Throttle Linearity Settings

[Unicorn]

There's a torque model/capability table in the ECU (at least one). There are then various tables which explain how much torque is wanted at different engine speeds and accelerator pedal positions. Through some transformation with ratios etc it ends up on a table which converts the torque demand vs capability into an end throttle butterfly angle.

Unfortunately it doesn't quite work as you'd expect and there is some behaviour you can't get rid of, yet, with the tools currently out there. I've only confirmed this with brzedit, but i've heard similar from other platforms.

[King Tut]

Originally Posted by Unicorn

There's a torque model/capability table in the ECU (at least one). There are then various tables which explain how much torque is wanted at different engine speeds and accelerator pedal positions. Through some transformation with ratios etc it ends up on a table which converts the torque demand vs capability into an end throttle butterfly angle.

Unfortunately it doesn't quite work as you'd expect and there is some behaviour you can't get rid of, yet, with the tools currently out there. I've only confirmed this with brzedit, but i've heard similar from other platforms.

Nice username change. I take it your old account got banned.

[AVOturboworld]

Originally Posted by AdrianG

RomRaider can see ~328 tables at this time, and I wouldn't doubt there's more that haven't been added to the definition yet.

We once paid to have a 2011 STI's maps completely defined so we could pass some testing in Japan, mainly in regards to cold start. There was 2500+ maps, 136 of them for cranking. And I have no doubt this ecu is even more complicated than that one.

[Unicorn]

Originally Posted by King Tut

Nice username change. I take it your old account got banned.

Dear King Tut,

Please don't be ridiculous.

Yours sincerely,
Unicorn

[Td-d]

Originally Posted by AdrianG

RomRaider can see ~328 tables at this time, and I wouldn't doubt there's more that haven't been added to the definition yet. It's a big complicated beast to wrap your head around. Luckily we have our factory ROMs as well as Shiv's open maps so we can identify the areas to focus on.

There are hundreds of 2D and 3D tables alone, of which only a fraction are understood/defined, and literally thousands of single byte, double word and float parameters.

What has been defined over the years is a very limited set of what's in the ecu - on the other hand, generally the most important levers.

Sent from my iPad using Tapatalk HD

[AdrianG]

Originally Posted by Td-d

There are hundreds of 2D and 3D tables alone, of which only a fraction are understood/defined, and literally thousands of single byte, double word and float parameters.

What has been defined over the years is a very limited set of what's in the ecu - on the other hand, generally the most important levers.

Sent from my iPad using Tapatalk HD

I really appreciate your definitions and the work you've done to get to this point. I'm learning a lot about this right now and I'd be pretty lost without your contribution. Thank you!

- AdrianG

[Td-d]

Originally Posted by AVOturboworld

We once paid to have a 2011 STI's maps completely defined so we could pass some testing in Japan, mainly in regards to cold start. There was 2500+ maps, 136 of them for cranking. And I have no doubt this ecu is even more complicated than that one.

I would kill to see that rom and definition


Sent from my iPad using Tapatalk HD

[Unicorn]

Originally Posted by AVOturboworld

We once paid to have a 2011 STI's maps completely defined so we could pass some testing in Japan, mainly in regards to cold start. There was 2500+ maps, 136 of them for cranking. And I have no doubt this ecu is even more complicated than that one.

Dear AVOturboworld,

That is nothing. I have A2L's here with in excess of 10k tables in them.

I can't think of a worse job than an OE calibration engineer.

Yours regretfully,
Unicorn

[mad_sb]

I posted how all the tables work months ago. I'm on my phone in the middle of moving so I can't dig it up right now, but if you search for threads started by me you will find it.

[arghx7]

I'd like to point out some principles of throttle design, pedal mapping, and torque control calibration: Throttles are inherently non-linear due to the laws of physics. Cable-driven throttles from "the good old days" were never linear either. Here's an interesting picture I pulled off Google image search of a cam-type mechanical throttlebody linkage.



Just like a camshaft to operate cylinder head valves is an eccentric circle, a cam to operate an old throttle has different ramp rates depending on where you are in the rotation of the cam. So let's dispel that myth of "my old blahblah had a linear/almost linear throttle."

Here's the physics behind why a throttle valve is non-linear. The throttle works on a pressure difference upstream and downstream of the valve.



When the valve is nearly closed, you have what's called "choke flow." The air reaches the speed of sound as it passes through the opening. This is called the "sonic area" in throttle body design and throttle calibration. That zone behaves differently than when the throttle is open further.



So depending on the opening angle of the throttlebody, the air could be flowing at a constant velocity and then suddenly as the valve opens further, the air velocity drops.

Originally Posted by SkullWorks

you will find if you actually make the pedal to TB a direct correlation that 50% is essentially the same as 100%

As with most modern engines the TB is simply WAY TOO BIG for the actual airflow demands...

For example, I have a S13 track slut, it makes 450hp 390 ft lbs, I have a ~70mm TB on that car...after ~60% you achieve the same exact power curve as WOT. I use this range to setup boost VS TPS...works great

Consider the basic relationship between opening area and volume flow and it will all make sense.



So in conclusion, even before drive-by-wire throttle valves never behaved linearly. The relationship between the gas pedal position and the opening of the throttle valve was never linear either.

Now, can we prefer a different response curve to the way the car came from the factory? Sure. That's what software tuning is for. Think about it this way... with cable throttle you were stuck with whatever non-linear response curve the engineer chose, but with modern tuning you can make your electronic throttle behave according to your preferences.

I'd also like to point out that engine torque control on a modern engine includes not just throttle valve but cam phasing/valve lift, spark timing, air-fuel-ratio, and wastegate position (in a car with a stock turbo). So this discussion of response vs gas pedal position really includes the whole torque model rather than just the opening of the throttle valve itself.

[Fizz]

Sorry to hijack, but can the throttle map be adjusted to get rid of the slight 'dead travel' I'm having at the top my pedal stroke? I reset the ecu last Tuesday and since then the first 1cm of throttle has gone dead...I mean you can press it that much without getting any revs whatsoever. Beyond that its all normal...

[Unicorn]

Originally Posted by Fizz

Sorry to hijack, but can the throttle map be adjusted to get rid of the slight 'dead travel' I'm having at the top my pedal stroke? I reset the ecu last Tuesday and since then the first 1cm of throttle has gone dead...I mean you can press it that much without getting any revs whatsoever. Beyond that its all normal...

Dear Fizz,

No.

Yours sincerely,
Unicorn

[Fizz]

Originally Posted by Unicorn

Dear Fizz,

No.

Yours sincerely,
Unicorn

So nothing can be done to rectify that?

[Unicorn]

Dear Unicornettes,

Please find below latest data from the rainbow with regards to stock torque control tables used with a Sprintex/Innovate supercharger (charge cooled):





I hope it is obvious that this is not ideal.


Yours sincerely,
Unicorn