[Harbir]

Originally Posted by sgiitk

4-ratios in a 3+1 are 4 ratios, except things like engine braking will be available only in the 'normal' part of the box. Since 3 is the highest ratio, it will be there on 2 & 1 only.

In an AT when driven normally there is no engine braking, the wheels freewheel if the engine speed is lower than what corresponds to the road speed. This is done to conserve fuel.

Are you guessing based on what the behaviour of the car seems to be or do you know this to be mechanically factual?

Here is why I ask. ATs that do not have clutches NEVER freewheel unless put in neutral. By freewheeling I mean that the wheels are disconnected from the engine.

But a car can SEEM to be freewheeling if it is in a high gear and the torque converter has a lot of slack in it. the wheels will speed up, and along with them the output shaft of the torque converter will speed up, but with the torque converter slack (rubber band), the input shaft doesn't accelerate as much, and with a tall ratio, the amount of back torque put on the engine is minimal, creating the absence of significant engine braking, and the FEELING that the car is free wheeling.

As I said before, engine braking in an AT is only a question of gear ratio and torque converter slack.

If what you are saying is correct, I would appreciate it if you could explain what you mean by the "the 'normal' part of the box" and WHY (in mechanical terms) engine braking is not available outside the "normal" part. I would also like to understand what are the mechanics by which a "true" 4 speed will provide engine braking in 3rd, and what are the mechanics that cause the 3rd in a 3+1 to not provide engine braking.

[sgiitk]

Originally Posted by Harbir

Are you guessing based on what the behaviour of the car seems to be or do you know this to be mechanically factual?

I have experience of three cars - Maruti Zen 1997, -3-speed AT, Hyundai Santro 2005 and Honda Civic 2010. I own the last two, and owned the first from 1999 to 2005!

So it is factual, also the incident from France I have stated. What can be more factual that having your brakes fade away totally, putting life & limb at risk.

Normal part of the box => the basic 3-speed part in a 3+1.

[Harbir]

I mean no offense, but there is nothing in your anecdotal evidence that indicates that the issue of engine braking is due in fact to the distinction you draw between a "true 4 speed" and a "3+1 AT", rather than about the gearing ratios and the behaviour of the torque converters in those cars.

If what you are saying is correct, there ought to be a mechanical engineering explanation for it, not just conjecture of what is happening based on what you observed in 2 or 3 or 4 individual cars.

[sgiitk]

Originally Posted by Harbir

I mean no offense, but there is nothing in your anecdotal evidence that indicates that the issue of engine braking is due in fact to the distinction you draw between a "true 4 speed" and a "3+1 AT", rather than about the gearing ratios and the behaviour of the torque converters in those cars.

If what you are saying is correct, there ought to be a mechanical engineering explanation for it, not just conjecture of what is happening based on what you observed in 2 or 3 or 4 individual cars.

If 13 years+ experience and only good fortune saving us from a major accident is anecdotal, the so be it.

The Civic manual also states explicitly that to get engine braking you must be in the S mode. Here the S mode automatically only cycles to 4 (no 5). I know from experience that I get reverse braking in my Santro in 2 or 1, and not in 3 even if I switch off OD!

Also, most cars (all afaik) do not provide for reverse braking in the highest ratio.

It is a matter of mechanical design and programming the ECU. Very likely when you go with OD on then the ECU switches in the 4th gear by actuating some EM components. The Mech portion is not designed to provide reverse braking in 3rd. This is exactly what I mean by the 3+1 not being a full 4-speed. I will not expect reverse braking in 3 either given the above.

[Harbir]

I just continue to not see what is the actual mechanical difference in a "true 4AT" and a "3+1 AT" that causes the differences in behaviour you claim.

But I'll drop this now since we seem to be making no progress.

thanks for the discussion!

[sarathlal]

Originally Posted by sgiitk

If 13 years+ experience and only good fortune saving us from a major accident is anecdotal, the so be it.

The Civic manual also states explicitly that to get engine braking you must be in the S mode. Here the S mode automatically only cycles to 4 (no 5). I know from experience that I get reverse braking in my Santro in 2 or 1, and not in 3 even if I switch off OD!

Looks like OT, but still confused over the 4 speed automatic comment. Is the absence of D4 that prompts you to the conclusion of Brio being a 4 speed CVT? Please let know if it is something else.

