Is there a reason to optimize true vs actual driver demand torque after mods?

[cmitchell17]

I know some people leave driver demand stock and they somehow make everything work when you do airflow mods to increase power.

I am trying to start over from scratch on a stock tune, and use stock torque tables along with as many stock parameters I can, and try to make a whole one from the ground up. I have increased driver demand tables around the 85-100% range around 25%. This apparently isn't enough since my "Throttle Desired Area" and "Throttle Desired Area Final" (I am assuming final is basically the same thing just with limiters already calculated in so it gives you the actual value going to the throttle blade?) is only going to 60.7% and 60.1%. So therefore my throttle blade is only getting to about 70% and timing is really choppy and rough (even though the immediate torque source is none), so basically it feels like my car has about half of the power as it does on my known working tune (that desires over 300% of desired throttle area).

So if anyone can help on these questions:

It looks like I will need to increase driver demand at high accelerator position again to try to get throttle desired area up and therefore actual throttle blade up, is there a point to optimizing it to keep going up and up right until I get my throttle to open 100% of the time?

Is there a analytical approach (histogram? use difference in stock and modified VVE or MAF?) to remodeling driver demand when you add mods? Or is doing it like this and guessing and checking the only way? Or does it not even matter since everyone just cranks up driver demand to some really high value and doesn't care if desired area is over 100%?

And these are the big mystery's of the GEN V:

My assumption is that PCM looks at driver demand and current accelerator position, then finds desired torque, uses some unknown table (we don't have access to) to convert a desired torque into desired throttle area, then uses another unknown table (we don't have access to) to convert desired area into throttle blade opening percentage?

During these calculations described above this is an "axel" torque source and also during this the PCM will never use any immediate torque control only predicted and only the throttle blade?

OR

Does the virtual torque tables step in and offer this translation from desired torque in driver demand to some airmass then some unknown table (we don't have access to) converts this airmass into a throttle blade percentage? (I know this last table has to exist since I know the relationship between throttle blade percentage and airmass is not linear)


There is just so many "Chicken or the egg comes first" type of scenarios going on here, so when does the airflow/airmass (VVE and MAF curve) come into play here, or is it independent?

[Higgs Boson]

start with a completely bone stock tune
put it in MAF only and tune the MAF exactly perfect
if you have a big cam making less torque at idle and off idle then go to your VTT and multiply the idle area by .7 and smooth one row/column out so that it wont pull lots of timing at idle and buck off idle while spark and throttle fight each other.

after that, things like running out of fuel, or adjusting individual limiters for specific scenarios will vary per application.

I tuned on the VVE for literally years and the cam'd whipple 1000 rwhp E85 DI/PI Z06 just never got exactly as I wanted until I just went MAF only and left almost everything else alone. stock DD, VVT except for idle, stock VVE, etc and the damn thing ran like stock until you hit it then it was 100% throttle and commanded spark , etc.

[TriPinTaZ]

If you're throttle won't open all the way then you likely need to increase your VT in those areas. Particularly if your cam phaser is locked. But in reality, Higgs has the best advice here. So many people thought in the beginning that so many parameters have to change. But the truth is, only change what is needed.


Driver Demand Torque Requested needs to be higher than Axle Torque Commanded and Actual Axle Torque. That is the only limiter that I am aware of when it comes to Driver Demand. Also, if you increase your VT areas at WOT you may have to increase your Peak Torque table as well because the VT increase will directly affect Max Engine Torque.




I would do as Higgs suggests and get that tune down. Then you can tinker with VVE later but will always have a base tune where the car drives good. Nothing is worse than modifying your car and not being able to enjoy it because the tune isn't working well.

[cmitchell17]

Thanks, I completely agree with what y'all are saying. Since I changed back to a stock VTT besides the small amount taken out at idle, I am getting much better and smoother torque transitions at shifts and part throttle shifts.

What confuses me is I tried last night by taking the driver demanded table (that has like crazy high torque amounts at WOT (>16,000 ft-lbs)) and pasting it into my new "stock based" tune and the same thing is happening as before. I copied over the peak and max torque tables too and the same thing. I think I'm about to give up again on trying to make my own tune and just run with what I know works (even though I don't like the idea of 30+ degrees of timing at WOT the tune has).

I don't know what else I can do to get MAF or VVE any better than I have. Trims are well under 5% and most of the time under 2.5%, sometimes at idle one bank will be at 0.3%. Then however, randomly I will get like 20% for a few minuets and then it will go back to almost zero, I think this is just the instability of combustion with a high overlap cam though.

One of my biggest wishes is if we could get a definite explanation of all the torque PIDs, whoever named them at GM (or either maybe they were mapped by someone else) did a pretty bad job haha, and I'm not sure why it was necessary to create so many of them?

[TriPinTaZ]

If you put your VT back to stock or near stock you don't need as much of an increase to your DD tables. In fact you only need to increase your Driver Demand above stock if Driver Demand Torque Requested PID is LOWER than Axle Torque Commanded or Actual Axle Torque PID's. You should be charting these PID's in your logs. Being this a Torque based system is why GM used so many. You should also be charting the Engine Torque PID's so you can look and see if you're hitting any limiters there. Only change what you need to based on the data you get from the logs.

