AT officially 1 sec slower than MT at Tsukuba
 

hopefully this isnt a repost. and even if it is. it deserves its own thread imo. there was some talk in the past about how the AT might actually be faster around a road course due to its very quick shifting. but the latest reports suggest that the gearing of the MT gives it an advantage over the AT. we now have hard numbers to confirm these reports.

[ame="http://www.youtube.com/watch?v=KvyDuLK4pm0&feature=relmfu"]CARトップ　ç*‘波サーã‚*ットテスト2012æ ˜¥ (å‰�ç·¨) - YouTube[/ame]

and some other cars from same test

[ame="http://www.youtube.com/watch?v=DE-Lzl3kI94"]CARトップ　ç*‘波サーã‚*ットテスト2012æ ˜¥ (後編) - YouTube[/ame]


Edit: fixed. looks like this site now has auto embed?

^ FYI - the embedded vids not showing up

The thing I don't understand is why charge $1100 more for the auto and get terrible gearing to the manual in comparison? The auto should've been geared the same or slightly better like how Hyundai did with 6 spd manual vs the 6 spd ZF auto.

im sure there will be shorter final drives available that can negate the effects of the overly tall AT gearing. clutch type LSD with shorter FD would be nice.

From what we can tell, the A/T seems to be the same basic unit as the A960 in the IS250, with identical ratios. The valve body is updated however. So the ratios are the same as the IS to save money it seems.

arghx7 or someone who knows about this subject, is it more expensive for the manufacturer to change AT ratios? I understand they use some sort of planetary gear setup but never really looked into that.

I have a few thoughts on this.

1) Like most things in manufacturing, production volume is a big element in determining costs. The FT-86 platform is intended to be a low-volume vehicle and relatively inexpensive vehicle. I don't think either Subaru or Toyota are expecting to make a lot of money off it, except maybe the fully loaded models.

2) "more expensive" relative to what? Well, is it more expensive to change a design or not change a design? A design change needs testing and validation. It needs a new supplier and/or new tooling for manufacturing the part if it's not an off-the-shelf piece. If the FT-86 is using the main gearsets and transmission architecture right from the IS250 (same ratios and everything), well that saves a ton of money on a low volume, inexpensive vehicle. The manufacturer has to ask, would the customer be willing to pay more money for different gears ratios? The normal customer doesn't see/understand gear ratios. They notice fuel economy, noise/vibration, driving characteristics, and of course price.

3)Is it more expensive to change gear ratios in a step A/T versus a manual? I imagine that's application dependent. Remember that gear ratio = # output teeth / # of input teeth . That's easier to understand on a simpler manual transmission because it doesn't rely on brakes and holding elements. On an automatic transmission this gear ratio calculation isn't always so straightforward because it depends on how the elements of gearset are being operated. You'd have to look through power flows in the transmission to see how everything is working.

I know that's a vague answer but in many ways automatic transmissions are so different from "regular car stuff." You can know a ton of stuff about engines and suspension but very little of that translates to an automatic transmission. Learning about an A/T requires you to start all over again in essence, at least for a lot of people.

Go to autoshop101 technical section http://www.autoshop101.com/autoshop15.html and it has old training documents for Toyota 3 and 4 speeds. You can also read about the powerflow of the A960 6 speed (in the IS250) from this document, http://avtopedia.ru/akpp/a960eaut.pdf , which was linked to in another thread.

Thanks arghx7!

Then what do you think of the manual transmission? Someone posted some article or something describing how the origin of the transmission is not clear because they changed "80%" of the parts. However, it keeps the exact same gear ratios as the late S15.

My question is, since the S15 is quite old, is it still a direct part share of some sort?

I really haven't seen enough about the manual transmission to have formulated many thoughts on it. I'm sure there are parts being shared for it to some extent--Aisin isn't going to design a new transmission family for a low volume car, especially since manuals are becoming less and less popular.

IMO, usually AT encourage to have more drivetrain loss and being heavier so it should be a bit slower than MT. Unless there are something like DCT (like SMG-II or GR6) equipped with the car so I don't see that a regular car with AT could be any quicker.

