Official BRZ MPG Numbers. Big Gear Ratio Differences?

[Jeff Lange]

Basically what it boils down to is that modern automatic transmissions do not have the huge losses that they had in the past, and the computer can select which gear to use for optimal fuel economy much more accurately than a person can without information like catalyst efficiency, A/F ratio, etc.

Yes gear ratios can play a part, but it's not the major contributor in this case.

Jeff

[davesrave]

Question: I don't have the technical understanding, but I wonder if you can predict which transmission will provide the best 0-60 and 1/4 mile times for average driving enthusiasts, taking into account human vs. computer reaction times?

[catharsis]

Originally Posted by davesrave

Question: I don't have the technical understanding, but I wonder if you can predict which transmission will provide the best 0-60 and 1/4 mile times for average driving enthusiasts, taking into account human vs. computer reaction times?

Consistency? the automatic. But for the absolute best numbers, the manual should pull out the best 1/4 mile times.

[davesrave]

Originally Posted by catharsis

Consistency? the automatic. But for the absolute best numbers, the manual should pull out the best 1/4 mile times.

Interesting answer. Never asked the question before because I'm not really that into those numbers. Would the answer be the same if one was auto crossing in the sport mode vs. manual?

[serialk11r]

A knowledgeable driver can shift better than a computer trying to interpret the movement of the gas pedal anyday...where are people getting the idea that a computer knows when to shift better when all it's doing is trying to read your mind?

[davesrave]

I guess it's because we hear the computer can shift in .1 seconds, lockup in 2nd thru 6th, and uses logic that is actually pretty good at predicting driving intentions while in sport mode. That kind of technology, if done right, at least raises the question.

[Dimman]

Originally Posted by arghx7

the thing that I'm concerned about is that very few people have more than a fuzzy idea of how a basic automatic transmission works. An old 3 speed is actually very complicated. If you don't know how that works, it's hard to appreciate the differences between that and a 6 speed. The thread could easily go over everybody's head... if you don't know what simpson gearset is, how a valve body works, how shifting works, what holding elements are and why you need them, what torque converter lock up is, then a lot would go over people's heads. So I would have to figure out what level of detail to provide, and what is a length that people would actually read.

The ones with the hydraulic check-valve mazes in them?

Knowledge is always good, but it's understandable that it would probably be a lot of wasted time on your part.

[serialk11r]

Originally Posted by davesrave

I guess it's because we hear the computer can shift in .1 seconds, lockup in 2nd thru 6th, and uses logic that is actually pretty good at predicting driving intentions while in sport mode. That kind of technology, if done right, at least raises the question.

I guess the most accurate way to put it would be that a driver paying attention to managing engine load and rpm will be able to coax much more fuel economy out of a manual than an auto could hope for. With manual gear selection on autos you can get close though. An automatic can get better economy for the casual driver whose thoughts on driving or fuel economy are limited to "push the gas, the car goes forward", or the manual transmission driver who doesn't bother to shift into the highest gear available.

An automatic transmission in auto mode has the inherent limitation of needing to translate the gas pedal input into commands for both the engine and the gearbox. It also has the inherent limitation of having a hydraulic system which consumes a small amount of power, and a bit more weight to lug around.

[davesrave]

Originally Posted by serialk11r

I guess the most accurate way to put it would be that a driver paying attention to managing engine load and rpm will be able to coax much more fuel economy out of a manual than an auto could hope for. With manual gear selection on autos you can get close though. An automatic can get better economy for the casual driver whose thoughts on driving or fuel economy are limited to "push the gas, the car goes forward", or the manual transmission driver who doesn't bother to shift into the highest gear available.

An automatic transmission in auto mode has the inherent limitation of needing to translate the gas pedal input into commands for both the engine and the gearbox. It also has the inherent limitation of having a hydraulic system which consumes a small amount of power, and a bit more weight to lug around.

I think I get what your saying. If I mimic the way I shift my manual car now on the paddle shifter of the BRZ, flooring it in the current gear, but only down shifting if I need more acceleration, will I get as good gas mileage with the Auto, or will the extra 47 pounds and friction losses of the auto provide poorer mileage than the manual transmission?

I do get better mileage out of my car than the EPA says I should without sacrificing acceleration or high speed highway cruising, and have never understood why. It was even more baffling when the EPA revamped their MPG estimates of my car downward ( I think the Celica went from 23/31 to 22/30 or something like that) and I'm getting 28 to 31 MPG in very mixed and occasionally sporty driving.

Now I'm considering getting the BRZ with an Auto, and using the paddle shifter. If the mileage is better than the manual driving it that way, then I want the auto, I think.

[arghx7]

One thing you have to think about is that a modern automatic transmission can skip gears. They can go from 1st to 4th for example, and with torque converter lockup you don't have a series of tip-out and tip-in as you change gears. Thus there is less transient operation, and the engine can be more efficient that way. Of course a manual can skip gears too, but most people don't drive that way.

I don't know about the transmission calibration on the BRZ so it's hard to say what the shift schedule is though. It could have been tweaked to show favorable results on the EPA test cycles. I would be surprised if you could actually get better mileage on an auto with it in paddle shifting mode.

I can pretty much guarantee you that the auto was not certified with the EPA separately using the paddle shifters. That costs millions of dollars. Usually the manufacturer will try to convince the regulator that most drivers never use the paddle shifters on normal commuting, and thus they get an exemption from having to certify it.

