DBW Pedal Positio Pen vs TB Throttle Position - LotusTalk - The Lotus Cars Community
Reply
 
LinkBack Thread Tools Display Modes
post #1 of 23 (permalink) Old 02-21-2018, 05:53 PM Thread Starter
Registered User
 
eldonz's Avatar
 
Join Date: Jul 2014
Location: Concord, NC
Posts: 910
DBW Pedal Positio Pen vs TB Throttle Position

Guys, those of you that have a data system on your car, have turned off the traction control, have you ever compared the throttle pedal position to the throttle body throttle position? My car is a 11' Exige and I'm curious if you see the phenome non that is in the graph that I have attached. Notice the difference between the two throttle graphs.

Thanks,
Eldon
Attached Images
 
eldonz is offline  
Sponsored Links
Advertisement
 
post #2 of 23 (permalink) Old 02-22-2018, 04:23 AM
Registered User
 
cyow5's Avatar
 
Join Date: Mar 2013
Location: NoVa
Posts: 3,859
With electronic throttles, you *typically* do not just have something that emulates a cable. Rather, it will use a table of pedal position vs. rpm to determine the throttle angle. Sometimes, this is used to get linear torque (50% pedal = 50% torque) or to reduce sensitivity at low rpm, or given the illusion of more power. For that last point, hop in any economy car, hit the gas half way, and then floor it - there's no difference since it already floored it for you to make it feel faster.

What's neat is that with a good setup, an electronic throttle can actually go from, say, 0-75% quicker and with far greater precision than a pro driver, but they aren't always set up that way.

"Lots Of Trouble; Usually Serious"
cyow5 is offline  
post #3 of 23 (permalink) Old 02-22-2018, 04:46 AM
Registered User
 
Rizzydee's Avatar
 
Join Date: Nov 2015
Location: Phenix City, AL
Posts: 367
Does anyone know how many steps are in the electronic throttle body motor? Trying to calibrate mine and need min and max effort (steps) so the ECU knows how far it's opening the TB. Does 30/200 sound right?
Rizzydee is offline  
 
post #4 of 23 (permalink) Old 02-22-2018, 05:12 AM
Registered User
 
cyow5's Avatar
 
Join Date: Mar 2013
Location: NoVa
Posts: 3,859
Quote:
Originally Posted by Rizzydee View Post
Does anyone know how many steps are in the electronic throttle body motor? Trying to calibrate mine and need min and max effort (steps) so the ECU knows how far it's opening the TB. Does 30/200 sound right?
It isn't a stepper motor, so it just uses voltage feedback from the two potentiometers

"Lots Of Trouble; Usually Serious"
cyow5 is offline  
post #5 of 23 (permalink) Old 02-22-2018, 05:17 AM Thread Starter
Registered User
 
eldonz's Avatar
 
Join Date: Jul 2014
Location: Concord, NC
Posts: 910
Quote:
Originally Posted by cyow5 View Post
With electronic throttles, you *typically* do not just have something that emulates a cable. Rather, it will use a table of pedal position vs. rpm to determine the throttle angle. Sometimes, this is used to get linear torque (50% pedal = 50% torque) or to reduce sensitivity at low rpm, or given the illusion of more power. For that last point, hop in any economy car, hit the gas half way, and then floor it - there's no difference since it already floored it for you to make it feel faster.

What's neat is that with a good setup, an electronic throttle can actually go from, say, 0-75% quicker and with far greater precision than a pro driver, but they aren't always set up that way.
Thanks for the feedback. I was wondering if they had used a logarithmic curve but a lookup table with simple interpolation would probably be easier to implement and tune. It explains why the car feels like it is really accelerating in areas of the track when I'm trying to modulate the throttle; ie, maintenance throttle. In your experience, does adjusting this table have much effect on the tune of the motor?

Later,
Eldon
eldonz is offline  
post #6 of 23 (permalink) Old 02-22-2018, 05:42 AM
Registered User
 
cyow5's Avatar
 
Join Date: Mar 2013
Location: NoVa
Posts: 3,859
Quote:
Originally Posted by eldonz View Post
In your experience, does adjusting this table have much effect on the tune of the motor?

Later,
Eldon
None - the fuel table uses throttle position, not pedal. If you get access to the pedal table (maybe Fastworks does?) then it is something you can definitely play around with, just keep in mind that runaway throttle can happen if you type "100" instead of "10" by accident. The PID controller may also have some room for improvement. depending on how it is calibrated and implemented. The worst thing that can happen with the tune is that any errors in the transient fuelling may be exacerbated if you are getting the throttle to move quicker.

With all that being said, engineers are human and free to make whatever stupid choices they want, so there very well could be some quirk in there. I'd be hugely surprised and disappointed though.

Edit: I don't know if the Exige uses speed/density or alpha/n fueling, but either way tuning the throttle-pedal relationship should not have any influence on the tune.

