Cafe, Carb = Hesitation

[wwest]

In doing some research on the internet I have found several engineering "white papers" and a few actual manufacturer documents which just might explain the problem.

Based on what I have found there can be no question that the manufacturers are using the lockup aspect of the torque converter to extend the fuel economy and thereby very likely reduce overall emissions.

Historically the torque converter has been used as a "clutch" to prevent the engine from stalling with the transmission engaged. It is also used as a torque multiplier, as high as a factor of 2, when accelerating.

While the torque converter is quite helpful in that multiplication mode it also results in the loss of about 10% of the fuel economy in acceleration mode and even a greater percentage loss when you're just cruising along.

So, according to the papers and documents some manufacturers are locking up the torque converter in all gears, even in 1st.

Obviously the torque converter cannot be locked up if wheel speed and selected gear ratio is such that the engine might stall.

So, you're in that merge lane, you slack off the gas slightly (throttle NOT fully closed) waiting for an "opening", the system says "ah-ha, we're cruising" and slips the transaxle into a cruise gear ratio and the torque converter into lock-up.

Now you suddenly ask for acceleration and the lockup must be released (has feedback "loop" closure indicated the previous upshift and lockup "completion"?) and the transaxle must be downshifted into the appropriate gear.

Unless Lexus has added another sensor in order to "know" the lockup sequence has completed the engine RPM and the torque converter turbine speed must be equal.

The problem, as I see it, is that these inputs can also be "equal" at times, strictly by happenstance, with the lockup disabled.

So, you re-apply that gas pedal to accelerate, the engine is "dead" as programmed during a gearshift or lock/unlock sequence, and in this case remains DEAD because the engine RPM and the turbine sensor speed are the same just because of happenstance.

Until these two speeds diverge there is no way for the controlling ECU to "know" the "unlock is complete.

Just another guess.

But.

If the torque converter lockup could be used full time then the fuel economy and emissions rating between manual and automatic transmissions would be eliminated.

But being able to use the torque multiplication and/or the lockup "at will" might just result in better fuel economy and emissions ratings than a manual transmission.

In any case it appears that Toyota and Lexus have come a LONG way down the "lockup" road and it's likely there's is not going back absent paying HUGE CAFE and/or CARB penalties.

[SW03ES]

NOW you're getting there...

[wwest]

Simpler explanation, maybe.

remember, I have NO personal experience except that which I have related regarding my 2001 AWD RX300.

But my guess is that the "key" to this whole thing is our/your specific "treatment" of the gas pedal.

You accelerate up to 35MPH in the acceleration lane and then you ease off the gas pedal waiting for, or to "match up" with an opening in the traffic flow. If you only ease off, still maintain even a small throttle opening, light pressure on the gas pedal, the system is very likely to switch to "cruise" mode, and engage the torque converter lockup clutch.

If, on the other hand, you lift off the gas completely the system will obviously assume a coastdown circumstance and keep the lockup disabled.

This theory also lends credence to the left foot braking having some unknown effect that leads left foot brakers to encounter the hesiation symptom more often. I would suspect that when one uses the left foot for braking the right foot may very well "rest" on the gas pedal just lightly enough to keep the throttle slightly open.

But back to the original slight pedal pressure theory.

You slacked off the gas and the system locked up the torque converter for low torque engine cruise fuel economy.

Now you suddenly want some serious acceleration from this beast, not WOT throttle, but an acceleration level that will let you merge safely into that "slot".

The secondary purpose of the design of the torque converter is to provide torque multiplication for just this scenerio. Of course you want the torque converter in "play", as does Toyota and Lexus because even with its 10% power loss it provides enough extra torque, 2 times that which is input, that the fuel economy net is a gain.

But how do we get that lock up clutch to release quickly?

Not only do we wish to use the torque multiplication capability in this instance, but because the lock-up clutch is not designed for high torque applications, it MUST be out of the loop before we allow the engine to build torque.

So the ECU releases the power to the lockup clutch control solenoid and that releases, in turn, the hydraulic pressure that was holding the clutch engaged.

But since the lockup clutch cannot long survive if it is engaged during high engine torque operations the ECU must be certain of lockup release before advancing the throttle servomotor, opening the throttle in response to the gas pedal position.

And if the feedback sensors in the newer models are the same as my 2001 the only method through which the ECU can tell that the lockup clutch has been successfully released is by comparing the torque converter input speed (engine RPM) and the torque converter output speed via its turbine speed sensor.

If these two speeds are not equal then the lockup clutch CANNOT be engaged.

But what if, strictly by happenstance, the lockup clutch is actually released but there is no torque converter input/output speed difference due to roadspeed and the particular gear ratio currently selected?

The engine is slightly above idle, say 1200 RPM, at ~30mph what gear would need to be selected to closely match the engine RPM?

If I were the programmer trying to solve the above problem I would likely add in a timer of enough period to assure that in 99.99% of circumstances the lockup clutch had enough time ( 1 second? 2 seconds?) to fully release before allowing the throttle to open.

The above theory also explains why my 2001 has burned fluid at 38k miles...

The lockup clutch is NOT fully released when the mechanically coupled throttle opens suddenly in exactly the same circumstance as above.

[ColinBarber]

Why speculate, it would be easy to monitor the voltage going to the lock up solenoid to see exactly when lockup occurs.

[wwest]

Confirmation....

You can PREVENT the hesitation symptom by fully releasing the gas pedal during coastdown!

The lockup clutch is disabled, cancelled, if the throttle position is at IDL ( idle) or brakes are applied.

See page 14 of:

http://www.autoshop101.com/forms/h16.pdf

Document reprinted by permission of Toyota.

[ColinBarber]

whilst those docs are good on autoshop remember that most are 10+ years old and therefore not all things would be accurate for newer models.

It is true that lockup is cancelled if the brakes are applied, that is stated in the RX workshop manuals.

[wwest]

The referenced document is an authorized copy of an actual Toyota document.

[SW03ES]

I've been telling people that for years, that you can prevent the hesitation by addressing the gas pedal differently.

[ColinBarber]

The referenced document is an authorized copy of an actual Toyota document.

←

Nobody said it wasn't, it's just that the document is over 10 year old.