Some info on the torque-based throttle follower

[smokeshow]

In Gen4 applications, the throttle follower is quite literally that - airflow that 'follows the throttle'. When you're on the pedal, you have control of the throttle (most of the time). When you're off the pedal, it should return to idle control. However if you just snapped the throttle shut and went straight back to your idle airflow target, say after a pull at high RPM...the amount of engine braking would probably knock your head into the steering wheel, among other things. So the follower exists to allow a blend from the on-pedal airflow to idle airflow.

Might be useful to define what those airflows are, seeing as there is no 'base running airflow' table, as in Gen3. On pedal, Gen4 airflow is just what results from the throttle position as determined by driver demand. (Gen5 expands to include on-pedal torque control as well as off-pedal speed control, a true closed loop torque system...but Gen4 just has the speed control.) Off-pedal idle airflow at steady state is a function of torque, which is a function of engine friction, accessory frictions, alternator load, AC load, idle reserve...you get the point. All of these are summed up and input to the inverse of the airmass torque model where a cylinder air mass can be determined. From that and RPM, you have a steady state airflow.

Meanwhile when you're on-pedal, the values from the throttle follower table are also input to the air mass torque model and checked against the steady state airflow. When you let off the pedal, the difference between them is added to the learned integral airflow correction (depending on the LTIT cell values) and then added to the steady state airflow. It then ramps out depending on the follower increment/decrement tables.

Here's a screenshot from a 2011 truck E38, which has most of the CAN variables for this defined. You may not be able to see these on a lot of cars, as HPT doesn't have many of them defined. The leftmost arrow shows the tip-out of the throttle, at which point the throttle position ramps down to throttle follower airflow. The middle arrow shows the integral getting the TF airflow added in to begin the ramp out, and the last arrow shows the end of the ramp.

tf.PNG


Keep in mind...you could get airflow pulled out by the throttle follower integral rather than added in...just depends on your calibration and operating conditions.

[LS ROB]

In Gen4 applications, the throttle follower is quite literally that - airflow that 'follows the throttle'. When you're on the pedal, you have control of the throttle (most of the time). When you're off the pedal, it should return to idle control. However if you just snapped the throttle shut and went straight back to your idle airflow target, say after a pull at high RPM...the amount of engine braking would probably knock your head into the steering wheel, among other things. So the follower exists to allow a blend from the on-pedal airflow to idle airflow.

Might be useful to define what those airflows are, seeing as there is no 'base running airflow' table, as in Gen3. On pedal, Gen4 airflow is just what results from the throttle position as determined by driver demand. (Gen5 expands to include on-pedal torque control as well as off-pedal speed control, a true closed loop torque system...but Gen4 just has the speed control.) Off-pedal idle airflow at steady state is a function of torque, which is a function of engine friction, accessory frictions, alternator load, AC load, idle reserve...you get the point. All of these are summed up and input to the inverse of the airmass torque model where a cylinder air mass can be determined. From that and RPM, you have a steady state airflow.

Meanwhile when you're on-pedal, the values from the throttle follower table are also input to the air mass torque model and checked against the steady state airflow. When you let off the pedal, the difference between them is added to the learned integral airflow correction (depending on the LTIT cell values) and then added to the steady state airflow. It then ramps out depending on the follower increment/decrement tables.

Here's a screenshot from a 2011 truck E38, which has most of the CAN variables for this defined. You may not be able to see these on a lot of cars, as HPT doesn't have many of them defined. The leftmost arrow shows the tip-out of the throttle, at which point the throttle position ramps down to throttle follower airflow. The middle arrow shows the integral getting the TF airflow added in to begin the ramp out, and the last arrow shows the end of the ramp.

tf.PNG


Keep in mind...you could get airflow pulled out by the throttle follower integral rather than added in...just depends on your calibration and operating conditions.

Could this cause the the idle timing tables to be referenced when light throttle is applied? I've had a problem forever with my ecm bouncing between on pedal and idle mode. When I log "Throttle Control Source" you can see it flip floping. I've gotten it really close now and barely can feel it, but I still know its there.

[smokeshow]

I've never seen the throttle follower cause that...but who knows what sort of failure modes you could find yourself in if the cals are too far off...

[LS ROB]

I see. I appreciate your response.

[5FDP]

Honestly, somedays having my head hit the steering wheel wouldn't be all that bad.

[Cringer]

Off-pedal idle airflow at steady state is a function of torque, which is a function of engine friction, accessory frictions, alternator load, AC load, idle reserve...you get the point. All of these are summed up and input to the inverse of the airmass torque model where a cylinder air mass can be determined. From that and RPM, you have a steady state airflow.

What torque value is used to define what is needed for off pedal idle? Is this exposed by HPTuners (Virtual Torque or something else)?

As far as I can tell the ECM calculates the target airflow to achieve the proper torque/desired idle RPM in the PID [Desired Idle Airflow]. So is the trick to bring Idle Integral Airflow to 0 g/s is as simple as accurizing the Virtual Torque Tables and commanding the proper base idle spark advance?

