Pedal Percent Power Request Modes

[Stoopalini]

On a vehicle with SRT drive modes, do these modes (1-7) relate to the different drive modes (Auto, Track, Sport, Snow, Tow, Eco)? I looked to se if there's a way to log which mode is active when, but don't see any channels in VCM Scanner which would correlate to the Power % Request Mode tables.

It seems a lot of folks tune all 7 tables to the same values, but when looking at OEM calibrations on vehicles with the SRT drive modes, I can see they are setup differently.

As I continue to refine the tune on my Supercharged Durango (which does have the SRT drive modes), I'd like to retain the different pedal feel for the various drive modes and I think this would be accomplished through these tables .... but am not sure which table relates to which drive mode.

Does anyone know how and when modes 1-7 are referenced?

[HaasExp]

How about you just cap them each off at different %'s and test which ones activate under which modes for your particular platform?

[LilSick]

for your application i recommend leaving them all at the oem settings

you have a supercharger and a smooth boost computer hooked up to the pedal

the car has modes...

sport is gonna be more sensitive than default

what else could you possibly want?

(what is it not doing that you want it to, or what is it doing that you do not want it to do???)

[Stoopalini]

How about you just cap them each off at different %'s and test which ones activate under which modes for your particular platform?

Yeah, I could definitely work up a set of data to test with, I just thought someone may already know how the 7 modes are utilized considering these tables seem to be very popular for dialing in throttle.

I actually did a bunch of testing today, for other purposes though ... and made good progress on refining the tune. It just gets extremely time consuming to not only test different configuration on known tables to see how my engine will react, but then to also run tests to see how the software is configured to use the various tables ... definitely can eat up several hours.

[Stoopalini]

for your application i recommend leaving them all at the oem settings

you have a supercharger and a smooth boost computer hooked up to the pedal

the car has modes...

sport is gonna be more sensitive than default

what else could you possibly want?

(what is it not doing that you want it to, or what is it doing that you do not want it to do???)

I want to understand how they're used so I can fine tune it to our liking. The wife likes a very sensitive pedal, but drives around in Auto mode most of the time. I've been able to get the throttle pretty well dialed in with the Throttle Body Model Airflow table (along with the large and small range tables), but it was still a bit jerky off the line. I was tweaking the airflow model to try and compensate, but then realized it was probably the power request which was causing it ... So I started looking at these tables to understand them better.

I ended up grabbing the entire Torque Management - Pedal Characteristics section of tables out of a 2021 Durango Hellcat and tried that. The pedal was smooth, but was kind of dead off the line. Most likely because the Power % Request and Expected Torque tables are setup for an engine with a ton more power than mine has. So I grabbed those two tables from my previous tune, and pulled them back in with a few tweaks down in the off-idle region, but left the Hellcat's Power % Request Mode and Expected Pedal Mode tables.

This is actually working really well. No more sudden jerk off the line, and the throttle pedal has a high sensitivity while still being easily controlled and predictable.

I've also recently overhauled the cam timing, did another round of ANN trainer tuning (with 14 neural nodes now), and have been testing out different configs for Tip In spark and IAT Spark thresholds.

I've got it where PT sudden throttle increases are snappy, but without any KR or misfires ... which took a few trials to get right. So part throttle is very smooth now with what feels like effortless power on tap. No lag on rolling tip in anymore. I still have a bit of lag from a standing start WOT stab, but no misfire or KR from it anymore. It seems the Desired Throttle Voltage takes almost a full second to reach 3.8v from the time the pedal voltage hit's 3.8v.

So I guess, what am I looking for it to do? Be the best it can be LOL. I know, subjective ... but the more I learn, the more I can fine tune it. Remember, my goal is to have it drive as smooth and refined as a factory Hellcat drives


Latest tune is here: TH_11262024_Base.hpt
And a log with that tune loaded: TH_11232024_Base_CAM-n-SPARK_After-Training_Ver-6.hpl


5:24 time marker has an example of a standing start. This was a 0-60 test, and it was my best thus far (4.97 seconds). Still seeing a tip in lean spike and the tip in still felt like it lagged ... I think the data shows why (ie: taking a full second to reach desired WOT)

The pedal voltage reached 3.8v at 5:24.609 but the Desired Throttle Voltage did not reach 3.8v until 5:25:600. The desired throttle voltage actually dipped down a bit into the tip in ... not sure why though.

