Understanding GPEC2A Trans Torque Management

[Stoopalini]

I'm starting this thread to capture my thoughts on this, and of course, accept any feedback and input you guys may be kind enough to offer

So after the latest beta release, it looks like my OS may now be supported properly by VCM Suite. Prior, some tables looked weird, I was missing Barometric characterization fields, etc ... in summary, HP Tuners corrected some mappings and also added 138 new mappings to my OS.

So I took a new log to evaluate, and am noticing some behavior during shifts I'd like to better understand.


Here's the log I am evaluating: TH_09272024_Base_Shift-Time_Test_Beta-2.hpl

And here is the tune file: TH_09272024_Base_Shift-Time_Test.hpt

Here is how the Trans Tq Mgt settings are set:

Trans_Tq_Mgt.JPG

And here is how the cat overtemp protection values are set (more on this later):

Cat_OT_Protection.JPG


So looking through this log, there are some WOT events and then some aggressive throttle events where WOT is not triggered.


During this particular WOT event, I notice the pedal voltage, throttle voltage, and accel position % all peak out and stay maxed during the trans shifting:

Tq_Mgt_During_WOT.jpg

In the above screenshot, I am evaluating Accel pos %, STFT/LTFT %, FA Enrichment value (in channel listing on the left), and cat temp (also in channel listing) to determine if WOT has been triggered .... and here, it obviously has.

During the shifting at WOT, it appears timing is being pulled as the Tq Mgt strategy (as expected). The RPM drop then also reduces injector DC. It doesn't appear to me that the electronic throttle is being pulled, nor is fuel being pulled. What's interesting to me in this WOT event is that STFT B2 carried over a positive trim throughout the event. I've read the GPEC2A can do this at times, when entering WOT ... if the STFT is showing positive it can carry through WOT. Not sure if that's what happened here, or if it's an artifact of VCM Scanner showing an anomaly as this is actually the 1st time I've noticed a value in STFT during WOT.


Now here is a look at an aggressive part throttle event:

Tq_Mgt_During_non-WOT.jpg

At first I thought WOT had been triggered because the STFT and LTFT's are showing zero, and FA Enrichment has been triggered ... but Accel Pos % is only 49% so I figured something else must be triggering this other than WOT. Then I noticed Cat temp is 1,740*F, so with an FA Enrichment value of 0.0150, it aligns to the enrichment value populated in table 34273.

So WOT was not triggered, but FA Enrichment was due to cat temp, and thus fuel trims went to zero while FA Enrichment was active ... correct?

Thinking I have that understood, I looked at the pedal V, throttle V and accel pos % values. The pedal voltage stays constant right around 2v, but the throttle voltage and accel pos % are tracking RPM. I assume this is normal and not an artifact of Tq Mgt pulling the throttle, but would love to understand it better.

I realize the controller is manipulating the throttle position to meet the requested airflow (based on a torque request, right?) ... so is this modulation of the accel pos % just a result of that activity happening, or is this a result of Tq Mgt kicking in .... or is it something else entirely?


I did find some clarification on what the "A" and "B" mean with regard to the Tq. Mgmt controls in Auto Trans TqMgt ... apparently A refers to the Trans being in control of Tq Mgt and B refers to cases where the Trans is not in control. I think it was a post by Steve at HP where I read this, and he stated the description is wrong when it refers to a Dyno flag.

So the Spark master switch is on, but both A and B switches for Spark are off .. but I still see spark being pulled during shift.

FSO is enabled for all 3, but I don't see evidence of fuel being pulled.

ETC is disabled for all 3, and I don't think I see evidence of throttle being pulled either.


So am I understanding this correctly?

[james1595]

Following for id like to learn as well

[Stoopalini]

So based on something I saw in this thread , I decided to try some different Max Demanded Throttle and Phi settings.

Here are the original values:

Max-dmd-th_and_Phi_Original.JPG

And here is what I changed them to:

Max-dmd-th_and_Phi.JPG

Here is a log I took with these settings: TH_10092024_Base_Max-Thr_n_Phi_Test-2.hpl

I started this log after running for a while, so the misfire counts start at non-zero.

The throttle definitely feels snappier and more intuitive, if that makes sense. More like what I'd expect a 550hp V8 to feel like. I do believe I am getting more misfires with these changes though. Not so much to be concerning, but something for me to evaluate a bit deeper. Last time I dealt with something like this, it was due to the fueling.