Regarding Civic on S, once you use the paddle shifters, its more or less fully manual. It will only downshift automatically and automatic up shift is restricted to 3rd gear. To go to 4 and 5, you have to use the paddle shift.

[sgiitk]

Originally Posted by sarathlal

Looks like OT, but still confused over the 4 speed automatic comment. Is the absence of D4 that prompts you to the conclusion of Brio being a 4 speed CVT? Please let know if it is something else.

See the shifter is D, D3, 2, 1. D is four speed and the rest obvious. Wait for another fortnight at most and all will be revealed. You may take D3 as S in the Civic, so remove the highest gear and you are left with another three.

Remember I also have a Civic AT since May 2010! I was quite aware of 5 being available on the Paddle but that was irrelevant to the discussion. That is a far more advanced AT than the CVT we are talking about. In fact D3, 2, 1 is more than I expected.

[sarathlal]

Originally Posted by sgiitk

See the shifter is D, D3, 2, 1. D is four speed and the rest obvious. Wait for another fortnight at most and all will be revealed. You may take D3 as S in the Civic, so remove the highest gear and you are left with another three.

Remember I also have a Civic AT since May 2010! I was quite aware of 5 being available on the Paddle but that was irrelevant to the discussion. That is a far more advanced AT than the CVT we are talking about. In fact D3, 2, 1 is more than I expected.

Oh ok, yea, as you said, in another few days we will get to know the detailed spec. Hope you dont mind asking this n number of times, but it is the urge to know how you were able to correctly identify it as a 4 speeder, that prompted me to keep asking.
Automobiles is one such intriguing subject

[sgiitk]

Originally Posted by sarathlal

Hope you dont mind asking this n number of times, but it is the urge to know how you were able to correctly identify it as a 4 speeder, that prompted me to keep asking.

I think you have the answer, D, D3, 2 and 1 make it pretty obvious. It is not rocket science.

If there is a 5 or something lurking then it will be a surprise. If it was more than 5 - speeds (6 or 7) then you would have a tiptronic type notch with + and -. I will not expect a paddle shifter at this price point. Also, I doubt whether the engine can support higher ratios.

[the.city]

Finally!! An interesting discussion worth joining here..

Originally Posted by sgiitk

In an AT when driven normally there is no engine braking, the wheels freewheel if the engine speed is lower than what corresponds to the road speed.

Originally Posted by Harbir

Here is why I ask. ATs that do not have clutches NEVER freewheel unless put in neutral. By freewheeling I mean that the wheels are disconnected from the engine.

I was taken aback when I read that first statement, but the more I think about it, the more I feel Mr.SG is right.

Looking purely from a driver's perspective-- When I drive any manual gear - be it bike or car - there's a sudden engine-induced slowdown/brake EVERYTIME I let go of gas. It's so much more evident on my Pulsar, a little less on my dad's i10 - but it's always there. As soon as I let go of gas, the engine RPM zaps down and the vehicle gets a minor jerk-ish reduction in speed.

But, in last 6+yrs of driving automatic I've never experienced that jerk. Of course, there's a "loss of acceleration" which could be interpreted as a jerk, but that's not what I'm talking about. As soon as I lift my leg from gas on my City AT the car goes into a smooth zone - the rpm goes down, the "live mileage" indicator goes to 40kmpl (which can be so misleading to some), and no jerk. The vehicle is indeed freewheeling at that point. May be, I shouldn't call it freewheeling. More like - Computer-Controlled-Temporary-Disconnect.

Off to Google to research more on this....


EDIT: I couldn't find much, but whatever I read points to what Harbir wrote. There's no real disconnect. The jerk should always be there, but it would be too small to notice because of higher gear ratio. The engine braking phenomenon is always there - both 4 and 3+1 - just less noticeable for higher gears. The important thing to understand here is that all these automatics don't have true mechanical couplings - they are hydraulic couplings (also called torque convertors). So, if I understand right, the transmission of torque is not same in both directions. Engine-to-wheel transmission would be different from wheel-to-engine transmission (= what we need for engine braking) due to higher gear ratio.

Again, off to Google to verify ...

