Originally posted by yaos+--><div class='quotetop'>QUOTE (yaos)</div>

Originally posted by SuperWhiteGT4@
<!--QuoteBegin-yaos

from what i understand, he is trying to make the t/b close all the way when shift. so the RPM drops faster. am i right?

Right, so why would you "lose torque" when it's doing the same thing it's always done, just faster?

ok, just to make sure i got the right idea.

so, do you know why the RPM would drop faster when the t/b is closed all the way?[/b]

Because the car is no longer recieving throttle input, and therefore not trying to accelerate, since with the thottle cable still able to govern it, it is unable to close completely, and thus intaking significant air, which continues the process of engine acceleration.

If the motor is still trying to accelerate (to an extent) when the SMT control unit wants it to deccelerate, it's not going to be able to shift until it gets its "ideal" rpm on both fronts.

 

Originally posted by SuperWhiteGT4+--><div class='quotetop'>QUOTE (SuperWhiteGT4)</div>

Originally posted by yaos+--><div class='quotetop'>QUOTE (yaos)

<!--QuoteBegin-SuperWhiteGT4

@
<!--QuoteBegin-yaos

from what i understand, he is trying to make the t/b close all the way when shift. so the RPM drops faster. am i right?

Right, so why would you "lose torque" when it's doing the same thing it's always done, just faster?

ok, just to make sure i got the right idea.

so, do you know why the RPM would drop faster when the t/b is closed all the way?[/b]

Because the car is no longer recieving throttle input, and therefore not trying to accelerate, since with the thottle cable still able to govern it, it is unable to close completely, and thus intaking significant air, which continues the process of engine acceleration.

If the motor is still trying to accelerate (to an extent) when the SMT control unit wants it to deccelerate, it's not going to be able to shift until it gets its "ideal" rpm on both fronts.[/b][/quote]

yea, i think what you are saying about the SMT control system is right.
but why the RPM would drop faster with the t/b is closed all the way? and the RPM drops slower when the t/b is not closed all the way? its cuz the power drops faster when the t/b is closed all the way. how many air gets in there = how many gas gets in there.

 

Originally posted by yaos+--><div class='quotetop'>QUOTE (yaos)</div>

Originally posted by SuperWhiteGT4+--><div class='quotetop'>QUOTE (SuperWhiteGT4)

Originally posted by yaos
<!--QuoteBegin-SuperWhiteGT4

@
<!--QuoteBegin-yaos

from what i understand, he is trying to make the t/b close all the way when shift. so the RPM drops faster. am i right?

Right, so why would you "lose torque" when it's doing the same thing it's always done, just faster?

ok, just to make sure i got the right idea.

so, do you know why the RPM would drop faster when the t/b is closed all the way?

Because the car is no longer recieving throttle input, and therefore not trying to accelerate, since with the thottle cable still able to govern it, it is unable to close completely, and thus intaking significant air, which continues the process of engine acceleration.

If the motor is still trying to accelerate (to an extent) when the SMT control unit wants it to deccelerate, it's not going to be able to shift until it gets its "ideal" rpm on both fronts.[/b]

yea, i think what you are saying about the SMT control system is right.
but why the RPM would drop faster with the t/b is closed all the way? and the RPM drops slower when the t/b is not closed all the way? its cuz the power drops faster when the t/b is closed all the way. how many air gets in there = how many gas gets in there.[/b][/quote]


you need to spend some time with basic engine theory... seems you dont quite get it. ](*,)

 

Originally posted by yaos
but why the RPM would drop faster with the t/b is closed all the way? and the RPM drops slower when the t/b is not closed all the way? its cuz the power drops faster when the t/b is closed all the way. how many air gets in there = how many gas gets in there.

You are correct that "power is dropping" when you let off the throttle. Furthermore, letting completely off the throttle drops power more rapidly than letting partially off the throttle. However, you've missed two key points: first, power drops to the same level whether it gets there slowly or quickly and second, when the clutch is disengaged there is zero torque being applied to the ground. It doesn't really matter if the engine makes 3000 hp and 4200 lb/ft of torque; if it's not connected to the ground, it won't do anything useful (maybe blow up and look pretty :twisted.

Letting the throttle plate close completely simply reduces the time spent at zero power. And all this mod does is let the plate close completely, every time, in a consistent manner.

