GMPP 376-525hp with Whipple, GM.SEMRET torque management crap....

[chevyspence]

*** ISSUE SOLVED*** (findings posted on page 2)
Man, this e67 has been kicking my butt in efilive. How the heck do I get GM.SEMRET to stop pulling 20-30deg timing on launch!?!?! I'm not exactly new to this whole thing and seems to me that there are obviously tables and/or parameters missing or being ignored, and there is only one applicable PID in the list (GM.SEMRET) which doesn't really tell me anything. I cant be the only poor soul trying to get a customers GMPP crate engine with supercharger back to them in a reasonable time and in reasonable running condition :(

I really hope I'm just missing something here, because I REALLY don't want to hack this thing and use Min final timing or something stupid to keep timing from going way negative; a backfire blowing the supercharger through the hood is definitely not what I need...

Tune and log files attached are from a while back but short and sweet showing the problem.

23659
23660

[izaks]

I have found only 2 ways to defeat GM.SEMRET
Switching off (no) B0515 to 0532 - no protection for drivetrain - not ideal
Set b5157 to safe values, but prevent timing going to huge negative numbers

[chevyspence]

Hey thanks for the quick input. A while back I did set all of those drivetrain parameters to no (built 4l80e and ford 9" so should be safe in this case) which still did not affect the SEMRET timing pull. That alone fixed that issue for you? I am working with a GMPP LSX376 engine controller, which did have some funky hidden idle rpm limits that I had to backdate the Operating System to get to work correctly, so thinking this may be another one of the GMPP controller gremlins :(
I had the same last-resort thought on using the b5157 table... really hate to do that on a customers car but sounds like it might be my only option, will just carefully explain what to be mindful of as far as detonation if they get a tank of bad fuel.

[pinstripebob]

It looks like you still have B0517 enabled. I know it's TCS reduction by throttle, not spark, but it might be worth turning that off and seeing if that's the final nail in the coffin.

A few of the Torque Control tables are not maxed out either, you might try maxing every one of them out and doing a log to see if SEMRET stays at zero.

Are there any other engine or transmission input/output shaft torque PIDs you could log? Maybe see if there's a calculated torque value being hit that's unexpected? With a supercharger on there, it's possible the torque model the ECU is using is no longer accurate, so it's pulling timing for traction control when you wouldn't expect it to.

[chevyspence]

Ah, youre right in that particular tune file (few iterations ago) I did have throttle limit active and lower values set in the higher rpm cells trying to do some torque clipping to keep the previous injectors at a safe duty cycle but for some reason that did not work on this application either like it has in past applications... So, as of current, have the larger injectors in and b0517 has been set to "no" for a few iterations now.
I initially had all of the torque control tables set to the absolute max of 8192Nm, but was curious if maybe being right on that hard limit was causing an issue (have seen a weird glitch like that a time or two) so all limits have been set to 8000Nm.
This OS contains engine only, without any transmission parameters available, and the spark and torque categories of the PID list in the scanner are very sparse compared to an LS1B. Basics like burst knock arent even an available PID on the e67 so now that I'm seeing this timing pull issue which I've usually been able to kill off via the typical means, it has be flying blind lol.

The application is MAF-Less, which hasn't caused any issues for me in the past, but this is the first GMPP aftermarket controller I've done with forced induction. So far hadnt seen this in "factory" applications when adding FI. I'm wondering if there is something fishy going on with the GMPP aftermarket OS and im hitting a hidden launch torque /airflow surge limit, and if maybe an LS9 OS might fix it. Although I've done a lot of OS and COS shenanigans with GenIII, with how complex GenIV is and the nature of the GMPP crate engine controller/harness setup i'm leary of other potential pitfalls of an OS swap...

23661

[izaks]

What are your settings in the TCM under Torque Limits (D9401 up) ?

[chevyspence]

A search for D9401 in the calibration file came up with no results.
If by TCM you mean transmission control; The GMPP e67 controller unfortunately doesn't have any transmission parameters, it just says "not used" where the transmission operation segment would normally be. They force people to either run a manual transmission or a stand-alone transmission controller (which is what this application has; a TCI controller running a 4l80e).
I'm starting to think what GMPP did is left some sort of transmission torque control coding in there but the trans segment is just not reachable anymore.

[joecar]

The E67, E38, E92 and all the ECM's after ~2007 are engine only controllers (with the mating transmissions having their own controllers).

Please post some log files.

[chevyspence]

The tune and log file in the first post are the cleanest ones I have on hand, Frame 310 in that log is where the issue in question starts. Is there another PID I should be logging? Not sure where to look as there don't seem to be any other relevant ones to look at

[izaks]

Apologies, I assumed it had a T43 and 6Lx transmission
The t43 has a number of torque control settings

[chevyspence]

SOLVED!!!! Well, made the problem go away at least, not sure exactly "why" but hopefully this thread now can serve to help someone else out too.
- Took the VSS signal away from the engine controller, and now no more torque (spark, throttle) reduction!

Based on doing a few datalogs of "launches" at different throttle application rates, there seemed to be a strong correlation between when tire spin started and when the torque reduction kicked in. I'm guessing it is some sort of traction control predictive model (vehicle speed, rear wheel speed, increases at an "improbable" rate, tires must be spinning so pull torque). Very strange that a GMPP aftermarket engine controller would have that kind of thing in it... maybe with no serial connection to TCM or EBCM modules that report actual wheel spin, there is some default fall-back scheme in there that GMPP forgot to remove.

Thanks for the help to everyone who posted up ideas!

[Tre-Cool]

I've been doing some testing recently on an e38 controller with trying to restrict throttle/power by using the TCS Throttle Limiter by gear table & relevant switches for throttle/spark.

In short regardless of having the TCS button on or off, my car was pulling timing down to negatives.

23662

23663

Good thing is it does limit the produced torque.