"Lots Of Trouble; Usually Serious"
cyow5 is offline  
post #7 of 23 (permalink) Old 02-22-2018, 06:21 AM Thread Starter
Registered User
 
eldonz's Avatar
 
Join Date: Jul 2014
Location: Concord, NC
Posts: 910
Quote:
Originally Posted by cyow5 View Post
None - the fuel table uses throttle position, not pedal. If you get access to the pedal table (maybe Fastworks does?) then it is something you can definitely play around with, just keep in mind that runaway throttle can happen if you type "100" instead of "10" by accident. The PID controller may also have some room for improvement. depending on how it is calibrated and implemented. The worst thing that can happen with the tune is that any errors in the transient fuelling may be exacerbated if you are getting the throttle to move quicker.

With all that being said, engineers are human and free to make whatever stupid choices they want, so there very well could be some quirk in there. I'd be hugely surprised and disappointed though.

Edit: I don't know if the Exige uses speed/density or alpha/n fueling, but either way tuning the throttle-pedal relationship should not have any influence on the tune.
That was my assumption, too; Pedal position should have no effect on tune since it should only be looking at the throttle position. I'll have to check with Phil and see if FastWorks has access to this table and if I can play with it or not.

Have you ever seen an ECU that looks at rate of change to predict where the throttle should be going to get there quicker? This would produce a quicker throttle response but still give the driver the slower response when trying to modulate the throttle.

Later,
Eldon
eldonz is offline  
post #8 of 23 (permalink) Old 02-22-2018, 07:01 AM
Registered User
 
cyow5's Avatar
 
Join Date: Mar 2013
Location: NoVa
Posts: 3,859
Quote:
Originally Posted by eldonz View Post
That was my assumption, too; Pedal position should have no effect on tune since it should only be looking at the throttle position. I'll have to check with Phil and see if FastWorks has access to this table and if I can play with it or not.

Have you ever seen an ECU that looks at rate of change to predict where the throttle should be going to get there quicker? This would produce a quicker throttle response but still give the driver the slower response when trying to modulate the throttle.

Later,
Eldon
My knee-jerk reaction is that would be a bad idea, but a good PID tune actually does this to some (minor) extent with the D term since it sees the error increasing and adds duty cycle accordingly. It doesn't add requested throttle angle, just tries to keep up. Remember, too, that since everything is digital, taking the rate of change has to have some filtering to get a smooth result, and this filter would create lag that partially undoes the whole the point.

I did a lot of custom DBW code in my Bosch days, and most of the time a lookup table and good PID tune were all that were needed on a non-turbo car. The speed benefit came into play when talking to the transmission for throttle blipping since it could heel-toe much better than any human. Otherwise, it was usually best to just give the driver what he asked for. Drivers are fastest when they are comfortable and in a repeatable environment, so an ECU that tries to guess what you will do next isn't usually the first place teams would want to spend money. With systems like TC or ABS, this was always the big hurdle - trying to distinguish what the driver wants vs. what he does. With DBW, the fastest way around the track is going to be something close to mimicking a cable throttle with some added considerations for traction management (decreased sensitivity at high power levels).

Keep in mind, too, that the TC looks at both the pedal and the actual torque, so changing this relationship *may* interfere with how smoothly that works. That goes back to the TC trying to guess your intentions.

"Lots Of Trouble; Usually Serious"
cyow5 is offline  
post #9 of 23 (permalink) Old 02-22-2018, 07:36 AM Thread Starter
Registered User
 
eldonz's Avatar
 
Join Date: Jul 2014
Location: Concord, NC
Posts: 910
Quote:
Originally Posted by cyow5 View Post
My knee-jerk ...

Keep in mind, too, that the TC looks at both the pedal and the actual torque, so changing this relationship *may* interfere with how smoothly that works. That goes back to the TC trying to guess your intentions.
Having written many control loops from scratch and designing many servo control loops, I am very familiar with PID loops. The "D" term will act like velocity feed forward but is not actually a good substitute for it. All I am saying is that the driver's foot can only move so fast which is why the table is programmed to be at 100% before the pedal reaches 100%. This is exhibited in the traces above. This also shows that you have to lift significantly in order to create deceleration. Now, if you would look at how fast the pedal is changing, you can detect that the driver wants full throttle. If this is the case, go to 100%. If not, stay linked to the pedal position.

From what I can see from the throttle traces, this ECU does not behave vary well for driver feedback. I have other traces where you actually see the driver lifting but the throttle is staying at 100%. On this particular trace, I am driving through turns 3, 4 & 5 at Sebring and modulating the throttle. If I look at the pedal trace, I would say that I left a lot on the table. If you look at the throttle trace, the car is near 100% throttle. I know where my foot is on the throttle and mentally, from the driver's seat, I don't think that I am pushing it hard enough. I believe this is why some drivers do not DBW and prefer DBC systems.