Originally posted by 1jzgte
any cons to this mod?

rougher shifting?
harder on the clutch? on transmission?

one you remove the tab is there any way to reverse it w/o havin to buy a new assembly.

I haven't done this yet, but I would guess that it's the same or smoother shifting, same or easier on the clutch, and same on the tranny. Since you don't have any more inconsistent "oops, I released the throttle too late" moments, shifting is smoother on average, but about the same normally. Since you don't have "oops, I pushed the gas too early and smoked the clutch" moments, it's easier on the clutch on average. I don't really see it affecting the tranny one way or the other though; the shift itself is done long before the clutch engages.

Maybe you could bend the tab out of the way instead of cutting it off? Then you could bend it back to reverse the mod.

Originally posted by LittleRocket
The system works by monitoring crank speed (rpm) and Transmission input shaft speed (RPM). When those two shafts are within 500 rpm of one another the smt will release the clutch to engage it and the next gear.

I assume other components also use at least the crank sensor, but is there a way to splice into the signal just before it gets into the shift computer? You could divide both signals to trick the computer into releasing the clutch earlier with a simple circuit. (I'm assuming at this point that it works by sending a "hit" signal every time a tooth passes the sensor.) For instance, to divide by 2 you simply output a hit every other input hit; for 1.5 you skip every third hit. Then the clutch would engage once RPM was within 1000 and 750 RPM, respectively.

Does this seem remotely feasable? You would probably have a control in the cockpit to turn it off and on ("sport mode"). Maybe even a dial to indicate how much division you want.

Also, does anybody know how the clutch mechanism works? If you could slow down the engagement rate (just a bit), you could engage the clutch from much higher RPM without shift-shock. As long as the throttle is still being controlled by the computer, you shouldn't be burning the clutch too much.

 

The rpm signal and the trans signal are very different. I don't know how to do what it is you seek. The crank signal has a bunch of teeth on the tone ring with one missing to indicate tdc. The trans signal has only a few teeth. If you split the signal the smt will try and rev up the engine and totally fubar on down shifts. The best bet is to lighten the rotating assembly as much as possible an 8 lb flywheel goes a long way to making the revs change faster.

 

Ah, so you'd have to somehow trick the ECU on upshifts, but leave it alone on downshifts, while accounting for the TDC missing tooth.

My point wasn't to merge the two signals though. Imagine that each is an analog signal for a moment (that is, at any given point it has a value, instead of computing a value based on hits per time interval). Now, let's say a normal shift looks like this:

Crank RPM       Tranny RPM

6000            6000

Shift -> Clutch disengaged -> Synchros force input shaft to wheel speed in new gear

6000            4000 (for example)

5500            4000

5000            4000

4500            4000

Clutch engaged -> Clutch forces engine to tranny speed

4000            4000

Accelerate -> VROOM!

What I want to do is divide both signals so the difference appears smaller. Here's what the signals would look like to the shift computer using divisors of 1.5 and 2 (real RPM at the left):

1.0 Crank  1.0 Tranny      1.5 Crank  1.5 Tranny      2.0 Crank  2.0 Tranny

6000       6000            4000       4000            3000       3000

Shift                      Shift                      Shift

6000       4000            4000       2667            3000       2000

5500       4000            3667       2667            2750       2000

5000       4000            3333       2667            2500       2000

4750       4000            3167       2667            Clutch engaged

4500       4000            Clutch engaged             2000       2000

Clutch engaged             2667       2667            Accelerate

4000       4000            Accelerate

Accelerate

This all assumes dividing the signal is this simple, of course.

 

Originally posted by fosley+--><div class='quotetop'>QUOTE (fosley)</div>

<!--QuoteBegin-yaos

but why the RPM would drop faster with the t/b is closed all the way? and the RPM drops slower when the t/b is not closed all the way? its cuz the power drops faster when the t/b is closed all the way. how many air gets in there = how many gas gets in there.

You are correct that "power is dropping" when you let off the throttle. Furthermore, letting completely off the throttle drops power more rapidly than letting partially off the throttle. However, you've missed two key points: first, power drops to the same level whether it gets there slowly or quickly and second, when the clutch is disengaged there is zero torque being applied to the ground. It doesn't really matter if the engine makes 3000 hp and 4200 lb/ft of torque; if it's not connected to the ground, it won't do anything useful (maybe blow up and look pretty :twisted.