If you set your MAF only mode your VVE table will not matter and then you can figure out what is causing the drivability issues. This will automatically rule out VVE so that once you fix the driveability and it runs great on MAF only mode, then you can spend time messing with VVE.

[castorjames]

"Driver Demand Torque Requested needs to be higher than Axle Torque Commanded and Actual Axle Torque."

"you only need to increase your Driver Demand above stock if Driver Demand Torque Requested PID is HIGHER than Axle Torque Commanded or Actual Axle Torque PID's"

im a little confused here, Driver demand torque needs to be higher then axle tq and actual axle tq or needs to be lower?

[TriPinTaZ]

"Driver Demand Torque Requested needs to be higher than Axle Torque Commanded and Actual Axle Torque."

"you only need to increase your Driver Demand above stock if Driver Demand Torque Requested PID is HIGHER than Axle Torque Commanded or Actual Axle Torque PID's"

im a little confused here, Driver demand torque needs to be higher then axle tq and actual axle tq or needs to be lower?

Oops I mean to type LOWER than. I corrected my post.

[cmitchell17]

So this is going back to the known good tune that "works" but actually it dosen't really "work" since I still have random timing pulls like this:

Capture.JPG

So how can we explain this?
Now when you say "driver demand" I am assuming "Driver Final Axle Torque Req" is what you mean in my GEN V case? This was kind of the point I was trying to make is we have such a mess of PID's and the only way for us to find out what's going on is guess and check the millions of combinations to try to figure out what all these PIDs mean.

Anyway in this case we can clearly see that the driver demand table I have in my tune is way higher than it needs to be and you can see all my "predicted" torque commands are all way higher than any "actual" PIDs. We can also see that my "Traction Control Desired Torque" PID is 0% I have always taken this as an indication that traction control is not active at the moment? Is this a good way of thinking?

Now when you look down a "Torque Mgt Advance" in the timing PID's you can see that its pulling 10 degrees of timing, this happens even time at this same point in 2nd gear for absolutely no reason. From all the PID's I have logged I see about 3 different PID's that contradict themselves and you could infer from that, that I am at some torque limit and its pulling timing?

Log 102 B Tune with Stock VT table.hplB Tune 2 (my os)(test).hpt

[Redline MS]

All engine torque values discussed in this procedure are engine brake torques (torque at the crankshaft) unless otherwise noted. Engine coordinated torque control (CTC) is conducted by first acquiring an arbitrated torque request from the driver or some other torque intervention (e.g. traction control, etc.). This torque request is then inserted into an inverse APC based torque model in order to calculate a desired APC. This desired APC is then inserted into a desired MAF calculation, which determines the desired MAF through the throttle body. The original torque request from the controller is also inserted into an inverse MAP based torque model to calculate a desired MAP. Desired MAP and Baro are used to calculate the desired pressure difference across the throttle blade for the compressible flow calculation. The pressure difference, air density, and desired MAF are then used in the desired area calculation using compressible flow. The compressible flow equation is used to calculate a desired throttle area. This desired throttle area is then given to the engine controller, which in turn opens the throttle body to achieve the desired area. Engine torque is calculated using the measured APC and the APC based torque equation.) The torque estimate is compared to the original torque request from the controller. If they both are equal the system is functioning correctly and the torque request will not be adjusted. If they are not equal, a torque error is calculated and closed loop gains, steady state flags, and filters are used to correct the torque error while controlling response time, overshoot, oscillation, etc. The desired APC input into the cam phaser schedule is also calculated during the Torque Control process. The desired APC is calculated from the inverse APC based torque model and then filtered. When an automatic transmission upshift is commanded, the phaser feed forward algorithm can be used to calculate the post-shift engine speed and desired APC for the phaser schedule lookup to have the phaser move to the required position slightly ahead of time, to allow the torque control system to accurately achieve the torque request as the shift completes. The filtered desired APC value and engine RPM are inputs to the cam phaser schedule. The cam phase position(s) are changed based on the cam phaser schedule and the actual phaser position(s) are then measured. The measured cam phaser position(s) are filtered and the filtered cam phaser position(s) arethen inserted back into the inverse MAP based torque model and the inverse APC based torque models. One of the main advantages of this particular type of Torque Control system is the torque control is independent of the cam phaser and spark calibrations. This allows parallel spark, phaser, and torque model development in that you can change the calibrations for cam phaser, spark, or both and they will not impact torque control within their control limits. This is just a skim over the process the torque based control system uses. Of course there are many other factor and trim tables that effect final calculations. Its also critical to remember that there is no dyno built into the car...all torque is derived from airflow. If airflow is off it's all garbage down hill. As Higgs has said a million times..... get airflow re-profiled properly from the mods done to the air pump.