I love how it sounds when a japanese person says hachiroku lol. All funny business aside i liked how the toyobarus were quicker than more powerful cars!! Tsukuba is a very technical track so this is where the toyobarus will shine!!

It's an off the shelf part from another car that Toyota stocks. Easier to just use that. I'd guess that they expected the MT cars to outsell the AT cars by a large margin (and worldwide that definately seems true so far) making sinking even more money into the AT an even worse idea.

It really has less to do with the weight and drivetrain losses through the torque converter and more about the ratios.

The OP mentioned that the AT would be quicker because of the fast shifts but that's nonsense when you look over the numbers.

Sure, an auto may be a 1/10 of a second faster than the manual in making the shift but for the 3-4 seconds you are in gear the auto is transmitting to the ground 10% to 30% less power.

Wow, that is just total b-shiite. Did you just time-warp in from the 60s? Go read a few of the other threads on the auto and learn something. Gears 2 to 6 - totally locked, virtually no loss relative to manual at same rpm. Any differences virtually entirely due to taller gearing in the AT.

Just because a torque converter is locked does not mean that it is sending the same among of power through the transmission during acceleration.

Sorry, but you are wrong. The last thing you said is correct, but that gearing difference changes the wheel TQ at the ground. This makes what the other guy said correct.

The lower gear ratios of the MT leads to more "wheel torque." I am talking actual wheel torque, which is the engine TQ multiplied by the gearbox, the rear end, etc and measured at the wheels.

Also, the TQ convertor is heavy with fluid compared to a MT clutch and FW, so slight loss there.

Not the wheel torque, but the ratio between the gears. When you shift in the manual transmission you end up at a higher rpm when shifting up, so you have more power available more often.

'Tire thrust'.

?
What I'm getting at is beyond 1st gear (we're talking reasonably geared cars) if you ignore shifting time and optimizing gears for certain speed turns on a track or something, closer ratio gears are what gives you more acceleration, and the absolute gear ratio is not very important.

Tire thrust is a general force that is basically the net of engine output, gearing multiplication, frictional losses and tire diameter. So in your example with the close ratio box we are adjusting the engine output by keeping in a better rpm range. In older autos, they would have much greater frictional losses.

So we are comparing tire thrust really.




And I like typing 'thrust'...

Er yea uh, more power at the same speed = more acceleration...

More thrust.

So you honestly think that the FR-S auto is sending 10-30% less power to the ground with equal gear and final drive ratios, as the "other guy" seemed to be saying? Even with a locked torque converter, which virtually eliminates the drivetrain loss? I'm speechless...

So where is this supposed significant power loss occurring then?

torque converter, hydraulic pump, planetary gearset, etc... :iono:

Compared to the 150-200hp that the engine is putting out, I don't think those items represent a significant parasitic load. If they did there would be more incentives to use dry clutch automatics.

Isn't that why there is a move to twin clutch semi-autos? Even in economy cars like the Fiesta?

The manual will still get more power down (less than the old days though).

tq converter filled with fluid is quite heavy. that's why mazda's skyactiv TC is way smaller. they only use it sparingly. unles toyota/subaru/aisin redesigned the TC to be smaller, i can only assume its still big and filled with fluid...

dry clutch auto? only AMG uses those. and when the tq gets to ludicrous levels like in the G65 AMG, they forgo the dry clutches and go back to the TC...

So the auto is just 1 second slower than the manual driven by a decent driver (whos no doubt pretty handy with his heel/toe action keeping the engine/power down maximised in the manual) despite a little longer gearing!

I can certainly live with that!
I dare say the average driver may not find any difference in lap times?
And even if they are pretty handy behind the wheel is such a small difference noticeable in real world driving? I may track mine once or twice for a laugh but that's all it'd be for.. A laugh.. not trying to break laptime records.