The other thing we don't know for sure is what shift schedule they were using when certifying the manual transmission vehicle.

Originally Posted by serialk11r

A knowledgeable driver can shift better than a computer trying to interpret the movement of the gas pedal anyday...where are people getting the idea that a computer knows when to shift better when all it's doing is trying to read your mind?

It depends on the situation and what you are comparing I suppose. To oversimplify: someone who is really good with a bow and arrow could be more deadly than someone who is ok with a gun, but that doesn't mean the gun isn't overall a better weapon in most situations.

[serialk11r]

Perhaps, I don't know how transmissions are programmed or anything but what I've seemed to notice is if you try to give it more than a tiny bit of throttle on most autos it kicks down a gear or 2. I guess you have a point about the faster shifts.

I have a suspicion that the gas pedal on modern autos does not control the load of the engine directly, but rather the computer interprets it as a demand for a certain level of power and then picks a gear and gives the engine appropriate load to match...is this correct? It would seem that such a strategy would help EPA city mileage a bit.

[Jordo!]

Modern sports car automatics typically offer both faster acceleration and better gas mileage than can be achieved by the average driver in traditional manual. They also offer approximately equal drivetrain losses.

Why?

Full lock up (generally, 2nd gear on) means a solid rather than viscous connection with engine, improved valve body calibration with electronic control means lightning fast shifts that will, on the average, far exceed what a human can do (and on supercars easily exceed without exception what a human could do), and the ability to select and hold any gear that you want, just as you would on a traditional manual, but with less opportunity for error (i.e., selecting the wrong gear).

You lose some specific qualitative experiences (in exchange for another set, BTW) and gain better performance and gas mileage.

If Toyota/Subaru fails to do this with the FR-S/BRZ it will not be for lack of technology but for lack of trying.

Reviews of the auto in action and the fact that the trans will be highly similar to the one in the 370Z and IS-F in design and performance means a lot of guys driving the MT will probably be grinding their teeth when the car is released, but we'll have to wait and see...

Anyway, ignore the brow beating, test drive both the AT and MT if you are unsure, and get the one that suits you.

[Jordo!]

Originally Posted by serialk11r

Perhaps, I don't know how transmissions are programmed or anything but what I've seemed to notice is if you try to give it more than a tiny bit of throttle on most autos it kicks down a gear or 2. I guess you have a point about the faster shifts.

I have a suspicion that the gas pedal on modern autos does not control the load of the engine directly, but rather the computer interprets it as a demand for a certain level of power and then picks a gear and gives the engine appropriate load to match...is this correct? It would seem that such a strategy would help EPA city mileage a bit.

Load is a function of ECU computations that take into account known values based on the VE of the engine, RPM, and MAF (or sometimes MAP). Throttle opening has an impact on this, of course, given that throttle position will affect how much air enters the engine.

[arghx7]

Originally Posted by Jordo!

Load is a function of ECU computations that take into account known values based on the VE of the engine, RPM, and MAF (or sometimes MAP). Throttle opening has an impact on this, of course, given that throttle position will affect how much air enters the engine.

You're on the right track, but that's kind of an outdated way of thinking about it. You are understanding the system as essentially reactive. It hasn't been that way for a long time. The driver or the control modules issue commands. The control modules process the commands in the context of models, then manipulate actuators in response to those commands to achieve targets.

Load is a bit of a nebulous term. There are different ways to calculate it in an onboard controller but they are all based on physics. They are usually related to pumping efficiency (and thus can exceed 100%) or they are normalized to the engine's max output (and thus peak at 100%). You can also use a physical quantity: on a Subaru or a GM it is based on the amount of air entering the cylinder. Other systems will base a load calculation on how much fuel would be required to achieve a stoichiometric air-fuel ratio under a given condition.

Originally Posted by serialk11r

I have a suspicion that the gas pedal on modern autos does not control the load of the engine directly, but rather the computer interprets it as a demand for a certain level of power and then picks a gear and gives the engine appropriate load to match...is this correct? It would seem that such a strategy would help EPA city mileage a bit.

winnar

There are different types of architecture used, but this is exactly right. And it's been that way for 10+ years now. Toyota calls their current architecture the Integrated Drive Power Control System, and it's been around for about 5 years. It is in essence based on achieving target wheel torque and power.

So the old way of electronically controlling a transmission (late 80s through 90s) was that you fed some signals to the transmission control unit and it reacted to rpm, TPS, and temperature inputs for the most part. Then electronic throttle was introduced in the late 90s, and you get an architecture like this:



You've got an accelerator pedal position sensor signal feeding into two control paths, and then compensation algorithms for shifting, traction/stability control, vibration concerns, etc. Well that gets really complicated, especially as you add more control modules. Integrated control provides a more elegant solution:


You utilize models and have the integrated controller coordinate the transmission and engine to achieve the target wheel torque/power. That also feeds into the engine dynamics control (stabilty control etc). The integrated engine controls help reduce vibration and fix the torque drop off between shifts if that's what the calibration calls for. You have to have a model to transfer target drive power into a torque request, and then a model to execute the torque request through actuator commands.



This system was implemented on the AA80E transmission used in the Lexus LS and IS-F and I imagine it has trickled down to just about everything Toyota makes for this market by now.