The e38 2-bar is killing me, Please Help! Log and Tune attached

[roastin300]

I have been battling getting this vehicle (2008 escalade) to run. I have completely lost me ass on this, but I will not give up on my customers. This is my first e38 and I am not liking it. Anyways, The vehicle is completely stock with the exception of a P1SC Propcharger that makes right around 160kPa at 6k RPM. I am running custom 2-bar OS, 2 bar sensor is in and scaled in the tune, 80lb Simens Deka with correct data that I scaled for this PCM. I am doing the 50% IFR scaling, Double Stoich, and half IFR. I failed the MAF, turned off DFCO, STFT and LTFT and tuned lambda error with the wideband over the new 2-bar VE table. When he left last time it was ripping pretty good. Within a day or 2 it would heitate badly. I assumed DCFO and LTFT were doing something. He cam back and I messed with alot of things. The result to get it to rev was to add quite a bit to the Fuel-->Transient-->Impact Factor-->Gas at the higher IVT's. It seems to be operating between 300 and 340 deg F IVT. The weird thing is this FIXED IT but ONLY WHEN IN PARK revving it....once you put it in drive and start moving, it falls on its face and the STFT and LTFT's go the opposite direction. In park is adds some fuel everywhere, in drive it removes fuel. I finally populated the Inpact Factory-->Gain table to ramp it up with increasing airflow and it helped a bit...but after an hour of driving it went back to driving aweful, almost as if LTFT's were active but they are OFF! Is there any other memory that is stored besides this? I dont understand what is going on, please any help is appreciated.

Attached is the last tune file and some logs where you can see me revving it in park (0mph) and then trying to drive it (>0mph).

[07GTS]

also have to set the maf dtc p101, 102, 103 to mill on first error for it to go into SD properly

[NJ_Phil]

I had a similar situation with my E38, SD and blower when I disabled P101, P102 and P103 or set them to "No MIL Light". It reported "Maf Failed" but was still using the MAF for airflow. I still had the MAF still plugged in, and it stalled when I unplugged it...even though it showed up as "Failed" in the scanner. Not sure if you still have the MAF plugged in though. You might want to plot Dynamic Airflow which is what the motor uses for VE cause your SAE Airflow is all over the place...another clue the MAF might might be in play.

Only way I got it stable was to set 101, 102 1nd 103 to "Mil on 1st Error" and disable SES. Then try a few starts where it would stall until the code was set.....then life was good

FWIW

[5FDP]

I'd also remove any channels you know you aren't using too, to prevent data skewing or just wasted space in the channels list. If you are running off VE only you can just remove any MAF related channels.

[roastin300]

also have to set the maf dtc p101, 102, 103 to mill on first error for it to go into SD properly

First, thank you so much for your guys feedback, you have no idea how much I appreciate your help!

That would be a drag if that was the only issue. Why is the e38 so particular? I knock the Gen 3's out of the park every time.

Okay, so if I put them on MIL on first error, do I check the SES enable button or no? I would assume this does nothing other than send the check engine light to the dash?

5FDP: I read many comments regarding it skewing the readings so I originally didnt have anything MAF related in the channels or graphs but then added the MAF status to be sure it was failed and it said it was so I thought life was good. I was also reading the airflow lb/hr to try and figure out where to add fuel in the Transient-->Gain table. If I get the MAF to actually fail now, maybe these transient tables can go back to stock? Its a stock vehicle, stock VVT cam (adjusted), etc.

That being said why would it rev just fine in park, idle, and in gear with the brake pressed down, but not once I get moving? Is the MAF ignored under a certain vehicle speed? I notice that the Fuel-->Open & Closed Loop-->LTFT Idle Cells Threshold Enable VSS switches off at a VSS of 2mph. It seems like right after the vehicle is moving is where I start having issues. Seems that MAF is ignored under this speed potentially? I couldnt find anything else airflow or fuel wise that was related to vehicle speed.

[roastin300]

Here are some screen shots of the Engine Diagnostics-->Airflow. Air the 4 tables fine or do they need to be maxed out so I dont hit any limits? I also raised the rpm to 8000 thinking that would eliminate these limits?