Letting the throttle plate close completely simply reduces the time spent at zero power. And all this mod does is let the plate close completely, every time, in a consistent manner.[/b]

i didnot miss any point at this idea. you are the one is missing points. different amount of air make different power, this is the #1 point you are wrong about. 2nd, yes, the car got no power to the ground when the clutch is off, but the engine itself is still making power. this is #2 point you are missing. put this 2 points togheter and think again.

 

Originally posted by yaos
i didnot miss any point at this idea. you are the one is missing points. different amount of air make different power, this is the #1 point you are wrong about. 2nd, yes, the car got no power to the ground when the clutch is off, but the engine itself is still making power. this is #2 point you are missing. put this 2 points togheter and think again.

Exactly, but the car NEEDS to lower it's RPM and thus it's power in order to shift the gear if either the car is stock or if you have this mod done. You're going to "lose" the power regardless, it's just a matter of how quickly the car is able to resume building power again.

You aren't affecting anything with this mod other than how quickly the car can drop it's RPMs that it needs to rev match between the transmission input shaft and the crankshaft of the engine. The engine HAS to drop it's revs because when you switch up a gear, you are slowing the rotation speed of the input shaft in the transmission, the engine needs to match that speed so that the shift process is smooth and effective as well.

You CANNOT keep an engine at a constant RPM when shifting, since once you let the clutch back in, it's going to slow down anyways (and roughly mind you). If it didn't, the clutch would be constantly slipping.

 

Originally posted by fosley
Ah, so you'd have to somehow trick the ECU on upshifts, but leave it alone on downshifts, while accounting for the TDC missing tooth.

My point wasn't to merge the two signals though. Imagine that each is an analog signal for a moment (that is, at any given point it has a value, instead of computing a value based on hits per time interval). Now, let's say a normal shift looks like this:

Crank RPM       Tranny RPM

6000            6000

Shift -> Clutch disengaged -> Synchros force input shaft to wheel speed in new gear

6000            4000 (for example)

5500            4000

5000            4000

4500            4000

Clutch engaged -> Clutch forces engine to tranny speed

4000            4000

Accelerate -> VROOM!

What I want to do is divide both signals so the difference appears smaller. Here's what the signals would look like to the shift computer using divisors of 1.5 and 2 (real RPM at the left):

1.0 Crank  1.0 Tranny      1.5 Crank  1.5 Tranny      2.0 Crank  2.0 Tranny

6000       6000            4000       4000            3000       3000

Shift                      Shift                      Shift

6000       4000            4000       2667            3000       2000

5500       4000            3667       2667            2750       2000

5000       4000            3333       2667            2500       2000

4750       4000            3167       2667            Clutch engaged

4500       4000            Clutch engaged             2000       2000

Clutch engaged             2667       2667            Accelerate

4000       4000            Accelerate

Accelerate

This all assumes dividing the signal is this simple, of course.

With your method are you still rev matching - I kinda got lost

 

Originally posted by yaos+--><div class='quotetop'>QUOTE (yaos)</div>

Originally posted by fosley@
<!--QuoteBegin-yaos

but why the RPM would drop faster with the t/b is closed all the way? and the RPM drops slower when the t/b is not closed all the way? its cuz the power drops faster when the t/b is closed all the way. how many air gets in there = how many gas gets in there.

You are correct that "power is dropping" when you let off the throttle. Furthermore, letting completely off the throttle drops power more rapidly than letting partially off the throttle. However, you've missed two key points: first, power drops to the same level whether it gets there slowly or quickly and second, when the clutch is disengaged there is zero torque being applied to the ground. It doesn't really matter if the engine makes 3000 hp and 4200 lb/ft of torque; if it's not connected to the ground, it won't do anything useful (maybe blow up and look pretty :twisted.

Letting the throttle plate close completely simply reduces the time spent at zero power. And all this mod does is let the plate close completely, every time, in a consistent manner.

i didnot miss any point at this idea. you are the one is missing points. different amount of air make different power, this is the #1 point you are wrong about. 2nd, yes, the car got no power to the ground when the clutch is off, but the engine itself is still making power. this is #2 point you are missing. put this 2 points togheter and think again.[/b]

I'm gonna have to agree with what SuperWhiteGT4 is saying...I have no idea how the heck you think a motor works Yaos...but's its cra-zay!