[TriPinTaZ]

Try this and see if it helps with Torque Management Advance pulling out the timing at WOT


B Tune 2 (my os)(test)_taz01.hpt


Also, your Driver Demand tables are way higher than they need to be. 20% over stock should be sufficient at 85-100% Pedal Position.

[cmitchell17]

Thanks for the detailed explanation, it looks like it answers some questions I have had about the torque control logic for a long time, I have been busy lately and its been raining nonstop for almost 4 days now, so hopefully I will get a chance to try the tune changes soon and I can look more into it.

[cmitchell17]

Try this and see if it helps with Torque Management Advance pulling out the timing at WOT


B Tune 2 (my os)(test)_taz01.hpt


Also, your Driver Demand tables are way higher than they need to be. 20% over stock should be sufficient at 85-100% Pedal Position.

I tried opening your file but I can't on either the beta or regular vcm editor. It says its a new or unsupported version and it can't open.

[TriPinTaZ]

I tried opening your file but I can't on either the beta or regular vcm editor. It says its a new or unsupported version and it can't open.

Are you using the latest Beta? You might need to update your Beta version.

[cmitchell17]

Are you using the latest Beta? You might need to update your Beta version.

I tried it with 4.5.14445 Beta (which HP Tuners says is supposedly the latest version)

and I tried it with regular version 4.4.4 (which HP Tuners says is supposedly the latest regular release)

So not sure what's going on with it, I've never seen this before and i thought they usually maintained compatibility with earlier versions.

[HeavyChevy305]

4.6.3 is already out

[cmitchell17]

I am assuming you have to manually download 4.6.3? I am checking through the about menu in the built in update checker, I will manually install and try. Thanks

[cmitchell17]

So its working for me know I went to the website and downloaded the latest beta from the link. So lesson learned here is that don't trust the check for updates thing.

[cmitchell17]

So TriPinTaZ looks like you fixed it, it still has a little torque reduction kick in but its a lot better than it was:
Capture.JPG

It still kicks in a little at around 6200rpms:
Capture.JPG

I need to get rid of some PID's because I think I am loosing sampling speed on some of the torque PID's since they don't seem to update at every sample frame, and also HP Tuners scanner gives you no indication on your allocated bandwidth either, so even though it says the pid's are checked to sample at the fastest rate available and shows the check mark on it, it doesn't appear that that is the case. It looks to me like the highlighted ones I can get rid of? According to what TriPinTaZ said that he was only aware of one limiter and I am assuming it would be when "Driver Final Axle Torque Req" shown at 5,144 lb-ft is less than either "Predicted Axle Torque Cmd" (we don't care about "Immediate Axle Torque Cmd" since it appears immediate torque methods have no Authority (even though the torque is being reduced with spark and not closing the throttle blade?)) or "Actual Axle Torque". When "Driver Final Torque Req" (which I am assuming is the value from driver demand once its gone through filtering (torque rate limits) and the closed loop torque error control as Redline mentioned) is less than these 2 you will have some form of torque reduction.

But then again my form of torque reduction is spark limiting not throttle limiting, either way I see no problems, which I would like to go back and look at my log before TriPinTaZ adjusted the VT and see if "Driver Final Axle Torque Req" was lower in that situation.

Also in the first picture I can't tell if the torque reduction early on before it downshifted to 2nd gear was a trans request? (there is a clear spot where its a definite trans request) or if it was a traction control request? or if it was an unknown request that is the same origin that I am fighting here? My strategy for monitoring traction control torque request is look at "TCS Desired Engine Torque" anytime it is not a constant -628 ft-lbs, I feel like it is trying to reduce torque? Anyone know how TC works?

[cmitchell17]

Also, I put my DD tables back to stock and then multiplied the entire 100-95% row by 1.25 then interpolated from 95% to 75%. I logged it and saw some throttle closure around 6000rpms+, so I went back to stock and multiplied the same way except multiplied by 1.3 and logged again and I no longer see throttle closure. Now is there any reason to vary the increase based on vehicle speed? It makes sense that if I have a basically stock engine with a cam and ported intake/heads that it would move torque up higher in the rpm, but since the table is based on vehicle speed (which torque output vs wheel speed is definitely not a continuous smooth function with the discontinuities when the trans shifts I am assuming the PCM know how to ignore the torque error spikes it gets when it shifts) I am guessing I still get a proportional linear increase in torque output with a cam/heads.

[TriPinTaZ]

I adjusted your VT a little more. Give this one a try. As a general rule, I have found that if you increase Virtual Torque you must also increase Driver Demand in those areas. 30% increase should still be adequate for this revision but if you see throttle closing again go ahead and bump the 100% Pedal Position by +5% or so.

B Tune 2 (my os)(test)_taz02.hpt

Usually if Driver Demand is too low, the ECU resorts to closing the throttle. But there have been some strange things the ECU does in certain situations when the Torque Model is off. To SUPER SIMPLIFY what Howard quoted above, if the Torque Model is off, the reverse lookup that the ECU does will show too large of an error to make sense of. As a response the ECU will resort to random shenanigans including closing the the throttle and random Torque Management Advance(timing pulls).