LT1 8051 tip-in lean? Anybody had success curing it?

[sherlock9c1]

Anybody in the 8051 world successfully dealt with the LT1 tip-in lean without scaling the MAF values, injector flow rates and cylinder volumes? I am tuning a bolt-on '95 motor with an intercooled Procharger setup with updated fuel system, GM MAF and fully characterized injectors. It's currently in MAF mode and the owner would like to keep it that way. The car starts instantly, idles perfect, and pulls like a freight train, but it's got a terrible tip-in stumble at idle that's particularly pronounced while cold. We've got a wideband on it and it's showing it goes immediately lean; no rich at all. During the rest of driving, the AFR is right where we expect it to be and the motor is well-fueled right up until redline (self-imposed around 5500rpm when we run out of injector). I have tried adding lots of fuel to the VE tables in the stumble areas but it doesn't seem to have made a difference.

Interestingly, I took a bone stock LT1 Roadmaster wagon, started it up, whacked the throttle hard and it seemed to have quite a bit of the same stumble!

I've read a number of other forums and many forced induction LT1s seem to have the same problem, and they all solve it by scaling the above values pretty severely, perhaps under the theory that the $EE has a tip-in enrichment that is nonadjustable, or at least nobody has found it yet. I did try a "half-MAF" tune where I scaled everything per above, and the car started, ran and drove decent, and seemed to have much more oomph during tip-in, but eventually the PCM learned it back out.

I'm wondering if perhaps any of the remaining mystery tables in the $EE definition might be something along these lines, perhaps a tip-in enrichment vs. coolant temp vs. rpm? Anybody playing with this? How many tables or areas in the $EE binfile are still "mystery?"

[steveo]

i did a bit of work in this area for sure

i did find "a transitional calculation of some kind" for snap throttle enrichment. it was tps bound.

seemed stupid simple, like if tps increases this amount since the last cycle add this much pw kinda thing, but i never pulled it apart and understood it completely (since im still a noob with this type of assembly), so it never made it into my xdf since i couldn't prove how it works. i'll see if i can find the relavent code again, it's been a while.

it seems 'pump shot' isnt a thing that needs as precise of calibration in a sequential port fi system, since the fuel is right there, not way up a wet manifold.

EE in default configuration seems to prefer a burst knock approach that retards timing on throttle stab so it kinda goes lean for a sec, but doesn't explode, and just goes from there. it also seemed from careful monitoring of O2 voltage that when in closed loop, the large o2 error from that momentary lean burst gets snapped back by CORRCL, and feeds it a small shot of fuel so the transition isnt so harsh.

actually the most successful approach i've found in MAF mode to improve responsiveness was to dick with PE qualifiers so you get a nice short burst of PE that doesn't last very long as MAP drops back down in short order. i got GREAT throttle response from that.

without the maf, you can get great results by adding fuel to the VE tables in the areas that need extra enrichment, and just slow the integrator down a bit so the corrections aren't as intense.

[steveo]

How many tables or areas in the $EE binfile are still "mystery?"

hundreds.

[1project2many]

it seems 'pump shot' isnt a thing that needs as precise of calibration in a sequential port fi system, since the fuel is right there, not way up a wet manifold.

There still needs to be an enrichment. The early ecm's used separate tables because response time of the ecm would not provide rapid enrichment. I was looking at one of the aftermarket ECU makers' manual and found a bit about calculated enrichment based on air density change. I believe it was an option, and it was probably an Accel unit. I remember it because they described it well enough that I thought a person might be able to find a similar calculation in $EE.

[sherlock9c1]

Thanks Steve! Yes please, if you could PM me that information, that would be much appreciated. With the supercharger, during throttle opening events, airflow increases a lot, really fast. Currently, once the PCM has learned the tune, I can stand there and just roll the throttle open with my hand and the motor would start to die like you had unplugged the fuel pump.

I was thinking of playing with the PE qualifiers as well. I will definitely go down that route for now. It's hooked to a manual transmission, if that makes any difference. I was strongly advocating going to Speed Density but supercharger cars only operate in two regions of the tables (part throttle and down the right hand side at WOT!). If it was a turbo car I would have been in 2bar SD from day one, no question.

I have disabled burst knock on the advice of someone else I trust but my rule of thumb is that the OEMs don't do things for no reason so make sure you have a darn good reason for changing what you're changing.

[sherlock9c1]

One clarification - The area where I need to make the biggest drivability improvement is during clutch engagement with a cold motor. Once the motor fully warms up, the stumble is mostly gone.

Okay so I looked in TunerPro; give me some pointers here:
PE Enable Min MAP is set to 15kPa - I'll look at the logs and see if I can leave this value as-is.
PE mode Coolant temp boundary is set to 151C, so only the cold mode tables are used.
PE enable TPS vs. RPM is set to 66% from 0-2800rpm - seems like I should drastically lower this for 0-1200rpm, and that way I can allow PE when coming off idle but not engage PE during part-throttle cruise.

Am I on the right track?

Anybody used both modes successfully on an LT1?

[sherlock9c1]

So I spent some time looking at data logs and came up with 9kPa as the MAP threshold and a TPS threshold of 3% for 0-1200rpm. My goal is to get the PCM to go into PE mode as quickly as possible when the driver presses the gas pedal, but only in the idle/clutch engagement area. I'll probably get another opportunity to tune it in a couple of weeks (are supercharged cars ALWAYS science projects?) and will report back.