 

Originally posted by SuperWhiteGT4+--><div class='quotetop'>QUOTE (SuperWhiteGT4)</div>

<!--QuoteBegin-yaos

i didnot miss any point at this idea. you are the one is missing points. different amount of air make different power, this is the #1 point you are wrong about. 2nd, yes, the car got no power to the ground when the clutch is off, but the engine itself is still making power. this is #2 point you are missing. put this 2 points togheter and think again.

Exactly, but the car NEEDS to lower it's RPM and thus it's power in order to shift the gear if either the car is stock or if you have this mod done. You're going to "lose" the power regardless, it's just a matter of how quickly the car is able to resume building power again.

You aren't affecting anything with this mod other than how quickly the car can drop it's RPMs that it needs to rev match between the transmission input shaft and the crankshaft of the engine. The engine HAS to drop it's revs because when you switch up a gear, you are slowing the rotation speed of the input shaft in the transmission, the engine needs to match that speed so that the shift process is smooth and effective as well.

You CANNOT keep an engine at a constant RPM when shifting, since once you let the clutch back in, it's going to slow down anyways (and roughly mind you). If it didn't, the clutch would be constantly slipping.[/b]

yup, we going to lose some power at the moment we change gears. lets back to the air. air is a big matter of the gas, and gas = power to a car. lets say, we are having 5 amount of air get in to the engine between first and 2nd gear with out this mod, and it makes power 5. we get 1 amount of air into the engine with this mod, so we make power 1 between first and 2nd gears with this mod. are we going to lose all the power right after the clutch is off? no, the power don't drop like that. and of course, the power 5 can hold better with the power we have made in the first gear than the power 1. this is the reason why the RPM drops faster with this mod. while this mod changes gears faster, but starts with a lower power point from the next gear. this is what this mod doing.

oh yea, you have guys been watching the japanese race video? look at how they change gears. they step on the gas, to have better power point for the next gear.

 

methinks you haven't had any fluid dynamics classes...?

that effect, though technically existant, is very much small.

 

Originally posted by Dude Man Bro
methinks you haven't had any fluid dynamics classes...?

that effect, though technically existant, is very much small.

And it's much smaller when you have the throttle body before the intake plenum. It's in fact negligible.

 

Wow...lots of thoughts here--it seems like the only one that has done this is LittleRocket. If he says that it results in better drive-ability and easier shifts, I think I believe him. I have been waiting for his pictures because it sounds like a great idea. Now that I've seen the pictures and understand what's involved I plan to give it a try.

If I understand your earlier thread on the same subject--the tab provides a direct linkage from the the throttle cable using the tab on the position sensor to engage the back side of a clutch mechanism (or maybe it's just a coil spring?) inside the throttle body. The clutch mechanism then manually operates the throttle plate.

On the opposite side of the throttle plate it looks like there's a motor that is operated by the DBW computer. During a shift if the driver doesn't let up on the throttle fast enough, the motor over-rides the clutch to close the throttle plate. Your theory is that this over-ride of the clutch may/should take longer than if the clutch/manual linkage is disconnected (by cutting the tab off the manual linkage).

I think you figured this out "by accident" because you were trying to use an SMT setup on the 2zz engine swap and had to keep the throttle/pedal position sensor.

Have I got this right? If so, removal of the clutch mechanism might allow the throttle plate to move faster (be more responsive) without the extra mass of the clutch--this assumes that the clutch can be easily removed and doesn't serve some other purpose.

 

Nahh. The clutch needs to stay in place as the car won't shift without it. Just remove the tab.

This wasn't by accident. I was looking for it while I was trying to figure out the 2zz t/b switch over without the clutch and backup cable pull.

 

OK...thanks.
...and thanks again for the pics.

 

Originally posted by CeeDapp
With your method are you still rev matching - I kinda got lost

I would only care about modifying upshifts, as the downshifts are plenty quick for me. So you'd probably have to tap into the upshift signal, and maybe the downshift signal. You start accelerating, get to 6000 RPM (or 6800 or whatever), then press upshift. The shift computer starts doing its shift stuff, and our divider box simultaneously starts dividing the speed signals that the shift computer sees. Since the computer doesn't do anything with the throttle during upshifts, I don't think it'll hurt anything to have the wrong signals coming in.