The loss on the FR-S auto has got to be pretty minimal given the big improvement in gas mileage it produces over the manual. It's the gearing that gives it the slower acceleration - the ratio differences in the upper end are very big. There is probably 1/2 second lost at the start due to the non-locked 1st gear as well, can't generate the wheelspin. Be interesting to see the rolling times, 5-60 or 40-70.

:word:

Like I said, less than the old days.

A dyno comparo would be nice to see what (if any) the difference is. The manual has shown that it is quite efficient at putting power down. But no auto dyno so far that I am aware of.

Yup there is still the weight penalty, but I still refuse to believe that the parasitic losses are significant. By dry clutch automatic I was referring to any automatic that uses only a computer controlled clutch to transfer power, so that include DSG, DCTs, etc. I think that there would be far more incentive to use these transmissions if parasitic losses under lockup were a problem.

they cost too much. that's why the TC auto still lives.

and exactly what do you mean by significant? having to turn the heavy TC filled with fluid will sap some power. exactly how much? we wont know until the dynos come out.

I think I forgot that even DCTs use hydraulics to control the clutches anyways lol... so all automatics have parasitic loss there. Long live the manual transmission? :D

What do I mean by significant? I am not sure how much power the hydraulic pump costs but it does not take much power to move some clutches and stuff around so in a setting where you're actually giving the engine 100% load at high rpm, I think it's safe to say the hydraulic pump itself is not doing much to hurt power output. As for the torque converter, I suppose it is still churning the ATF, but the turbine speed and impeller speed are matched under locked conditions, and the losses would be rpm dependent. If say 7000rpm vs 3000rpm is a 10 fold difference in this loss (kinda just cubing the difference), and the torque converter was costing say 10hp, then at 3000 it costs you ~1hp, which for rolling around town/on the highway is a MASSIVE parasitic load and highly unlikely, or we'd be seeing much worse fuel economy numbers. Of course I'm making some very strong assumptions but I would be very surprised if the locked converter were losing more than a few hp.

well.
given the same tires and FD ratio:

3rd gear
1.541 vs 1.404 is around 10% more torque
4th gear
1.213 vs. 1.0 is around 21 % more torque

Whatever the engine is putting out, the gears multiplying it determine the torque where the "rubber hits the road."

Since the MT will more ofter be at a lower gear ratio it is putting down higher torque more of the time. So the quote above is somewhat true (just exaggerates a bit). Even with a super fast AT shift (less time at zero torque), the MT will deliver higher torque wile in gear.

And for more anecdotal evidence. The Toyota/Lexus 2GRs are putting down WAY less power with their (locking) ATs than those swapped into MR2s with MT (and Fidanza FW).

^ I said "with equal gear and final drive ratios". The differences you are talking about are all due to different gear ratios, they are not the result of a differential in the inherent inefficiencies of the 2 transmissions.

That's not necessarily true... In mileage tests modern autos can often outperform their manual counterparts, due partly to their ability to keep the engine operating in it peak efficient specific fuel consumption range. I'm with you though I want to see dyno results and rolling times...

It's almost definitely due to poorer gearing (and maybe a little bit due to the weight penalty), unless not all of the nannies were switched off.

The drivetrain losses should be essentially equivalent on a dyno pull. Hopefully someone will confirm (or refute) this with data soon.

Pity it isn't a 7 speed so the ratios could have been made closer (that's how they did it on the Z!).

Hopefully with tuning, shift speed can be increased (on modern trans like this, no need to change valve body -- line pressure is controlled electronically) and with additional power ups, the difference will shrink.

Anyway, I definitely owe someone on here a beer -- I had thought the AT might be marginally faster than the MT, but that would have only held true if the gearing were closer.

So I stand corrected... I don't mind being proven wrong, but it sucks that I had to be wrong on this issue in particular...

Oh well.

It kind of shows the parts-bin vs ground-up nature of the car. Off the shelf transmissions. Sounds like they tried their best with what they were given but there are limits.

Just looking at peak torque to the wheels comparisons, 1-3 are reasonably close, but 4-6 not so much. I'll try to compare with rpm drop over a dyno graph later...