Later,
Eldon
eldonz is offline  
post #10 of 23 (permalink) Old 02-22-2018, 07:52 AM
Registered User
 
cyow5's Avatar
 
Join Date: Mar 2013
Location: NoVa
Posts: 3,859
Quote:
Originally Posted by eldonz View Post
Having written many control loops from scratch and designing many servo control loops, I am very familiar with PID loops. The "D" term will act like velocity feed forward but is not actually a good substitute for it. All I am saying is that the driver's foot can only move so fast which is why the table is programmed to be at 100% before the pedal reaches 100%. This is exhibited in the traces above. This also shows that you have to lift significantly in order to create deceleration. Now, if you would look at how fast the pedal is changing, you can detect that the driver wants full throttle. If this is the case, go to 100%. If not, stay linked to the pedal position.

From what I can see from the throttle traces, this ECU does not behave vary well for driver feedback. I have other traces where you actually see the driver lifting but the throttle is staying at 100%. On this particular trace, I am driving through turns 3, 4 & 5 at Sebring and modulating the throttle. If I look at the pedal trace, I would say that I left a lot on the table. If you look at the throttle trace, the car is near 100% throttle. I know where my foot is on the throttle and mentally, from the driver's seat, I don't think that I am pushing it hard enough. I believe this is why some drivers do not DBW and prefer DBC systems.

Later,
Eldon

Your second paragraph supports exactly what I am saying about it not being good for the throttle to do anything but what is asked. It seems like your first paragraph could be abated the same way - it just needs to do what you asked first and foremost. Looking at your trace, there is one point 2/3rds through where the rpms have a subtle blip upwards, and the throttle decreases - all while the pedal is matted. Are you positive TC was off?

"Lots Of Trouble; Usually Serious"
cyow5 is offline  
post #11 of 23 (permalink) Old 02-22-2018, 07:53 AM
Registered User
 
cyow5's Avatar
 
Join Date: Mar 2013
Location: NoVa
Posts: 3,859
@eldonz here is where I have circled what I am referring to
Attached Images
 

"Lots Of Trouble; Usually Serious"
cyow5 is offline  
post #12 of 23 (permalink) Old 02-22-2018, 08:20 AM Thread Starter
Registered User
 
eldonz's Avatar
 
Join Date: Jul 2014
Location: Concord, NC
Posts: 910
Quote:
Originally Posted by cyow5 View Post
@eldonz here is where I have circled what I am referring to
The top trace is the pedal position and the bottom trace is the throttle position and "Yes" the traction control system is off. I started turning it off when I started noticing the difference between the two. I wanted to see how much of this was being effected by the traction control system. In those traces, you can actually see the throttle position lag the pedal position like you would expect. What I did not like was the fact that lifting on the pedal was not being reflected in the throttle. This, I now believe, is the pedal vs throttle lookup table.

Both of my paragraphs combined would make a better system, I believe. As I stated, the trigger would be the rate of change. Keep the pedal and the throttle tightly coupled if the the acceleration is below a threshold. Exceed that threshold and go to the max of the desired direction.

Later,
Eldon
eldonz is offline  
post #13 of 23 (permalink) Old 02-22-2018, 08:21 AM Thread Starter
Registered User
 
eldonz's Avatar
 
Join Date: Jul 2014
Location: Concord, NC
Posts: 910
Quote:
Originally Posted by cyow5 View Post
@eldonz here is where I have circled what I am referring to
What you are seeing in the top trace is lifting slightly during a high speed corner known as Bishop's Bend. Farther in, I lift to maintain speed but desire no more acceleration. You can do this corner flat if you hit the turn in point correctly. Miss it a little and you cannot get back all the way to the left for the next right. You can enter the next right, mid track, but it does hurt the corner speed a little.

Later,
Eldon
eldonz is offline  
post #14 of 23 (permalink) Old 02-22-2018, 09:03 AM
Registered User
 
cyow5's Avatar
 
Join Date: Mar 2013
Location: NoVa
Posts: 3,859
I'm not saying it wouldn't work in theory, I am saying it would not make a *human* driver faster. As soon as it overshoots one time unexpectedly, the driver will not be 100%. And it will overshoot every now and then such as when the driver decides he wants to back off quickly but was previously headed for full throttle. The ECU, with your approach, will say "here's that 100% you asked for a moment ago!" even though the driver now wants less. Remember, the throttle blade has physical inertia, the electric motor has limited torque, and there are digital filters and delays throughout the system.

That's why I say a comfortable driver is faster. When I set up ABS, a driver could beat it one of ten times for max braking force, but the ABS was fastest overall when it did the same thing ten out of ten times, even if the max force was slightly lower.

There are aftermarket piggyback systems that do what you are looking for though, so you could always try attaching one and trying for yourself. I wouldn't, but you can.

