Originally Posted by titaniumkingpin

evo 9 ya, please let us know if your car is fixed and where that restrictor is. pictures would be great.
keep us updated

Originally Posted by Turbocake

EDIT - please read the whole thing before posting a question.

I believe some folks are having different versions of the same problem, but I think that for the vast majority, my diagnosis is correct:

Unless the induction parts upstream of the throttle body are somehow manipulating the EFI's sensors in such a way that the Idle Air Solenoid is immediately cracked wide open (doubtful, because IAS's typically can't raise RPM's above 3k and they don't open all that fast), then none of the induction parts upstream of the throttle blade matter whatsoever.

The ultimate dictator of engine RPM is (air, via) the throttle body. If it's not getting in through this part (in an otherwise sound engine), it's not getting in, and RPM's will not climb. PERIOD.

BOV, Piping, all that stuff = Irrelevant - EXCEPT for mechanical issues - that we would ALL mysteriously be having in perfect unison - like the throttle spring, bearings of the throttle body, throttle cable, gas pedal, etc., there is nothing else that could possibly cause these specific conditions (RPM jumping between shifts) other than the throttle body being open to some significant* (<---more on this in a moment) degree to allow it to happen.

The gentleman, for instance, who said his gas pedal stayed down while the car pulled on it's own - that's definitely a distinctly different problem than what most of us are experiencing. I would say that car has something wrong with the aforementioned hardware-based issues - possibly the throttle body. Some throttle bodies made back in the day by BBK had a bearing tolerance issue that would only make the throttle stick when the car was running. Turn the car off and the problem went away. This was heard about, but not widespread.

With Evos, the fact that the symptons are so widespread amongst car and people types lends credit that it's an operation issue rather than an oddball problem for a few.

That said, let's get back to airflow. I've noticed that I had this problem after putting my mods on (Buschur stage II) and that it's gotten much worse since winter's cold air arrived.

I then decided to watch my AFC-II while driving around to get some hard numbers. (Boy do I miss datalogging with good software like EFI Live) and I noticed that when and where I thought was off the gas, I was not. I couldn't tell at all by feel, I had to see the numbers on the AFC-II to believe it. I tried modifying the timing of my shifts to what I thought would make the car buck from letting off the gas too soon, and it was perfect. Driving the car barefoot also helped my accuracy because I could perfectly feel the pedals.

Reasoning:

Boost is an incredible factor with our cars because the boost pressure is so high and the rotating mass of the engine is so low.

Check this out... start your car and rev it up. Watch your boost guage and see how much boost you get before you bounce the limiter... not much, right? That engine still revved way up and hit the rev limiter, right? Did you even have it all the way floored? Probably not.

Now, imagine the few moments in time when you've just "let off" the gas under boost. The concentration/density of air is much higher and much more air will flow through a given throttle blade angle. If you had magical boost just waiting, at full pressure all the time, your engine would free-rev so fast that you wouldn't be able to, at your FASTEST, snap the throttle all the way to the floor (nevermind all the way back up) before the engine spiked to redline. It's easy to see how even a few thousandths of a second could matter here. Face it, most of us aren't that precise at shifting, and this nuance exploits it perfectly.

Additionally, when you've let off, the density of that air may even slow the throttle blade down all on it's own - that's over two atmosphere's of air density in there and lots of flow forces acting on the blade actuation. TB bearings must be precise and smooth (under operating loads). The throttle spring is pulling against those forces and doing the chore of pulling your throttle cable (try rerouting it under the hood. I've noticed that the clip closest to the firewall is way out of place and at the wrong angle. Just pull the cable out and let it hang free for smooth operation.)

You could also install a secondary throttle spring. Better automotive and hardware stores will have springs you can use. This will give your foot a more consistent pedal position feel.

The weather matters because the air just gets more dense (and the problem worse) when it's cold outside - the air is cold to start with, then it gets boosted, and THEN cooled by an even colder-than-normal intercooler. This is incredibly dense air, especially compared to the 6-9 psi most "turbo" cars see.

Meanwhile...

At the exact same time this compressed air is sneaking around the on-it's-way-to-being-closed throttle blade, your clutch is being pressed by your left foot with the intention of having it be timed correctly with the let-up of your gas foot.

Problem is, the timing you have selected for your clutch pedal is based on the percieved release point, which is most often affecting/writing your cranial habit-code when you're putzing around, driving nice and slow and smooth.

The point is that identical clutch timing from a "non-boosting" shift to a "boosting" shift will show huge differences in actual clutch-release times just because there is a huge amount of power breaking that ***** loose in one instance, and in the other, hardly any force at all.