[Harbir]

the absence of the "jerk" is because of the torque converter. I think lots of people don't understand what exactly that is.

in a manual transmission, when a car the car is in gear and the clutch pedal is not pushed in, there is a direct connection from the engine's crankshaft to the contact patch of the driven tyres. If you release the throttle, the loss of driving force is IMMEDIATELY transferred through through the transmission to tires which slow down immediately in a proportion that is DIRECTLY related to the ratio of the gear the engine is in, and to a lesser extent, the engines compression characteristics.

Torque converter automatics will not behave like this (unless the torque converter is capable of full lock up which I will describe later). To understand this you have to understand what a torque converter is and how it works. The first thing you have to understand is that there is NO DIRECT MECHANICAL CONNECTION between engine and transmission.

A torque converter is a device that allows a vehicle to come to a complete stop while in gear without turning off the engine. The input shaft is connected to an impeller or pump. On the output shaft ( which feeds the transmission) is mounted the turbine. Both the impeller and the turbine are in a bath of fluid. They are NOT directly connected to each other.

As the engine turns the pump, the pump forces fluid into the turbine. This causes the turbine to turn and thus cause the transmission to turn. This is known as a fluid coupling.

This torque converter is responsible for a number of characteristics of automatic transmissions.

1. Since there is no direct mechanical connection between the engine and transmission, it is possible for the car to be in gear while stopped while the engine runs. In this case, the pump continues to spin and force fluid at the turbine, but the turbine is held stationary by the vehicle's brakes. The brakes prevent the wheels from turning, which prevents the transmission from turning, which prevents the turbine from turning.

In this situation, the engine faces a drag from all the fluid churning around the stationary turbine. It is also what causes the car to start creeping forward if pressure on the brakes is reduced or released. The turbine starts to turn slowly, turning the transmission and the wheels.

2. When the car is travelling at a steady speed, the pump and turbine reach equilibrium speeds. The turbine will still turn a little slower than the pump due to the loss of energy into the fluid so the engine will be turning a little faster than it would at the same speed and gear ratio in a manual transmission car.

3. If the driver travelling at steady speed suddenly pressed the accelerator down, a lag develops. The engine speeds up along with the pump, but turbine does not accelerate at the same rate beause of the fluid coupling. It takes time for the turbine's speed to catch up. This is experienced by the driver as a lag in the response of the vehicle. THe RPM climb but the vehicle does not accelerate in proportion. The taller the ratio of the gear the transmission is in, the bigger the lag. This is the "rubber band" or "slush" effect of the torque converter. It is this behaviour that is so offensive to driving enthusiasts in the west, and the reason torque converter automatics are contemptuously refered to as "slushboxes".

4. A similar thing happens in reverse when the driver lifts off the throttle from a steady speed but with an additional factor in play. When the driver lifts off the throttle, the engine slows down and the pump slows down, but there is no way for the turbine to slow down in direct response. The momentum of the car causes the turbine to keep spinning. It churns up the fluid as the speed differential increases between the pump and turbine. There is an additional factor. the pump is more efficient when driving rather than when being driven by fluid. THe turbine is more efficient when it is driven rather than when it is driving fluid. THis means that when the driver lifts off, and the momentum of the car wants to make it keep going, the turbine cannot transfer the energy through the fluid back to the pump and thus the engine very efficiently. This means the engine RPM and pump speed drops but the turbine keeps spinning. This is why automatics do not show a jerk when you lift off and why they have weak engine braking.

5. Automatics produce strong engine braking at low ratios because the low ratios cause the transmission to REALLY spin fast compared to the vehicle's speed. Such a huge differential in the speed of the pump and the turbine causes a large enough drag to start delivering a strong reverse torque on the engine, producing engine braking.

This means that the idea of "3+1" transmissions vs "true 4 speed" is complete nonsense. If a 4 speed automatic feels like its not producing engine braking in 3rd and 4th gears, it only means that the ratios are too tall to produce adequate back torque in their torque converters and/or the torque converter is poor at tranfering back torque to the engine. One torque converter design may be very efficient at transfering back torque to the engine from the transmission, another one might not. If a converter is not very efficient at transfering back torque, and is already getting only a small amount of back torque from the transmission due to the gear ratio, the amount of back torque that reaches the crankshaft would be so small as to not even be felt to exist when you're really in need of it.