[HaasExp]

The pedal voltage reached 3.8v at 5:24.609 but the Desired Throttle Voltage did not reach 3.8v until 5:25:600. The desired throttle voltage actually dipped down a bit into the tip in ... not sure why though.

I would avoid looking at the "desired throttle voltage" because that is not what you are actually getting at the TB. I'm sure this was brought up in one of your other posts.
Log the "relative throttle" voltage as that the actual voltage at the TPS on the TB.
You logged the "throttle position sensor" voltage, which has some weird scaling but will be proportional to the relative throttle voltage, and that parameter indicates that the TB is virtually fully open at 5:24.787. So that's a whole 0.178sec actual delay. Hardly an issue.
Furthermore, your polling rate seems quite slow on your desired throttle parameter anyway. This will accentuate any perceived delays.

[HaasExp]

I've also recently overhauled the cam timing, did another round of ANN trainer tuning (with 14 neural nodes now), and have been testing out different configs for Tip In spark and IAT Spark thresholds.

I assume that you have noticed it leaning out to near 0.9 lambda up top at WOT?

[LilSick]

I assume that you have noticed it leaning out to near 0.9 lambda up top at WOT?

that is a BS PID that is there for the cars that actually come with a wide band

<channel ParameterID="36" />

i would not pay any attention to any data populated by that channel in a 5.7 car or truck

[LilSick]

***

[HaasExp]

that is a BS PID that is there for the cars that actually come with a wide band

<channel ParameterID="36" />

i would not pay any attention to any data populated by that channel in a 5.7 car or truck

I'm looking at the "WB EQ Ratio 1" parameter which will be coming from an external wideband. Definitely not a BS parameter.

[Stoopalini]

I have noticed the reported AFR leans out as the RPMs climb, which was one reason I added the fuel pressure sending unit.

Here's an example right at the top of 2nd, ~5600 RPM, fuel pressure of 57psi (down from 62psi at idle), injector duty cycle is only 54% ... but Lambda is reaching 0.87 or so when being commanded at 0.77.

1st_to_2nd_WOT.JPG

I'm not quite sure why it's happening though. Should I be adjusting the PE table and adding more fuel? Lambda seems to float in the mid 80's during WOT. I was attempting to target 11.7:1 AFR, so 0.80 lambda, but I have it set for 0.77 since it seems to hit an actual target a bit higher than commanded. At least according to my AEM x-series wideband.

[Stoopalini]

I would avoid looking at the "desired throttle voltage" because that is not what you are actually getting at the TB. I'm sure this was brought up in one of your other posts.
Log the "relative throttle" voltage as that the actual voltage at the TPS on the TB.
You logged the "throttle position sensor" voltage, which has some weird scaling but will be proportional to the relative throttle voltage, and that parameter indicates that the TB is virtually fully open at 5:24.787. So that's a whole 0.178sec actual delay. Hardly an issue.
Furthermore, your polling rate seems quite slow on your desired throttle parameter anyway. This will accentuate any perceived delays.

Thanks, makes sense ... I actually am logging relative throttle (as well as TPS, commanded throttle, etc ..) and I see relative throttle hitting WOT (76.1%) at 5:25.366 so that's 0.757 seconds after the pedal hit 3.80v at 5:24.609. I've no idea if that's what is causing the lag on take off though. Maybe it's just the torque converter and I should try power braking it?

[Stoopalini]

***

Yeah, DFCO is enabled, but should be disabled above 60kPa, right?

I left DFCO enabled because I thought this allows decent fuel mileage for a daily driver. Is this influencing the AFR as the RPMs climb though, and I should be disabling it ... or adjusting the settings for it?

[LilSick]

Scanner lags

With DFCO enabled sensor see’s hella lean soon as you lift

[HaasExp]

I have noticed the reported AFR leans out as the RPMs climb, which was one reason I added the fuel pressure sending unit.

Here's an example right at the top of 2nd, ~5600 RPM, fuel pressure of 57psi (down from 62psi at idle), injector duty cycle is only 54% ... but Lambda is reaching 0.87 or so when being commanded at 0.77.

