Lean transition into PE and lean spot during WOT

[Tre-Cool]

Dyno Graphs as promised. Rpm vs DC & RPM vs Duration. you can see around the 3800-4k rpm mark it was doing the weird stuff

power-v-duration.jpg

power-v-dc.jpg

I did another graph of a run time and for this car it was always doing it around 3 seconds into it. I'm generally at WOT from a 2200rpm hold. Entire run is about 8.5 to 9 seconds.

[cc-rider]

Absolutely. I've been using it for years. I setup the pe enrichment table to something like .87 lambda, but then use the open loop tables to run up to 1.25 EQ for the higher map/boost ranges.

HOWEVER, this only works if the car is permanently in open loop. the second you enable closed loop again the tune ignores the table entirely.

Are you certain that the open loop table stops working after enabling closed loop? I know that the computer will look for whatever the richest setting in all the tables gives it and use that. For instance, if you have the PE table set to deliver 12:1 afr but the open loop table set to deliver 11:1....then it will default to the open loop table and deliver 11:1. I know for a fact that I've seen this happen during some tuning runs when I was playing around with this. However, I may have been open loop tuning.

But I think I've also seen this phenomenon occur while running with closer loop enabled. It's why there are all the temp tables and every other table in there. You may have to be in PE/open loop for them to read the tables and use them...but I don't believe you have to keep closed loop permanently disabled in the tune to take advantage of these tables.

I can understand that the lean spike and injector pulse phenomenon you posted will only work in open loop. But I think the map table will work regardless.

Perhaps I'm wrong...

[cc-rider]

Absolutely. I've been using it for years. I setup the pe enrichment table to something like .87 lambda, but then use the open loop tables to run up to 1.25 EQ for the higher map/boost ranges.

HOWEVER, this only works if the car is permanently in open loop. the second you enable closed loop again the tune ignores the table entirely.

Dyno Graphs as promised. Rpm vs DC & RPM vs Duration. you can see around the 3800-4k rpm mark it was doing the weird stuff

power-v-duration.jpg

power-v-dc.jpg

I did another graph of a run time and for this car it was always doing it around 3 seconds into it. I'm generally at WOT from a 2200rpm hold. Entire run is about 8.5 to 9 seconds.

Very interesting charts. Thanks for sharing that data. These damn ecu's seem to have something coded in at the 4000rpm area that we don't yet know or have access to.

What did you have your dynamic airflow disable rpm set to for this run? Was it the stock setting of 3800 or was it lowered?

[Tre-Cool]

Are you certain that the open loop table stops working after enabling closed loop? I know that the computer will look for whatever the richest setting in all the tables gives it and use that. For instance, if you have the PE table set to deliver 12:1 afr but the open loop table set to deliver 11:1....then it will default to the open loop table and deliver 11:1. I know for a fact that I've seen this happen during some tuning runs when I was playing around with this. However, I may have been open loop tuning.

But I think I've also seen this phenomenon occur while running with closer loop enabled. It's why there are all the temp tables and every other table in there. You may have to be in PE/open loop for them to read the tables and use them...but I don't believe you have to keep closed loop permanently disabled in the tune to take advantage of these tables.

I can understand that the lean spike and injector pulse phenomenon you posted will only work in open loop. But I think the map table will work regardless.

Perhaps I'm wrong...

You are correct it will use whatever is richest between the 2, but for some reason as soon as you turn CL back on. It ignores the Open Loop Map enrichment table. I kinda use a similar technique on the LS1b platform using efilive's cos. I set the map enrichment table upto 105kpa then use pe enrichment at a designated higher map at say 125/130.

E38/E67 just wont play ball with CL on.

All cars i tune are Speed density only, none have a maf fitted.

[Tre-Cool]

Very interesting charts. Thanks for sharing that data. These damn ecu's seem to have something coded in at the 4000rpm area that we don't yet know or have access to.

What did you have your dynamic airflow disable rpm set to for this run? Was it the stock setting of 3800 or was it lowered?

This particular car it's set to 8000rpm. It's also the only car i've ever felt/experianced this issue with. I was starting to think it was a potential issue with the LPG system the car also has wired in until i found this thread. But I also couldnt see why or how the LPG system would effect the tune every run at that particular rpm but if i start the dyno run at a higher rpm it doesn't happen at the previous rpm.

[Niemer]

Yeah, it's not a function of RPM, TPS, MAP but rather time after PE is enabled. It always seems to be ~2 seconds after PE kicks in.

[BigDaddyCool]

So i take it this was never solved? I now have this same issue after changing from my stock cam to a ls9 cam in na. Pretty much right smack on 2.1-2.3 seconds.

[GHuggins]

Is everyone having this problem on 6.0's? Or is there any similarity on the setups?

[BigDaddyCool]

How about is everyone only having this problem on the Custom OS? ..... Has anyone who's ran into this problem switched to the stock os on a maf only tune to see if it made any difference? Because it really seems like PE doesn't drop out according to the scanner but it's awfully like it switches back to stoich, then pe enables causing the transient to kick in as well...... Case and point my scan below.

[GHuggins]

I was leaning towards a torque command causing this for that reason, but I can see it being the custom OS possibly too... I was wondering about 6l's because of the torque model - right around the rpm area where everyone seems to be having this issue - the torque model drops off - I was wondering what would happen if the torque model was changed... I can rework yours and post the tables to try if you like, but it would just be a test... That's if switching the OS back didn't prove effective...

I also assume you've tried killing all of the tm? Including Tc?

[BigDaddyCool]

the cars own calculations/torque model etc could be the issue sure...... canning TM doesn't fix it. By all means send some tables etc, and i'll give them a go.