A good torque converter will be efficient at responding quickly to changes in torque levels and direction, but it will also be not as smooth, and may in fact be down right jerky, so those kinds of TQs are rare. After all, anybody who wants a really good performance in this regard should choose a manual transmission (or now a DCT as well).

The exception here are torque converters that can achieve full lock up. These are sophisticated units that can engage pawls to mechanically lock the pump and turbine together. The computer determines when to lock and unlock the torque converter, allowing the transmission to eliminate the rubberband or slushbox effect under many circumstances.

[the.city]

Perfect. I was thinking of writing a big explanation. I don't think I could have done a better job than you.

The idea that it's not a mechanical coupling is the key. This translates to the fact that overdrive gears are as bad in engine braking as any other higher gear in "true" 4-speeds (as it was christened earlier in the thread).

It appears to me that the only reason overdrive is being associated with engine braking is because a "button enabled" overdrive has been present in several (mostly older) automatic models. It was always suggested to disable overdrive when going uphill/downhill. Add to it, the fact that overdrive was/is physically a different piece of hardware on several models (to avoid investing money into building a whole new gearbox). Considering both things together, we've evolved an opinion that overdrive gear behavior "something different". We're thinking that turning off the overdrive button is the only effective way to introduce engine braking - In other words: Engine braking will not be effective in the overdrive mode / Engine braking is not available on the 3rd or 4th or whatever gear of 3+1/ gearbox.

With current technology and ubiquity of overdrives, the behavior is no different. The mechanical response & functionality is the same. 3+1 or 4, both are same for all practical purposes today. The engine braking will be equally bad on higher gears of both cases.

[Akshay1234]

Originally Posted by Harbir

The exception here are torque converters that can achieve full lock up. These are sophisticated units that can engage pawls to mechanically lock the pump and turbine together. The computer determines when to lock and unlock the torque converter, allowing the transmission to eliminate the rubberband or slushbox effect under many circumstances.

That explanation was a good read. But now tell me why this happens. Comparing my Accord with the 3 series, both are automatics, and both have lockable torque converters. In the Accord whenever I lift of the throttle, the the rpm drops down to 1500 odd, not idling speed. And when I press the accel again it goes back upto the normal rpm vis-a-vis the speed and gear.

Now in the Bimmer, even when my foot is off the throttle, the rpm is steady at what is should be according to the corresponding speed. And this is not only in sports mode, but in normal mode too.

I understand both torque converters are lockable, so does this have to do with just the way the transmissions are programmed or is it something else?

[Mpower]

Originally Posted by Harbir

The exception here are torque converters that can achieve full lock up. These are sophisticated units that can engage pawls to mechanically lock the pump and turbine together.

A wet clutch is used for locking, almost universally as opposed to a pawl. (pawl is actually is used for parking)

[Harbir]

Originally Posted by akshay1234

That explanation was a good read. But now tell me why this happens. Comparing my Accord with the 3 series, both are automatics, and both have lockable torque converters. In the Accord whenever I lift of the throttle, the the rpm drops down to 1500 odd, not idling speed. And when I press the accel again it goes back upto the normal rpm vis-a-vis the speed and gear.

Now in the Bimmer, even when my foot is off the throttle, the rpm is steady at what is should be according to the corresponding speed. And this is not only in sports mode, but in normal mode too.

I understand both torque converters are lockable, so does this have to do with just the way the transmissions are programmed or is it something else?

I am not an AT expert by any means, but I would conjecture that in the accord, the torque converter is unlocking (by design or failure) in the situation you describe but the BMW's is not. You have to keep in mind that the locking converters are unlocked a portion of the time. The slushy effect is often apparent in cars with locking TQs.

Most good cars have locking torque converters now, but cars at the bottom of the range (such as i10 AT) may not due to the demand for simplicity and low cost.

Originally Posted by Mpower

A wet clutch is used for locking, almost universally as opposed to a pawl. (which actually is used in P)

Wet clutches sound like the right solution. I was speaking from an article around 1990 with an illustration of a mercedes torque converter that did use pawls to lock the impeller and turbine together. I had a nagging feeling that pawls would be tricky because of the speed differential and the precision required, but since I have mostly contempt for slushboxes for my own use, I didn't bother to do any further research. Thanks very much!