The reason the problem is more pronounced at mid-range RPM is because this is where boost swell is most often coupled with relatively low piston-pumping by the engine's internals. A little bit of leaked-by air at these RPM's will go further to free rev the engine than the same amount of air applied to an engine that is already gobbling just-about that much already. When you shift at nearly redline, your engine is at the far end of operating efficiency, rather than being right in the middle (at ~4k, for instance).

The cumulative point is that you're getting burned by both pedals when boost is involved, and especially in the middle RPM range. The simple (and beautifully CHEAP solution) is to proactively and intently train your shift timing to compensate for the clutch's want to slip sooner, under the load - and to lift your right foot sooner AND higher. If you don't hear that BOV cracking open hard, you're not doing it right.

I've since concluded that this is a very lame byproduct of small displacement/high-boost turbo cars, but one I can live with.

For the folks who are competent and have pursued the problem with the above in mind, and still have a problem, I would look at some cause of unmetered air entering the engine between the throttle blade and intake manifold gasket. There could be something disconnected, cracked, or otherwise letting air in.

Part of why I'm fairly certain that this is about driver error is because so few alternatives exist. Engines don't just rev up "because". They need air.

Also, the guys with the aftermarket EMS's are potentially in their own category. Odd that so many platforms can share the same problem, eh?

Cheers,

Originally Posted by 90GSX-03EVO

It's very strange to me that this restrictor valve would work like this. It's designed to slow engagement of the clutch, but I guess it could work to slow disengagement, as well. With that said, I don't understand how or why the engine would increase rotational speed when the throttle is lifted and the clutch is disengaged. That would be like if you were turning a wrench, and the wrench continues to turn on it's own - and with more torque than when you were turning it by hand.

I'm not saying that what you did is not the cure. I'm just not understanding how the laws of physics are working on this one.

Very strange problem we are seeing here.

Originally Posted by titaniumkingpin

is this the same thing they call the clutch pill restrictor? if so, mine is already out.

Originally Posted by evo 8 ya

Yea it is the same thing. Today I went for a drive and it is still happening . So back to the drawing board I guess.

Originally Posted by 90GSX-03EVO

It's very strange to me that this restrictor valve would work like this. It's designed to slow engagement of the clutch, but I guess it could work to slow disengagement, as well. With that said, I don't understand how or why the engine would increase rotational speed when the throttle is lifted and the clutch is disengaged. That would be like if you were turning a wrench, and the wrench continues to turn on it's own - and with more torque than when you were turning it by hand.

I'm not saying that what you did is not the cure. I'm just not understanding how the laws of physics are working on this one.

Very strange problem we are seeing here.

Originally Posted by mister_pants

I am still trying to understand how many came to the assumption that a clutch throw-out bearing could have anything to do with this problem. As a matter of fact I dont see how a clutch could cause this at all. Early on in this thread the person that was having the most trouble with this problem clearly stated that with his foot off the gas (at high RPM and boost) when he depressed the clutch the ROM's would raise. If this were a clutch related problem, with the throttle shut and the engine decelerating suddenly engaging the clutch would not cause the RPM's to raise the reported 1500~2000rpm's as being reported. I would definatly start looking elsewhere. Another thing that I must disagree with is when EVOVII_SWE stated that an Idle Air Valve could not raise rpm's above 3k. IAC's regulate air, plain and simple, with the throttle SHUT the IAC may only be able to let enough air in the raise the RPM's to 3k but with the car high in the RPM range if the IAC were to go full open than it is conceivable that it would allow sufficient air to raise the RPM to the reported levels, especially under a "no load" situation like between shifts. I am not saying the IAC is the culprit, I just wanted to correct EVOVII_SWE about the function of the IAC. For you guys running aftermarket angine management check your tables. I have not had any experience with AEM but I have used both Accel and Electromtive in addition to spending countless hours on GM port fuel injection. If you have a decel table or function check that. One last thing....and this is mainly inspired by 90GSX-03EVO's arguement on driver error. Is it possible that your kicking the floormat up enough to hold the throttle? Once upon a time during a race that happened to me where my f**king floormat held my car at damn neat WOT!! I know nobody wants to admit to driver error, and in some of these cases it would seem like something else is going on, whatever the case I doubt it is in the clutch.
-okay my rant is done

Originally Posted by mister_pants

I am still trying to understand how many came to the assumption that a clutch throw-out bearing could have anything to do with this problem. As a matter of fact I dont see how a clutch could cause this at all.