So maybe changing the ETC Min Phi also requires an adjustment to fueling? In each of the areas in the tune where a misfire was logged, the fuel trims seem to be negative. Not much, like 5-7% or so, but always negative.

Here's an interesting point in the log ... notice the pedal voltage is constant here, but the throttle voltage and Accel Pos % flattened out, or dipped ever so slightly, before ramping up.

Throttle_Dip.JPG

This is the only place in the log where this happened. Everywhere else I "tipped in the throttle" looks like the changes I made have the throttle voltage and accel pos % following more closely with the pedal voltage


One interesting observation is it appears during non-WOT spirited acceleration, the throttle voltage and accel pos % are no longer dipping during the trans shifting as much as they were before this change

Non-WOT-Shifts.JPG

I'm still going to keep digging in and testing, but found these results interesting.


I am experiencing some weirdness in the 2-3 shift, and on rare occasion in the 1-2 shift ... where it feels like torque is being pulled significantly for too long. All other shifts are quick, snappy, and seamless, but that 2-3 really dips the front end down during the shift followed by what feels like it slipping into 3rd. You can see it in the log by looking at the torque plot. I have the Shift TqRed Time field set to 0.02s, which is the OEM setting. Whipple had changed this to 1.0s, which was worse. Maybe I should try 0.0s? Or maybe I need to look at the trans settings instead ... more to learn!

[LilSick]

Yes

Tq mgmt is where you need to be adj

[HaasExp]

Not entirely sure which actual PIDs (there are many) you are logging for the throttle and the pedal voltages, but you should be logging the channels that max out at approx 3.84V for both as these are the ones that directly correspond to the TB and pedal voltages in the various tables and parameters in the tune.

I believe the ETC min phi has to do with the TB airflow modeling (to convert the requested torque to a desired airflow as thus TB opening). It may be making the TB more responsive which could be indirectly affecting the fueling due to the quicker throttle tip in.

[LilSick]

Not entirely sure which actual PIDs (there are many) you are logging for the throttle and the pedal voltages, but you should be logging the channels that max out at approx 3.84V for both as these are the ones that directly correspond to the TB and pedal voltages in the various tables and parameters in the tune.

I believe the ETC min phi has to do with the TB airflow modeling (to convert the requested torque to a desired airflow as thus TB opening). It may be making the TB more responsive which could be indirectly affecting the fueling due to the quicker throttle tip in.

throttle desired position vs relative pedal. although i never see the whole 3.84 being desired

yeah i agree - therefore stop changing the phi

[Stoopalini]

Thanks guys, I appreciate the input.

So I made some adjustments to the torque settings:

Trans_Tq_Mgt-adjusted.JPG

and also increased the TCM's Max Avail Torque table:

Trans_Max_Trq.JPG

and gave it a test. Here's the log from the above changes: TH_10112024_Base_Trans-Torque-Test.hpl

The torque dips during PT shifting are now gone, so that's good news! I was still seeing misfires counts during tip in though and the 2-3 shift still is acting strange. So I put Phi back to stock values, and adjusted the Max Demanded Throttle some more.

Max-dmd-th_and_Phi-Adjusted.JPG

And tested again ... here's the log from that test: TH_10112024_Base_Trans-Torque-Test2.hpl

Misfires are now back to the normal 1 or 2 every once in a while, so it appears it was the Min Phi table which was causing it.

The 2-3 shift is still weird, but now only seems to be doing it when accelerating moderate to aggressive. With casual driving, the 2-3 shift is now good as it is not causing you to lean forward in your seat during the shift like it was before.

Obviously there's still something going on with the 2 -3 shift which needs addressing. Not sure it is the trans tune though? Is there a way to adjust the timing pull which happens during the shift? I've looked through trq mgt in both the engine ECU and the TCM, but don't see any tables which allows control over how much timing is pulled, or how long it lasts. I'm sure I'm missing something though ...

Or maybe this 2-3 shifting issue isn't torque related at all? Not sure ... Even before the Whipple, 2-3 was always sloppy. Dodgeboy tuned it for me back then, and he tried to resolve it, but it's always remained .... so I just let it be since I knew I was adding a Whipple in a couple of months. He has sinced revised the tune, which is great except for the troublesome 2-3.