"Lots Of Trouble; Usually Serious"
cyow5 is offline  
post #15 of 23 (permalink) Old 02-22-2018, 11:26 AM Thread Starter
Registered User
 
eldonz's Avatar
 
Join Date: Jul 2014
Location: Concord, NC
Posts: 910
Actually, I have a few other things that I want to try but I need to finish a couple of other projects that pay the bills first.

Later,
Eldon
eldonz is offline  
post #16 of 23 (permalink) Old 02-22-2018, 12:38 PM
Illegal Alien
 
Julian73's Avatar
 
Join Date: Oct 2011
Location: Longview, TX
Posts: 4,439
Garage
Quote:
Originally Posted by eldonz View Post
Thanks for the feedback. I was wondering if they had used a logarithmic curve but a lookup table with simple interpolation would probably be easier to implement and tune. It explains why the car feels like it is really accelerating in areas of the track when I'm trying to modulate the throttle; ie, maintenance throttle. In your experience, does adjusting this table have much effect on the tune of the motor?

Later,
Eldon
The T6 ECU, at least, used in the V6 cars uses a look up table with separate 'curves' for normal and sport modes. BOE's editor is able to access and revises these tables.

'17 Evora Sport 410 GP (No. 1 of 9) LMS alloy pan, BOE headers, SSC harness bar
'12 Evora S GP (No. 1 of 14) BOE Prototype Stage 2 - 420+bHp **SOLD**
Julian73 is offline  
post #17 of 23 (permalink) Old 02-27-2018, 07:25 PM
Registered User
 
Join Date: May 2004
Location: San Jose, CA
Posts: 58
Hey eldonz, it looks to me just like you're overflowing the bottom graph. It looks like the numbers are shifted left by a bit. Anything over 50% is being clipped.

I don't think there's something fancy going on here.

[EDIT] It looks like your max range graphing the throttle body position is 100. What happens if you set it to 255 instead?

2010 Lotus Exige S260
DanKeen is offline  
post #18 of 23 (permalink) Old 02-28-2018, 03:50 AM
Registered User
 
cyow5's Avatar
 
Join Date: Mar 2013
Location: NoVa
Posts: 3,859
Quote:
Originally Posted by Julian73 View Post
The T6 ECU, at least, used in the V6 cars uses a look up table with separate 'curves' for normal and sport modes. BOE's editor is able to access and revises these tables.
Can you get to the PID or filter parameters, too? Those are really what he needs since that's where the lag comes from

"Lots Of Trouble; Usually Serious"
cyow5 is offline  
post #19 of 23 (permalink) Old 02-28-2018, 03:52 AM
Registered User
 
cyow5's Avatar
 
Join Date: Mar 2013
Location: NoVa
Posts: 3,859
Quote:
Originally Posted by DanKeen View Post
Hey eldonz, it looks to me just like you're overflowing the bottom graph. It looks like the numbers are shifted left by a bit. Anything over 50% is being clipped.

I don't think there's something fancy going on here.

[EDIT] It looks like your max range graphing the throttle body position is 100. What happens if you set it to 255 instead?
It clips at 100 because that is the largest value. His labeling is confusing - "throttle" is the pedal, and the "auxiliary" is throttle. In post #11 you can see where he lifts briefly but too quickly for the ECU to respond. This is due to a filter, PID loop, and lookup tables

"Lots Of Trouble; Usually Serious"
cyow5 is offline  
post #20 of 23 (permalink) Old 02-28-2018, 04:51 AM Thread Starter
Registered User
 
eldonz's Avatar
 
Join Date: Jul 2014
Location: Concord, NC
Posts: 910
Quote:
Originally Posted by DanKeen View Post
Hey eldonz, it looks to me just like you're overflowing the bottom graph. It looks like the numbers are shifted left by a bit. Anything over 50% is being clipped.

I don't think there's something fancy going on here.

[EDIT] It looks like your max range graphing the throttle body position is 100. What happens if you set it to 255 instead?
The graphs are auto-scaled by the software, there is no graphical clipping occurring.

Later,
Eldon
eldonz is offline  
Sponsored Links
Advertisement
 
Reply

  LotusTalk - The Lotus Cars Community > Motorsports > Track and Technique

Quick Reply
Message:
Options

Register Now



In order to be able to post messages on the LotusTalk - The Lotus Cars Community forums, you must first register.
Please enter your desired user name, your email address and other required details in the form below.

User Name:
Password
Please enter a password for your user account. Note that passwords are case-sensitive.

Password:


Confirm Password:
Email Address
Please enter a valid email address for yourself.

Email Address:
OR

Log-in











Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page
Display Modes
Linear Mode Linear Mode



Posting Rules  
You may post new threads
You may post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On

 
For the best viewing experience please update your browser to Google Chrome