TP and WOT
 

I've noticed in a lot of logs that at WOT the TB is actually closing up to about 10% as the revs climb. Is this the ECU trying to limit power/load, or is that actually the optimal opening?

I would think that fully open would always be the best for power?

Here's one log that shows it clearly:

http://www.datazap.me/u/wparsons/mar...log=0&data=1-2

Pretty sure thats normal, seems to happen on everyones logs, throttle position tapers to around 90% - seems strange though

It is strange. ECUtek hits about 80%, mine is 80.4% and a friends hits 80.6%, and stays steady at that the whole way through while OFT seems to show it slowly reducing.


Wonder what would happen if you change the throttle opening tables on OFT? AFAIK my tables are set for maximum opening.

I have read about this in a number of posts but no tuner has said anything other than this is normal. No one have offered an explanation. I wonder whether it is a hardware limitation rather than software. On my old car with twin weber carbs and a big cam, the resistance in the throttle changes noticeably when the cams/power starts to kick in. It maybe the increased airflow at high RPM causes the throttle to close slightly on the FA20 but that is pure speculation.

this is a function of the DBW control tables. 100% accelerator input != 100% throttle plate actuation. if the throttle plate closes early, it's because the dbw tables are telling it to.

Holy shit, I'm seeing ghosts! :D

Has anyone done any testing to see if opening the throttle plate to 100% improves hp up top or not? I find it counter intuitive that a smaller opening would make more power high up. Down low I can believe it, but I have nothing to back that up with beyond my gut feeling.

having the throttle plate close early, at least on an FI car where I've tested it, will cost you power. this is a common mistake made on 86's that goes something like this:

• person sees table labeled 'WOT torque threshold' that seems to read in nm.
• person says 'wow... i want more torque.. awesome!' and bump up the values without adjusting the other dependent tables
• person gets upset when they log their throttle plate angle and see it closing (usually through the midrange because they thought that taking the 'dip' out of that table would eliminate the torque dip)

basically, increasing values in WOT torque threshold causes the throttle plate to open *less*, which is counterintuitive. in order to maintain full throttle angle with a reasonable WOT torque threshold curve you also have to edit the dependent tables.

for a more detailed explanation of how this all works: http://forums.nasioc.com/forums/show....php?t=1537010, though it doesn't apply directly to our ecu, it'll give you an idea of how the tables relate to one another.

To be more application-specific: The 'WOT Torque Threshold' value is divided by the current value in 'Accel Pedal to Desired Torque', which is then used as an input to 'Desired Torque to Throttle Opening' to resolve a desired throttle plate angle.

Ok, so I'm not going crazy! I'll give that link a read later, seems a lot more complicated than I originally expected.

it is. DBW is probably the most convoluted part of tuning an 86. 'principle of least surprise' doesn't apply here.

@jamesm is the max opening on ECUtek about 80% or have you managed to get it to open more? The tables suggest that it's requesting 86% so I figured that 80% was the maximum the physical could manage.

i typically see 80-81%, hovering between 80.4 and 80.8 most commonly.

But the question still remains: why is the throttle plate closing early?

Is it an engine-saving measure, so that the cylinders don't get starved for air b/c the air can't flow fast enough to get the proper amount into each cylinder quickly enough at those high RPMs? (b/c the higher the revs, the less time there is for the cylinder to fill with the right amount of air before the valve closes). As I'm just speculating, that's only my off-the-cuff WAG.

Btw, @jamesm, welcome back :).

I don't recall seeing this behavior with the stock roms. It's been a while since I've seen a stock car, though. Next time I have one in for a tune I'll be sure to take note of it.

I know that in the STI/WRX, the throttle plate begins to close before redline. I saw this in the throttle maps COBB showed as part of their "See what we did with the SI-drive selector" descriptions, by giving all 3 modes the throttle mapping of S#, but only changing the boost target instead. The curves showed at a 6700rpm redline the throttle plate beginning to close somewhere between 6k and 6.4krpms, back to full close at redline.

Maybe that's a better way to limit revs than the harsh fuel cut in our 86's? B/c I never remember getting the driveline lash that the 86 gets when it hits the rev limiter when the STI would hit its rev limit.

EDIT: Actually it was in Subaru's own marketing description of SI-drive that shows the changing throttle map. http://www.iwsti.com/forums/1795632-post4.html

See how the throttle actually dies off before pedal reaches full open in every mode.