Now the whipple is installed, I'd really like to get this ironed out. The best way to describe it is during the 2-3 shift, with Normal clutch shift type active, the vehicle pulls back quite a bit during the shift and then slips into 3rd pulling the vehicle forward. Sometime it really dips the front end down, kind of like a manual transmission with someone shifting slowly.

3-4, 4-5, 5-6, 6-7, 7-8 are all really quick and transparent, at both light cruising and WOT (as well as anything in between), but 2-3 just gets more disruptive the more throttle I'm giving it. In sport or track mode, it's even more pronounced.

Maybe I need to start looking at the Forced and Main Tq Intervention tables?

Trans_Forced_Tq-Intv.JPG

Trans_Main_Tq-Intv.JPG

[Stoopalini]

Oh, and related to pedal and throttle voltages ... I now have throttle desired position in my channel listing. It does get to 3.80v quite frequently, even when pedal voltage is not maxed out at 3.8. This is actually kind of nice considering I have a Smooth Boost electronic bypass controller which is feeding off the APPS connector at the pedal. So even though the desired throttle goes to 3.8v and accel position goes to 76% (indicating WOT), the pedal voltage can be in the 2's, which means I'm basically controlling the electronic bypass position with the pedal at this stage, which in turn is driving the PRatio.

It's a very transparent feeling, to where the accelerator pedal is responsive and predictable. No light switch effect, no dead pedal feeling at any point in the travel, etc ...

Here's a good example:

Pedal-Throttle_Volts.JPG

So Throttle Desired Position is 3.8v, Throttle V is 4.29, and Accel Pos % is 76.1%. All of these are WOT values, but the pedal voltage is only 2.49v, and notice the PRatio is only 1.27.

As I increase the pedal, the bypass closes more, causing PRatio to go up, and I get more torque ... all while throttle voltages stay the same.

[LilSick]

timing IS torque

Decrease the torque reduction and it will pull less timing

Log it n see



I doubt that’s all it is tho…

My 2-3 needed more help than any other upshift

[Stoopalini]

Yeah, I realize timing = torque in this case. What I was looking for was the mapping which translates between the two.

I see the TCM tables direct a Nm torque reduction based on "Engine Torque by TOS delta speed" for Forced Tq Intv, and "Engine Torque by TOS" for Main Intv.

2-3_Tq_Intv.JPG

But something is translating these Nm torque numbers in the table into a timing pull amount, and I was looking to see if I could find that somewhere.

I've a feeling trying to get the 2-3 shift transparent could take a LOT of trial and error, especially if the root cause is fill time, pressures, etc ... related and not just torque mgt related.

I thought about trying to copy the 3-4 shift variables over to the 2-3 table, since it is always very smooth and transparent ... but something tells me it won't be that easy and probably isn't a good idea.

[james1595]

There is a torque reduction spark table, its the minumum spark the engine will pull during torque reduction, that is the table you are looking for. As for the tables, i wouldnt copy and paste shift pressures, but the torque intervention tables is fine. You may just need to increase pressure and pull more timing with the faster shifts. Then when its good, you can try reducing it little by little

[LilSick]

Yeah, I realize timing = torque in this case. What I was looking for was the mapping which translates between the two.

I see the TCM tables direct a Nm torque reduction based on "Engine Torque by TOS delta speed" for Forced Tq Intv, and "Engine Torque by TOS" for Main Intv.

2-3_Tq_Intv.JPG

But something is translating these Nm torque numbers in the table into a timing pull amount, and I was looking to see if I could find that somewhere.

I've a feeling trying to get the 2-3 shift transparent could take a LOT of trial and error, especially if the root cause is fill time, pressures, etc ... related and not just torque mgt related.

I thought about trying to copy the 3-4 shift variables over to the 2-3 table, since it is always very smooth and transparent ... but something tells me it won't be that easy and probably isn't a good idea.

Yep

Nope, that is not the solution.

It’s just trial and error

[LilSick]

There is a torque reduction spark table, its the minumum spark the engine will pull during torque reduction, that is the table you are looking for.