Also screenshot of changes made to the DTC's. The P0111 was No Error Reported...I changed that so the IAT could be monitored...is that a bad idea? Thanks guys

Aiflow_Screenshot_MAF_Fail_e38.PNG

DTCs_MAF_Fail_e38.PNG

[NJ_Phil]

Here are some screen shots of the Engine Diagnostics-->Airflow. Air the 4 tables fine or do they need to be maxed out so I dont hit any limits? I also raised the rpm to 8000 thinking that would eliminate these limits?

Also screenshot of changes made to the DTC's. The P0111 was No Error Reported...I changed that so the IAT could be monitored...is that a bad idea? Thanks guys

Aiflow_Screenshot_MAF_Fail_e38.PNG

DTCs_MAF_Fail_e38.PNG

That's close to my airflow tables but none of that should be checked until 8000 rpm so defaults would also be fine.
I fail my MAF L0 at 15,000 Hz just in case the E38 needs a little more of a difference between HI and Low fail, not sure how important that is

Odd thing I encountered is the car stalled 2 seconds after startup after I did a "Write Entire" and then started fine once DTC P102 registered. Then I cleared the codes and they never came back......until the next Write entire

[10_SS]

That's close to my airflow tables but none of that should be checked until 8000 rpm so defaults would also be fine.
I fail my MAF L0 at 15,000 Hz just in case the E38 needs a little more of a difference between HI and Low fail, not sure how important that is

Odd thing I encountered is the car stalled 2 seconds after startup after I did a "Write Entire" and then started fine once DTC P102 registered. Then I cleared the codes and they never came back......until the next Write entire

I've had weird things happen with the MAF fail due to the 1 or 2 second delay that you see there is still stock and not many people mention changing that to 0.0 or 0.1.

Basically would fire up on startup fuel or whatever then stall... found out this is because the times you see in your pic are still set to 1second so the MAF test wont actually run during that 1st second so it wont fail until that time is expired.... and write entire clears the MIL which re-enables the MAF then makes the test wait that time the 1st time you start after write entire... then once it's failed usually it starts up fine until it clears itself for whatever reason then it does the 1 sec thing all over again. Set them to 0.1 or less and that part should be fixed.

With the timers stock at 1second, it still uses the MAF during that time, wont fail it until after 1-2 seconds runtime.

And like someone else said, they like to put all 0's in the MAF FREQ table to make sure it wont run on MAF even if you think it's failed. This is a really helpful piece of advice, yes one more thing you forget to copy over and cant figure out why your MAF tune wont run lol

[NJ_Phil]

I've had weird things happen with the MAF fail due to the 1 or 2 second delay that you see there is still stock and not many people mention changing that to 0.0 or 0.1.

Basically would fire up on startup fuel or whatever then stall... found out this is because the times you see in your pic are still set to 1second so the MAF test wont actually run during that 1st second so it wont fail until that time is expired.... and write entire clears the MIL which fails the MAF then makes the test wait that time the 1st time you start after write entire... then once it's failed usually it starts up fine until it clears itself for whatever reason then it does the 1 sec thing all over again. Set them to 0.1 or less and that part should be fixed.

With the timers stock at 1second, it still uses the MAF during that time, wont fail it until after 1-2 seconds runtime.

And like someone else said, they like to put all 0's in the MAF FREQ table to make sure it wont run on MAF even if you think it's failed.

Thanks SS. I'll zero out my MAF and set the delay to .1.
I'm tuned SD only but like to show MAF frequency on the scanner because that's raw air coming in and I like to see what the puter does when generating dynamic.

[roastin300]

10_SS: Aaahhh Okay that makes sense. I will lower the delay time. Thank You! As far as putting all zeros in the MAF table, I suppose that in the event the MAF was not failed and you thought it was, this would tell you because the engine would really not want to run because it would be cutting fuel from the zero values?

NJ_Phil: Thanks for your discussions here. I assume if your monitoring your MAF readout even though your in SD, you have tuned the MAF Table? Can you lend me some insight towards the Dynamic parameters? Can you not use the estimated mass airflow (I think its in maths) to acquire the value? I am thinking that my stock fuel transient values will remain as is considering its a stock engine/cam/throttle body etc.

That being said, I had him CUT the yellow wire on the MAF so we dont have to deal with that asshole MAF Failing issue anymore. That way we still have the IAT in there to use until I can remove the MAF and install a standalone IAT and rewire it. He confirmed it runs way different but bad still. I am assuming this is from the huge amount of transient fueling I added to the tune to compensate for the MAF trying to pull fuel out. It was lean before, now its probably pig rich. See any issues with just chopping the signal wire as an extra level of MAF Failing?

