin and i'll see if i can help :)

gimme your email

just upload somewhere and pm me easier ;)

https://drive.google.com/file/d/0B4w-boswX0J3eHhlaXFzU0pFSTQ/view?usp=sharing

Going by the name I'm assuming not stock, can you send the ori also?

STOCK:

https://drive.google.com/file/d/0B4w-boswX0J3eWNaM3lPdjdYTWM/view?usp=sharing

You need to look at the conversion maps....

URL/PRG.

alright and what am I looking for? I know they're there to compensate for residual exhaust gas due to cam timing, back pressure, intake pressure, and what not.

Sorry mate i thought I had info/full map pack on this ecu but alas not!   It's to time consuming to follow code as I habe a protect to finsh

ok thanks for trying

I would posit that the factory calibration has made a mistake here, and these were in fact responsible for similar issues with mine own conveyance.

even if they did, I'd be stabbing in the dark to make changes to those.

You should try to familiarize yourself in that case.

the rabbit hole gets deeper...

Bump, this is still a serious issue

Do you have a recent log with MAF voltage?

I can't measure MAF voltage with my ecu

Going to be impossible to tune unless you have a more useful logging setup ..

I mean what can I do? I'm using visual me7 logger my ecu just doesn't have a MAF voltage block.

seems here the requested load just drops.

why not use ecuxplot to actually graph that?

ok sure

Couldn't you just use a multimeter, and run the wires into the cabin? Put it next to the dash, and have someone take a vid during your run.

if I was capping out MAF voltage I'd see the g/s flatline correct? Cause according to the log they're still increasing

That log doesn't show a MAF dip.

You're trying to diagnose two separate things at once. Pick one at time, focus on that. Also, post the logs.

BTW I don't think you're capping MAF. I just want to see voltage during the dip you saw.

The requested load issue is a totally different problem.

Also, what is the hardware difference from stock on this setup?

I'm not trying to diagnose two difference things. I'm running a stock MAF and housing. I think/assumed there is a correlation between the MAF and requested load drop. It seems to happen right at 420g/s. I don't see a voltage dip. Never claimed to. I solely want to diagnose the req load drop.

(http://nefariousmotorsports.com/forum/index.php?action=dlattach;topic=9438.0;attach=20073;image)

The topic says, MAF reading normal and rising, but one of your other logs show the MAF dip.

Please post the logs

Preferably one that has ALL the load request variables.

You may have to do some IDA work if ME7Info doesn't find enough of them.

Also, what are the hardware changes from stock?

Ah, thats because of i hadn't dialed in my boost curve yet so i'd spike and then level out some and then start dropping boost in the upper rpms.

I only have TWO load variables i can log. Link me to this "IDA work" you refer to. I'm not familiar...

here's the log.

Hardware changes include:

one twin scroll turbo charger
one air/water IC

stock everything else.

here are my me7info variables list with rough translations...

Wait, you are adding a turbo to a NA V8 w/o knowing how to disassemble the ECU code with IDA Pro?

i'm not the one tuning it.

So shouldn't your tuner be able to provide you at least a few RAM locations to help logging?

probably, i'll ask. but how do i add those ram locations to the logging program?

RAM locations are specified in the .ecu file generated by ME7Info, where else? :P

AH, DUH. so open in notepad and edit the bish. ;D

I was hoping OP would make a new thread since in the beginning there was 2 different issues to begin with and now that I'm in involved there is only one.

Kinda hoping this forum could help OP on that since I'm working pro bono on this. What should OP log?

Ya, i think it is quite clearly slow path intervention via cutting load request.


The load request path, specifically.

was digging thru the FR and found something but maybe it's nothing...

"At higher speeds it is possible that the available theoretical maximum injection time is not sufficient to obtain the required target torque. Therefore, an injection time timx_w that is larger than the maximum possible injection time timxth_w is deployed until the desired torque is withdrawn and timx_w is not larger than timxth_w. For this purpose, the control error dtimx_w is assigned to a PI controller. When the controller is active, the output controlled variable mitibgr_w represents the desired torque. When the controller is inactive, mitibgr_w receives the value 100%. The desired torque in %MDBGRG is obtained by initializing with mifab_w and mitibgr_w. In order to avoid jumps in the nominal torque, the integrator of the integral component is initialized with mifab_w."

MIFAB_W  =  Limited indexed driver-desired torque

Or is it possible it has to do with my ecu's version of "PS_W'?? Hitting a cap...?

Anyone?

Without logging the load request path, you're not going to find any answers...

Add a turbo to an NA ECU, you'd better have a full complement of tools at your fingertips, or forget it.

which variable(s) cover the load request path exactly?

Backtrace rl_w in the FR.

minimally:

ok, cool I was digging thru the FR for hours last night, hence how I found that fuel injector time meeting target torque I posted about.

err I mean rlsol_w (requested) not rl_w (actual)

sadly i only have rl, rl_w, and rlsol_w in this wonderfully NA ME7.1.1

SOLVED!  ;D

Finally finding an a2l really helps.