1st_to_2nd_WOT.JPG

I'm not quite sure why it's happening though. Should I be adjusting the PE table and adding more fuel? Lambda seems to float in the mid 80's during WOT. I was attempting to target 11.7:1 AFR, so 0.80 lambda, but I have it set for 0.77 since it seems to hit an actual target a bit higher than commanded. At least according to my AEM x-series wideband.

The PE table is the FA adder for your commanded (target) lambda. If you are commanding 0.77 and you are actually getting 0.87 then you need to tune the ANN so your actual lambda matches your commanded lambda.
Given that your IDC is only 54%, you still have heaps of headway remaining. This also implies that your fuel pressure is still satisfactory up top even without reading the actual pressure.

[LilSick]

Yeah, DFCO is enabled, but 1) should be disabled above 60kPa, right?

2) I left DFCO enabled because I thought this allows decent fuel mileage for a daily driver.

3) Is this influencing the AFR as the RPMs climb though?

4) and I should be disabling it

5) ... or adjusting the settings for it?

1) yes, it is only enabled when the engine is pulling vacuum

2) dunno how much difference it makes with efficiency but i bet it reduces emissions when it cuts

3) no

4) no

5) try this:

[HaasExp]

Thanks, makes sense ... I actually am logging relative throttle (as well as TPS, commanded throttle, etc ..) and I see relative throttle hitting WOT (76.1%) at 5:25.366 so that's 0.757 seconds after the pedal hit 3.80v at 5:24.609. I've no idea if that's what is causing the lag on take off though. Maybe it's just the torque converter and I should try power braking it?

If you look at the data more closely, you will see that the actual polling rate for the relative throttle % is quite slow (something like 2-4Hz). This slow data rate will mislead you into thinking that you have lag.
Log the relative throttle voltage and set the polling rate to 10ms otherwise you will just end up seeing issues that don't really exist.

The actual lag on takeoff looks to have something to do with the descent lean spike you get. Almost looks like it causes a stumble.

[Stoopalini]

Thanks guys, I appreciate the input and assistance.

I had previously changed all the channels in my scanner config to be the fastest polling, so I double checked this particular channel and it turns out Relative Throttle isn't actually in my list. It must be one of those channels it captures by default? So I explicitly added it and set the polling to 10ms.

I can try the DFCO adjustments as well, thanks Don.


For ANN training, I would think I set up an EQ error table with filtering to ensure I'm only capturing EQ error when DFCO isn't enabled.
Does the below look right? The two variables are "When WB-1 EQ is below 1.4 lambda and when relative pedal is above 0.1v.

EQ_Error_Graph_Setup.JPG

I set the axis based on what I see in the NN trainer:

NN_Trainer_EX-117.JPG

What I'm unsure about is how to make adjustments in the virtual VE and take into consideration the cam position. Do I need to set the VVT tables to be each of the NN cam positions, one at a time, and take 5 logs? Then use that data to train by adjusting the virtual VE for each of the exhaust positions (and selecting the appropriate intake position based on 228.5 - exhaust)?

I haven't done this because I'm not sure if having the VVT tables set across the board to each of the 5 values (85, 95.75, 106.5, 117.25, and 128) would cause an issue when logging PT and WOT runs. I have my cam position pretty much centered around 114, based on some input from Hemituna saying for supercharged applications, running the cam around the "straight up" position, and then retarding, advancing a bit for key areas, works well. After I made this adjustment, it does run well like this. But it means the cam doesn't move much at all. It does still park at 125 at idle sometimes, but for PT and WOT, it's staying between 112 - 116 ... But of course, the NN trainer has fixed values for the training.

[Stoopalini]

Or ... is it viable to setup some filtering on the existing data, and make adjustments from this?

For Cam position 117, I basically filtered to show data when the cam is +/- 5 degrees from 117, and also when actual EQ is below 1.4 and relative pedal voltage is above 0.1v

EQ_Error_CAM_117_Setup.JPG

The result seems to track what I'm seeing in the data, excepting the 0.21 PRatio row ... but the upper RPM columns seem to be about the right adjustment I think.

EQ_Error_CAM_117_Result.JPG

[LilSick]

looks like you are telling the trainer the cam has a lobe separation angle of only 114 degrees...

we both know it does not

wouldn't that throw all of it's calculations off???