44681

it is the floor

but it is the floor for EVERY gear

therefore not what you want to change to get where you want to go

to tune the amount of timing pulled 2-3 first isolate by zeroing 47142

then log to determine which of the 4 tables is being used

norm perf sport or max

start decreasing the values in that table/those tables (just the lower most and far right cells) and you will see it reflected in the log during wot 2-3 shifts

where you will see it depends on those switches we were talking about earlier

[Stoopalini]

There is a torque reduction spark table, its the minumum spark the engine will pull during torque reduction, that is the table you are looking for. As for the tables, i wouldnt copy and paste shift pressures, but the torque intervention tables is fine. You may just need to increase pressure and pull more timing with the faster shifts. Then when its good, you can try reducing it little by little

I believe that table is minimum spark allowed for any torque reduction scenario though, not just trans shifting. Like for knock retard as well. So I'm not sure I'd want to use that to try and dial in the reductions during shift.

to tune the amount of timing pulled 2-3 first isolate by zeroing 47142

then log to determine which of the 4 tables is being used

norm perf sport or max

start decreasing the values in that table/those tables (just the lower most and far right cells) and you will see it reflected in the log during wot 2-3 shifts

where you will see it depends on those switches we were talking about earlier

I'm actually logging "Clutch Shift Type", which tells me which of the 4 tables is being referenced, as well as "Trans Shift Torque Mgt State", which tells me when 47142 (Normal Forced Torque Intervention for 2-3) is being used vs when 47107 (Normal Main Torque Intervention for 1-2) is being used.

Clutch Shift Type stays on Normal unless I engage a different drive mode. Track mode seems to use Perf clutch shift type. Not sure what sport mode uses, I'm guessing sport? I'll have to test it and confirm though.

Looking at the log:

• Time marker 44.205: Spark is 8.5 degrees and Tq Mgt State changes to "US Forced Torque Intervention". This is when it references 47107, since Clutch Shift Type = Normal
• Time marker 44.395: Actual spark reduction begins
• Time marker 44.515: Spark is now 4.5 degrees and Tq Mgt State changes to "US Tq Intv". Here is where is starts referencing 47142, again since Clutch Shift Type = Normal
• Time marker 44.568: Spark is now all the way down to the minimum (3.0) as defined in 44681
• Time marker 44.825: Spark is still 3.0 and Tq Mgt State changes to "Post Tq Intv"
• Time marker 44.875: Spark starts to come back up
• Time marker 44.995: Timing is now back up to 7 degrees with no more reduction
• Time marker 45.025: Trans Shift Torque Mgmt State goes back to "No Intervention"

So the entire torque reduction process seems to take 0.82 seconds. For comparison, the 3-4 shift right after takes 0.71 seconds to go through the torque reduction states. So I don't think what I am feeling is a shift time thing, but is more related to the way it's shifting.

[LilSick]

so much of your understanding and logic is off, i am not gonna sit here and correct it all

but...

sport mode should use the perf tables and track should use the sport tables

the aggression counter might not have been high enough yet so while yeah it was in track mode in the bcm, it was still accessing the perf tables in the tcm

[Stoopalini]

Yeah, I mixed up tables 44681 and 44679. Where 44681 is minimum spark floor for torque reduction while 44679 is overall minimum spark ... my mistake.

I hear you on the trans mode, but it's very apparent in the logs I have. Hitting the "Track" button causes the "Clutch Shift Type" to change from Normal to Perf, and it stayed on Perf the whole time it was engaged, even during throttle stabs and WOT events. I actually saw something similar before the Whipple install, where Track mode was using shift schedules 2 and 7 instead of 4 and 9 (or 3 and 8). It did engage 9 very rarely with track engaged, but only at WOT and only a few times.


Another oddity is that “Limp/Protect Mode” continually flips on and off for some reason. No impact to vehicle behavior though. I wondered if this may have been another mismapping by HPT?

[LilSick]

dodge's logic does not always make sense

[HaasExp]

throttle desired position vs relative pedal. although i never see the whole 3.84 being desired

Relative pedal and relative throttle voltages are the key ones to log as they show the actual pedal and throttle body voltages that line up with the voltage parameters in the tune file.
Desired position is what you want but may not necessarily get, which can then be misleading.

[HaasExp]

Even before the Whipple, 2-3 was always sloppy. Dodgeboy tuned it for me back then, and he tried to resolve it, but it's always remained .... so I just let it be since I knew I was adding a Whipple in a couple of months. He has sinced revised the tune, which is great except for the troublesome 2-3.

What did Dodgeboy say about the issue?

[james1595]

Sounds like my issue, excpet mines the 1-2 shift. No matter what you do, it never seems 100%. Most likely a bad clutch from the factory. Ive just accepted it and try to make it better, but am just waiting for it to blow to replace it.