[NJ_Phil]

10_SS: Aaahhh Okay that makes sense. I will lower the delay time. Thank You! As far as putting all zeros in the MAF table, I suppose that in the event the MAF was not failed and you thought it was, this would tell you because the engine would really not want to run because it would be cutting fuel from the zero values?

NJ_Phil: Thanks for your discussions here. I assume if your monitoring your MAF readout even though your in SD, you have tuned the MAF Table? Can you lend me some insight towards the Dynamic parameters? Can you not use the estimated mass airflow (I think its in maths) to acquire the value? I am thinking that my stock fuel transient values will remain as is considering its a stock engine/cam/throttle body etc.

That being said, I had him CUT the yellow wire on the MAF so we dont have to deal with that asshole MAF Failing issue anymore. That way we still have the IAT in there to use until I can remove the MAF and install a standalone IAT and rewire it. He confirmed it runs way different but bad still. I am assuming this is from the huge amount of transient fueling I added to the tune to compensate for the MAF trying to pull fuel out. It was lean before, now its probably pig rich. See any issues with just chopping the signal wire as an extra level of MAF Failing?

Chopping the yellow wire shouldn't cause any trouble. As for me, I just plot the frequency of the MAF on the scanner and scale it in with the Min/Max values to have it track somewhat with Dynamic Airflow (which is an available channel you can enable in the channel list) Gave up on dialing in the table since it's pretty ragged with a large 4" custom MAF pipe.

My opinion is you had the MAF affect VE when you tuned it and tried to tune VE around it, then changed transients to help out the weirdness. Now with MAF out of the picture, the VE map is out of whack. Trying to dial in transients with VE not right is going to be tough because transients can cover up a bad VE table so you think it will be ok in one area then get bizarre lean/rich tip in or tip outs. I'd put all transients back to stock, retune the VE, then adjust Impact factor or gain if AFR undershoots or overshoots during tip in and out. Increasing manifold area as some do, has never done anything for my setup except get fuel in a little quicker on tip ins but cause big delays on tip outs followed by lean overshoots before trims brought it back to stoich.

[roastin300]

Yup that is exactly my train of thought. I am hitting it early tomorrow morning, putting my wideband back in it and will see where the VE is at and then work on dialing it back in with the factory transient tables restored. I will report back after I tune and post up a log. I am feeling confident about this now. Thank you very much for your feedback!

[roastin300]

Quick thought...anyone know if the commanded EQ Ratio/AFR changes when a fuel transient is present? I dont recall ever seeing this and would assume so if the error associated with it in the VE and MAF tables show up.

Also, do you guys ever feel the need to smooth or set everything constant in the open loop EQ Ratio table and scalar tables when your tuning? Typically on NA setups I will make ALL tables including the PE have a Commanded EQ Ratio of like 1.13 to 1.15 so its not moving around to different targets. Yes its a calculated ratio but I have accomplished a rock solid VE and MAF table faster when I do that and less smoothing is required after the fact. On this boosted setup I would have to drive around rich (EQ Ratio 1.25+) all the time to safely dial it in.

[5FDP]

I never change the commanded open loop or stoich value while driving to help dial in the fueling. I just keep the stoich ratio for the fuel you are using and the commanded AFR in PE at what I desire.

When you are logging an EQ error it's always going to correct the error for whatever is commanded. I believe I am saying that right.

So if you had 14.68 as your stoich or lambda 1.00 it would be correcting to lambda 1.00 while outside of PE. Even when you dip into PE at say lambda .85 it's still going to correct for that lambda value the entire time the error is being logged. The error always corrects the fueling for whatever the commaned EQ ratio is.

That is why changing the fueling to run much richer is not needed in my opinion.

[roastin300]

Ya I get that. Just have had better and faster results with less iterations when you dont have a varying base table and more improtantly a bunch of scalars moving target AFR constantly. Either way you can come to the same result.