[10_SS]

How about is everyone only having this problem on the Custom OS? ..... Has anyone who's ran into this problem switched to the stock os on a maf only tune to see if it made any difference? Because it really seems like PE doesn't drop out according to the scanner but it's awfully like it switches back to stoich, then pe enables causing the transient to kick in as well...... Case and point my scan below.

this car's Voltage is all over the place for some reason, 14.8 then 0.3 sec before the lean spike it drops to 12.4v, mine never does this and I have a boost a pump. There's also a lot of burst knock being registered right before and during the lean spike, that is strange. Then the lean spike occurs and VE Airflow and Cyl Air are low/constant right when the lean spike starts, 0.59/135, then right after the lean spike almost as it's being corrected by a spike in both.. jump up to 0.75/185... Is everyone seeing that at the lean spike?? AFR also is 13.3 before the spike with the lower airflow, and richer at 12.6AFR after the lean spike with the higher airflow...

Are these all automatic cars having the lean spike? I wonder if it's something due to tuning process maybe shifts are getting in there on some of these, messing something up?

I didnt see a tune file, but I would really look at why the airflow jumps way up right at the end of the lean spike and carries on properly, the MAP says constant and raw MAF HZ is steady as well so unless a table is switching or something like that, airflow should not be changing like that.

Lean Spike.PNG

[GHuggins]

This log is a little different than the other one he posted, but voltage doesn't change enough - I don't think? - to cause this issue / air fuel goes overly rich afterwords - possibly because he over compensated with the MAF - didn't remember it doing that in the other log?

It is interesting that air mass doesn't increase until after the lean spike and then it starts going up...

It reminds me of tm kicking in where the engine will go to stoichio, but not necessarily show it being commanded?

[LSxpwrdZ]

I've seen this on nearly all the blown G8's that I've done. Personally haven't figured out what is causing it. And like you said it's always 1.9-2.0 seconds after PE enables.

I also have a dyno dynamics it didn't matter what RPM was on the ones I've tested. I can hold the RPM at WOT steady and it's always that time frame after PE enables that the injector pulse spikes lean.

Only thing I can think of is there is an airmass filter or compensation table based on time that we do not have access to.

[BigDaddyCool]

Alright........... I have it fixed, rather I have a work in progress fix. I got an idea of what it might be.

Years ago, I found on the stock OS running VE only my map sensor would randomly mid RPM in PE drop all signal (you can imagine how it feels). I replaced it but it didn't fix the issue so I just went back to the custom OS which didn't seem to have the issue. So I think it's a bug in the operating system or an electrical issue? .... Besides the point. I'll workout exactly what the fix is shortly......... but it's one of three things.

1. PE Map Requirement (even at 15KPA. Set it to 0, same with Hysteria) - So if the map drops signal it doesn't pull PE. The maps so slow it may not show up on the scanner (guess). - BandAid or Officical Fix? Depends if it's a OS bug or an actual issue either way.
2. It may be 100% transient fuel...... Stomp Compensation (and the decay) ...... Impact overall vs the evap (boiling point) .....

I tried the torque model, it has absolutely nothing to do with that, leave it stock.

So I'll figure out what it is...... and I'll come back with a definite so we can solve this. I figure it's either zeroing out the MAP requirement for PE, or just transient fueling needing to be tuned (the whole lot).

[BigDaddyCool]

Sure enough in my case, zeroing out the PE MAP Requirement/Hysteria solved this for me. So either there is a issue with my map sensor/electrical issue or there's a bug in my OS. Either way ..... controlling PE via TPS works. There's a chance you could just put the MAP and Hysteria the same so they both align to zero if it drops out not pulling PE....... But I wont trial that, I will be moving on now I've solved it.

I've decided to can MAP requirements for Transient as well..... in case it's causing a delay with transient etc.

So that's it for me..... I don't know if it fixes other peoples problems like mine.... If someone else has this issue thats running in open loop to get around the PE situation, perhaps they should try my fix and see if it works.

[GHuggins]

Hopefully some others with this issue will give it a go and hpt can look into it

[Tre-Cool]

Is everyone having this problem on 6.0's? Or is there any similarity on the setups?

The car i mentioned is a stroked 6.0lt. so 402ci

How about is everyone only having this problem on the Custom OS? ..... Has anyone who's ran into this problem switched to the stock os on a maf only tune to see if it made any difference? Because it really seems like PE doesn't drop out according to the scanner but it's awfully like it switches back to stoich, then pe enables causing the transient to kick in as well...... Case and point my scan below.

I use the efilive stock OS version. So it's not a problem isolated to HPT. (however this car has been modified by both software)

[BigDaddyCool]

My last scan the problem returned. So more investigation is required.

[Rawr256]

I only see this on my G8 with boost with different injectors then LSA/LS9. Friends G8 with same cam I have doesn't have any issues with this, but another friend's G8 with a different cam is fine. Difference between the two blown cars is the injectors, friend's is running the stock LSA/LS9 injectors while i'm running the Deatchwerks 95 lb injectors. I've changed all the transient fueling and it doesn't seem to have an effect one or another.

While thinking of this my friend's V with the same injectors I have had a similar issue that only happened at the drag strip we chalked up to possible fuel delivery. He would launch the car and shortly after the 60' mark it would go lean then pick up right away just like mine does. I don't think it is anything with the injectors as much as maybe a table that isn't changed to accompany the extra fuel flow possibly, possibly hard cap somewhere in the background? My friend's V we were able to apply the fuel flow rate without needing to do the half/double Flow/AFR trick like on my G so i tend to think it isn't anything with that. He just upgraded to a ZR1 with the same injectors and we put in the data, haven't done any logging yet but will def watch no the dyno and street to see if it catches the same thing.