GOLFMK8
GOLFMK7
GOLFMK6
GOLFMKV

Carista. Throttle response behavior - meaning?

anotero

Autocross Champion
Location
Hither and thither
Car(s)
Mk7 GTI
Ok cool! What I figured when I did my logging is accelerator % and throttle %. Then you need to do some logging - reving it up in park/neutral, and some driving with some light and hard accelerations...at least that's what I did when I logged it. If I'm doing this wrong, by all means folks, let me know but this should give you the data and you can compare it to what I did in my post link above.

I'll log the box on and off this week. Will post log here.
 

tigeo

Autocross Champion

tigeo

Autocross Champion
I'm waiting.
Here you go...not my finest video but it serves the purpose. It's late. I'll redo at some point with a better explanation but if you can't understand it from this and all the other info and data I presented I give up ?

 
Last edited:

SRTie4k

Ready to race!
Location
NH
Regarding data logging, your methodology of "pressing the accerator to some arbitrary position a few times" is not really the best way to ascertain whether the rate of change is different between the "incremental" and "direct" throttle settings, that is, unless you have a very accurately calibrated achilles tendon.

The best way to come to an accurate conclusion about this is to rig up a fixture of some kind with a motorized arm connected to something like an Arduino that controls it at a specified rate. That way you can depress the accelerator pedal accurately and consistently.

Short of rigging up a fixture, I suppose flooring the throttle as quickly as possible would also work to some degree, although you really should have a finer resolution on data logging to examine rate of change.
 

shovelhd

Autocross Champion
Location
Western MA
Regarding data logging, your methodology of "pressing the accerator to some arbitrary position a few times" is not really the best way to ascertain whether the rate of change is different between the "incremental" and "direct" throttle settings, that is, unless you have a very accurately calibrated achilles tendon.

The best way to come to an accurate conclusion about this is to rig up a fixture of some kind with a motorized arm connected to something like an Arduino that controls it at a specified rate. That way you can depress the accelerator pedal accurately and consistently.

Short of rigging up a fixture, I suppose flooring the throttle as quickly as possible would also work to some degree, although you really should have a finer resolution on data logging to examine rate of change.

Flooring the throttle won't show the difference, as it's mainly in the first 1/4 of travel.

That video proves nothing, yet according to Tigeo, we are all idiots..
 

SRTie4k

Ready to race!
Location
NH
Flooring the throttle won't show the difference, as it's mainly in the first 1/4 of travel.

That video proves nothing, yet according to Tigeo, we are all idiots..

Well to be perfectly fair to Tigeo, it is really those who claim it does make a difference who are the ones that should be proving their claims, not the other way around. The way the car comes from the factory should be considered the "baseline", and any theorized modifications in VCDS/OBDEleven should be what is proven with actual data.

In any case, from the brief amount of data and charting I've seen, the sampling resolution is simply just not up to the task of proving it either way. In metrology, our sampling is typically set to 10x the base resolution you want to see/use, and in this application my assumption is that you'd want data sampling in the rate of ms or 1000Hz.
 

tigeo

Autocross Champion
Flooring the throttle won't show the difference, as it's mainly in the first 1/4 of travel.

That video proves nothing, yet according to Tigeo, we are all idiots..

At this point I'm out on this - I've spend a bunch of time on it and if you think you are getting a more 1:1 ratio for your accelerator to throttle, by all means good for you. The data from the pedal tuner will be interesting when it gets posted and I suspect it will more look like I think it will, a more 1:1 ratio vs. the graphs I showed which BTW were from a ~1 mile drive (not just revving which I have also done - did you see those? What more do you want to show this?) including both hard acceleration and softer...I just don't know how to collect any better data for folks. It still comes down to this - you are making a change in the steering control unit and saying it impacts the engine control unit - what other change does this within the VAG computer system? I've showed clearly what changes when you change from indirect to direct. I take it folks think that there is another channel being changed that I don't know about or that isn't shown in my live data. Which one is that? The only thing I've found through all my research on this that could be at least plausible to make a change in that ratio is the "Audi direct throttle mod" - you can search the Google for that one. When you go to long coding in the engine control unit, you can change the brand from VW/Seat/Skoda to Audi. When you do this, you immediately throw out your cruise control. I've been wanting to do the accelerator/throttle logging on that one to see if makes a change. It's plausible b/c it's 1) in the right place within the system and 2) impacts cruise which could be related to the accelerator/throttle channels. I'll check back for the pedal tuner graphs. I don't think anyone is an idiot - I think that folks are experiencing placebo which is rampant on so many minor car mods (think butt dyno on K&N filters and the like). You think you changed something so you feel something different. That's all this is. The data don't support it. But again, if you think they do, by all means ok! What really propagates this is that OBDEleven/Carista call this "direct throttle" in their apps that make this simple coding change.
 
Last edited:

tigeo

Autocross Champion
Well to be perfectly fair to Tigeo, it is really those who claim it does make a difference who are the ones that should be proving their claims, not the other way around. The way the car comes from the factory should be considered the "baseline", and any theorized modifications in VCDS/OBDEleven should be what is proven with actual data.

In any case, from the brief amount of data and charting I've seen, the sampling resolution is simply just not up to the task of proving it either way. In metrology, our sampling is typically set to 10x the base resolution you want to see/use, and in this application my assumption is that you'd want data sampling in the rate of ms or 1000Hz.
Good stuff and I agree. I still think if folks step back and just ask themselves if any of this even makes sense w/r to where this change is located/what it's called/what is changing when you do this (live data shows it clearly). The simple test I did with the graphs in one of the links I posted is as you say, limited and in the end, I don't need to prove it to myself. The way to do it is to drive around for a week with each setting, record throttle/accelerator %s, then calc the ratio, and do the simple stats (min/max/mean/median) for comparison. If I came on here and posted "Hey everyone, I found a great new performance mod using VCDS/OBDEleven...it improves stopping distances/increases brake pressure. It's called driving profile switchover, you change it from incremental controlled over time to direct controlled over threshold value. It's located in the dashboard control unit 17." Some folks would probably say "Wait a minute, shouldn't brake adjustment should be in the brake control module?" But some would say "Oh man, it really improved it!". I guess I could do this as an experiment (pick a mundane change and say that it changes something that it doesn't) and see how many folks report that it worked (placebo)....kind of like those Phycology 101 experiments I participated in in college.
 

tigeo

Autocross Champion

Subliminal

Autocross Champion
Location
Vegas
Car(s)
Slow FWD VW Hatch
Well my entire computer screen isn't being eaten up by one single post of ranting so ya that makes it much easier.

Brevity is the soul of the wit, young padawan.:)
 

Subliminal

Autocross Champion
Location
Vegas
Car(s)
Slow FWD VW Hatch
It's already 27"... and using my 4k 60" TV to read a forum post is a bit overkill imo
 
Top