Now, the point is that we've tricked the computer into thinking the sub-500 RPM difference has been reached, even though we're actually 750 or 1000 RPM away. That means quicker upshifts at the expense of smoothness.

For downshifts the divider box won't do anything, and if it receives a downshift signal while in upshift mode, it'll cancel upshift mode and tell the shift computer what's really going on so the throttle blip will be accurate.

 

Originally posted by fosley+--><div class='quotetop'>QUOTE (fosley)</div>

<!--QuoteBegin-CeeDapp

With your method are you still rev matching - I kinda got lost

I would only care about modifying upshifts, as the downshifts are plenty quick for me. So you'd probably have to tap into the upshift signal, and maybe the downshift signal. You start accelerating, get to 6000 RPM (or 6800 or whatever), then press upshift. The shift computer starts doing its shift stuff, and our divider box simultaneously starts dividing the speed signals that the shift computer sees. Since the computer doesn't do anything with the throttle during upshifts, I don't think it'll hurt anything to have the wrong signals coming in.

Now, the point is that we've tricked the computer into thinking the sub-500 RPM difference has been reached, even though we're actually 750 or 1000 RPM away. That means quicker upshifts at the expense of smoothness.

For downshifts the divider box won't do anything, and if it receives a downshift signal while in upshift mode, it'll cancel upshift mode and tell the shift computer what's really going on so the throttle blip will be accurate.[/b]

I guess I would agree with everything you've written except for the "expense of smoothness". It would seem to me that the syncros are going to be put to task at 750 to 1000 rpm differential.

 

Maybe I'm wrong, but I was pretty sure the synchros had already done thier thing well before then, and it was just the clutch slowing down the shift.

 

You can't mess with the crank position signal without doing seriously adverse things to the engine. The crank position signal is what the ecu uses to fire the coils injectors and refference vvt-i. So split that signal in half and in one revolution you have the injector firing on the exhaust stroke and no ignition until tdc of the exhaust stroke IE mega muffler exploding back fire into the exhaust. You will also get vvt-i codes because the crank signal is the base that the ecu uses to target cam position.....

I would just lighten things up as much as possible. If you want to go one step further get a monster clutch so that is slips less.

You might be able to jack with the transmission signal to get the clutch to release the clutch faster. My guess is the smt ecu will throw an error code due to the signal's drastic change. It may also change how the clutch is engaged as it is a learning unit....

 

Originally posted by fosley
Maybe I'm wrong, but I was pretty sure the synchros had already done thier thing well before then, and it was just the clutch slowing down the shift.

you are not wrong

 

I know the ECU uses at least the crank signal for all kinds of stuff, but I was hoping there was a point between the ECU and the TCU that you could splice into and only affect the TCU without confusing the ECU.

I've only had my Spyder for a couple weeks now, and haven't looked further into the engine bay than the oil dipstick really (ok, so that's not true, but it might as well be). Is the TCU separate, or is it part of the ECU? It'll be a lot harder to hack if it's all one unit, but I don't think it's impossible.

And I completely agree that throwing lightweight stuff on there will be easier and more reliable--just throwing out other ideas.

 

Has anyone else tired this MOD yet?

 

Originally posted by fosley
I know the ECU uses at least the crank signal for all kinds of stuff, but I was hoping there was a point between the ECU and the TCU that you could splice into and only affect the TCU without confusing the ECU.

I've only had my Spyder for a couple weeks now, and haven't looked further into the engine bay than the oil dipstick really (ok, so that's not true, but it might as well be). Is the TCU separate, or is it part of the ECU? It'll be a lot harder to hack if it's all one unit, but I don't think it's impossible.

And I completely agree that throwing lightweight stuff on there will be easier and more reliable--just throwing out other ideas.

The ecu and smt ecu are two seperate untis. They communicate via canbus nework. If you canhack it the mechanicals can shift faster.

 

Finally did this mod today-

Here's my writeup/review of it. I highly recommend it

http://spyderchat.com/forums/showth....php?20130-SMT-Quickshift-Mod-quot-writeup-quot-and-quot-review-quot&highlight=