Another question. Apparantly the e38 PCM is a SOB in SD with regards to dialing things in because the IAT heat soaks and causes issues (not sure why I have never had this issue on older PCM's). Have any of you guys adjusted the cylinder charge temperature bias (increase the values) to stabilize the calculation for charge temp? If you lean on the coolant temp more then the artificial IAT's reported by a heat soaked sensor could be reduced. I know this needs to be set prior to tuning the VE because I did this on my nitrous s10 when I tricked the IAT to pull timing when nitrous activated, and it shifted my AFR alot (as expected) and I had to re-dial in my VE table to get fueling back where it needed to be.

[10_SS]

Thanks SS. I'll zero out my MAF and set the delay to .1.
I'm tuned SD only but like to show MAF frequency on the scanner because that's raw air coming in and I like to see what the puter does when generating dynamic.

I agree I use the raw HZ for sanity check... never trust one sensor!

[10_SS]

10_SS: Aaahhh Okay that makes sense. I will lower the delay time. Thank You! As far as putting all zeros in the MAF table, I suppose that in the event the MAF was not failed and you thought it was, this would tell you because the engine would really not want to run because it would be cutting fuel from the zero values?

NJ_Phil: Thanks for your discussions here. I assume if your monitoring your MAF readout even though your in SD, you have tuned the MAF Table? Can you lend me some insight towards the Dynamic parameters? Can you not use the estimated mass airflow (I think its in maths) to acquire the value? I am thinking that my stock fuel transient values will remain as is considering its a stock engine/cam/throttle body etc.

That being said, I had him CUT the yellow wire on the MAF so we dont have to deal with that asshole MAF Failing issue anymore. That way we still have the IAT in there to use until I can remove the MAF and install a standalone IAT and rewire it. He confirmed it runs way different but bad still. I am assuming this is from the huge amount of transient fueling I added to the tune to compensate for the MAF trying to pull fuel out. It was lean before, now its probably pig rich. See any issues with just chopping the signal wire as an extra level of MAF Failing?

I have also messed with transient fueling to the point somehow it was always adding fuel, like PE fuel, that's when I decided to go back to stock and just slight changes here and there... so you should probably go back to stock Transients - ALL OF THEM - your problem might be fixed... aside from having to completely retune the VE table (and MAF once you fix your cut wire). DOnt cut the wire.. just do the steps described above. 0'ing the MAF table will essentially be the same as cutting the yellow wire.

Like most people say.. only change one thing at a time and drive it a while... you change two things at once and this (my transient rich fueling condition during normal driving) can happen. If you make small changes one at a time and drive, your more likely to catch something if it messes something else up. But what tends to happen is one day you realize something is all messed up and you dont know when it started. Sickening feeling I know.

[NJ_Phil]

But what tends to happen is one day you realize something is all messed up and you dont know when it started.

Thought it was just me that thought like that .

I got to a point where I changes so much shit, the only thing that worked right was idle. Reverted back to the stock tune, reused my VE table and changed the basic stuff and car runs better than ever.
Got caught up so many time trying to tune around a bad VE map for areas I thought weren't important (like 4000 rpm @ 15kPa) that transients were fighting trims and I'm starting to see why a load bearing dyno is so important.

[roastin300]

Hey just wanted to report back that the issue was those 3 DTC's....and once I put the transient tables back to stock it drove pretty nice, but was just rich, so I dialed in the VE table over the course of a few hours and its running pretty good. A few things I noticed though:

1. At idle in PARK, changes to the VE tables (there are 3 in the 2-bar OS) does not change the idle AFR....I still cannot figure out how to adjust this. I changed base running air flow and throttle follower tables ALOT (reduced them to reduce AFR was my thought) and it made NO change to AFR. Anyone know how I can dial the AFR in at idle? It runs very rich when in Park. Put it in Drive and it appears to reference the VE table.

2. With the STFT and LTFT's shut off (zero'd/maxed appropriate tables) to dial in the VE, it seems that once they were turned back on, the correction from them did not follow the reported error that remained in the VE. Wondering if just because the scanner shows 0% for all fuel trims, if the dam e38 PCM is still doing something. I removed passenger O2 for the wideband btw, so only drivers side was there.

[roastin300]

Also there are thresholds for "Idle LTFT's". From the factory it says below 0% TPS to engage them. At idle I am reading between 8-14% TPS. So I raised the TPS threshold up and the LTFT's started to pull out the fuel at idle. Is this how it needs to be done? It clearly is not looking at the VE table at idle. I practically Zero'd out the idle cells in the VE and the AFR didnt change.