[hjtrbo]

Thanks Jake, best sleep ever

[Allen Vos]

Could this cause the the idle timing tables to be referenced when light throttle is applied? I've had a problem forever with my ecm bouncing between on pedal and idle mode. When I log "Throttle Control Source" you can see it flip floping. I've gotten it really close now and barely can feel it, but I still know its there.

Sorry to revive an old thread, but Did you ever figure any of this out?

I am having a similar issue where if I give it a little pedal, the ECM bounces between pedal and idle mode. In my case it brings in the adaptive spark timing and tries to pull timing to reach the target idle RPM. It only at idle , not when driving.

Is there any settings for this at all?

I know you had said, you got things close now and was hoping you could offer some insight into this?

[LS ROB]

Having fueling spot on seems to help minimize the effect, but i believe my BRAF and Scalar had something to do with it. Not saying thats your problem, but when i set my scalar back to stock, it helped some, so i feel like its in the throttle/throttle body realm of things, but cant say for sure.

[Cringer]

so i feel like its in the throttle/throttle body realm of things, but cant say for sure.

So perhaps Allen you can try setting this table back to stock as a test
[ECM] 33030

[Allen Vos]

Having fueling spot on seems to help minimize the effect, but i believe my BRAF and Scalar had something to do with it. Not saying thats your problem, but when i set my scalar back to stock, it helped some, so i feel like its in the throttle/throttle body realm of things, but cant say for sure.

Appreciate the response Mr. I'm up for trying what I can to try and lick this lol. Its not an end of the world problem and the car drives fine, but my OCD wont let me leave it alone lol.

When you say Scaler, I don't mean to sound dumb, but which table are you referring too?

I do have fueling from what I feel is pretty spot on and have felt the difference when its not been spot on.

[Allen Vos]

So perhaps Allen you can try setting this table back to stock as a test
[ECM] 33030

I'm not sure if you caught my log and tune from the other post at all but I should say that I put a few tables back to stock to try and figure my issue out.

the ECM 33030 Desired throttle Area was one of them, I returned to stock, along with the torque tables and the throttle follower table.

Changing these did nothing for my issue unfortunately.

I'm not sure what your thoughts are on modding the ECM 33030 table are but I had done this when the car was bone stock and it made a big difference on throttle response so I would prefer to keep this table modified unless there is a very good reason not too.

I feel I am going to just start a fresh thread on trying to understand the relationship between idle and pedal for throttle control source.

If we can understand how it works and what it does then it might be easier to diagnose and figure this out.

I was kind of wondering if this had much to do with the torque values of the MAP/Air mass torque tables.

I know some guys say they need to be changed, well at least on the idle , but other guys say not to touch them.

I am still trying to get a better understanding and the relationship of the torque tables to all this .

[hjtrbo]

Could this cause the the idle timing tables to be referenced when light throttle is applied? I've had a problem forever with my ecm bouncing between on pedal and idle mode. When I log "Throttle Control Source" you can see it flip floping. I've gotten it really close now and barely can feel it, but I still know its there.

There is a "Throttle threshold used to determine an idle engine condition" scalar. I don't know if it's referencing TPS or effective area. Is your TPS a bolt on? You may be able to give it the slightest of tweaks then nip the fastners back up. OR, now I have no fkn clue, but hit the baro tables. Just in the idle transition area where you see the PID change, try 5-10% each way. Side effects and likely hood of success unknown.

[GHuggins]

There is a "Throttle threshold used to determine an idle engine condition" scalar. I don't know if it's referencing TPS or effective area. Is your TPS a bolt on? You may be able to give it the slightest of tweaks then nip the fastners back up. OR, now I have no fkn clue, but hit the baro tables. Just in the idle transition area where you see the PID change, try 5-10% each way. Side effects and likely hood of success unknown.

Can you post a pic of that? His problem is his engine is so efficient now that it requires very little throttle input and is hitting much lower loads than it ever did stock. If memory serves when he drops below 4ish percent pedal it goes to idle mode. I was able to help him by changing his tq model so as the airmass reading of the spark tables were affected to the point that it would think it was getting a touch more pedal and then by massaging some of his timing tables at the same time, but it still hasn't fixed the root cause which has been the idle accelerator pedal percent. I also already shifted his baro tables around, but it didn't seem to do anything other than calm the bigger tb down.

[hjtrbo]

Edit: Definitely effective area. Interpolates to 4.06% tps. Your memory is good.

[GHuggins]

Edit: Definitely effective area. Interpolates to 4.06% tps. Your memory is good.

You my hero once again

Rob, looks like there may be a permanent solution.

[hjtrbo]

It'll take a few hours to dig through the bin to get it for Robs OS. Add it to my list (or if Verlon has nothing on )

[LS ROB]

You my hero once again

Rob, looks like there may be a permanent solution.

That's great news! I'm interested to see what comes of this!

[LS ROB]

It'll take a few hours to dig through the bin to get it for Robs OS. Add it to my list (or if Verlon has nothing on )

Hey brother I appreciate that! No rush here, so, just whenever, if ever is fine. Thank you!

[hjtrbo]

Don't hold your breath, Greg's still waiting for his job.