Subaru Ascent Forum banner
1 - 20 of 71 Posts

·
Registered
Joined
·
31 Posts
Discussion Starter · #1 ·
I don't normally post on auto forums, and especially don't start threads, mainly because of the propensity for discussions to get way off in the weeds. But have done lots of thinking about the CVT on the Ascent especially regarding the "fake shift points", and want to see what all you think. This may be really long, so hang on.

First, some background, definitely not to intimidate or think I'm smarter than any of you, but to lend some insight into where these theories are coming from. I'm a 20 year practicing mechanical design engineer, that has been working on cars for 35 or so. I have lots of background in control systems, and design of power transmission devices, but NOT automotive. So take my insight with that grain of salt.

My theory is the fake shift points in our transmissions are a necessity of the engine power. Lets go back to how a typical transmission works. There are gears, of varying size, one side attached to the torque converter/clutch/engine, and the other side to the wheels. The gears are clutched/banded into each other, with positive engagement 99% of the time. This means everything is so locked down that there is no slippage in normal operation. Clutches and bands in normal transmissions are sized to not slip at maximum engine torque on a stock car. (Yes, when we up the torque of an engine, we have to sometimes upsize clutches, pressures in auto trannys etc.) When gears switch in a "normal" transmission, manual or automatic, the engine power backs off. In a manual transmission, we let off the gas a bit and declutch when shifting, in an (modern) automatic trans, the engine backs off the spark and fuel for a brief few milliseconds to enable smooth shifts.

Now along comes the CVT with it's "infinite" ratios. This is achieved by pushing, or pulling (ours) a metal chain/belt hybrid between two cones. These cones vary in width by hydraulic force to change "gear" ratios. Smaller (farther apart) on the input pulley, and larger (closer) on the output pulley gives a higher (taller or numerically lower) ratio. Hydraulic pressure, thus cone width is controlled by a computer that is programmed to decide the "best" ratio given the situation.

(warning lots of physics/engineering begins here)
The nature of the CVT design is that the belt relies only on friction to connect the two cones. And at that, some of the time its sliding friction. The belt/cone interface is only static friction at a very small line contact for a brief time. Therefore, to keep the proper friction force, the hydraulic pressure has to be maintained pinching the belt as tightly as possible, but also still not too tight or noise and efficiency will go down. (pinch too tight for the situation and the pulley doesn't want to let go of the belt creating noise and excess heat) In my reading up on this, the computer does have knowledge of engine load demand and varies pressure in the hydraulics based on this.

Now the super tricky part of a CVT: CHANGING RATIOS! To change ratios, the two pulleys must move in a coordinated manner, one farther apart, and one closer together. Here's a link to a simple animation on how this looks. https://commons.wikimedia.org/wiki/File:GearBoxRotRotVar.gif#/media/File:GearBoxRotRotVar.gif

In order to smoothly change, the friction forces must be low enough that the belt can transmit power to your wheels as you're accelerating the car, AND slide "up and down" on the pulley SIMULTANEOUSLY. In order to not feel/hear "chain noises" (as stated in the owner's manual can happen) the pulleys pinch a little less.

So finally the conclusion of all of the above! When the engine torque on our Ascent is low, 0-50% or so throttle, the power output of the engine is low enough to allow the "normal" CVT feel, smooth as butter ratio changes -- and little or no engine RPM variation as acceleration happens. However, over a certain power, this would result in belt slippage therefore the computer knows this, and briefly: backs off the engine torque, changes ratios ASAP, and reinstates requested torque. It acts like a traditional manual transmission. We feel this and perceive this as the "fake" shifts.

What all of this means? I think it means the engineers would have rather been able to smooth shift the CVT throughout the power band. But given the high engine output needed to move our big Subies, and given the current design of CVT, they couldn't handle the torque any other way.

TLDR -- My theory is the "fake shifts" are due to the high engine output/high weight of the Ascent, and the engineers were forced to add "fake shifts" when we really get on the power to avoid belt slippage and improve CVT life. The engine backs off power, changes ratios, and turns the power back on in order to keep our CVTs alive.

Discuss.
 

·
Registered
2020 Ascent Touring
Joined
·
1,553 Posts
Interesting. I'm just not convinced that the Ascent is so powerful to require the shifts to save the transition. As shown by the CVT recall, the pressure balance between the variators can cause chain slap that can break the chain guide. The more shifting the more likely to get a "mis shift". Aren't there infiniti's with more powerful engines that don't take shift.
 

·
Registered
Joined
·
31 Posts
Discussion Starter · #3 ·
I think you're right on both counts Titanrx8, Infiniti's are more powerful: a quick google shows the QX60 (295hp) compared to the Ascent (265), but I think we really have to talk torque. As 245ft-lbs, of torque, which is less than our 277ft-lbs. (13% more torque) That is the force we are really trying to resist with friction.

On the chain slap thing, absolutely 100% agree with you. Dynamics of something like a moving chain are hard -- very hard, to simulate, and it seems that Subaru got it wrong under certain drive regimes.

Interesting. I'm just not convinced that the Ascent is so powerful to require the shifts to save the transition. As shown by the CVT recall, the pressure balance between the variators can cause chain slap that can break the chain guide. The more shifting the more likely to get a "mis shift". Aren't there infiniti's with more powerful engines that don't take shift.
 

·
Registered
2019 Ascent Limited, 2015 WRX, 2022 OB Onyx
Joined
·
4,935 Posts
I don't normally post on auto forums, and especially don't start threads, mainly because of the propensity for discussions to get way off in the weeds. But have done lots of thinking about the CVT on the Ascent especially regarding the "fake shift points", and want to see what all you think. This may be really long, so hang on.

First, some background, definitely not to intimidate or think I'm smarter than any of you, but to lend some insight into where these theories are coming from. I'm a 20 year practicing mechanical design engineer, that has been working on cars for 35 or so. I have lots of background in control systems, and design of power transmission devices, but NOT automotive. So take my insight with that grain of salt.

My theory is the fake shift points in our transmissions are a necessity of the engine power. Lets go back to how a typical transmission works. There are gears, of varying size, one side attached to the torque converter/clutch/engine, and the other side to the wheels. The gears are clutched/banded into each other, with positive engagement 99% of the time. This means everything is so locked down that there is no slippage in normal operation. Clutches and bands in normal transmissions are sized to not slip at maximum engine torque on a stock car. (Yes, when we up the torque of an engine, we have to sometimes upsize clutches, pressures in auto trannys etc.) When gears switch in a "normal" transmission, manual or automatic, the engine power backs off. In a manual transmission, we let off the gas a bit and declutch when shifting, in an (modern) automatic trans, the engine backs off the spark and fuel for a brief few milliseconds to enable smooth shifts.

Now along comes the CVT with it's "infinite" ratios. This is achieved by pushing, or pulling (ours) a metal chain/belt hybrid between two cones. These cones vary in width by hydraulic force to change "gear" ratios. Smaller (farther apart) on the input pulley, and larger (closer) on the output pulley gives a higher (taller or numerically lower) ratio. Hydraulic pressure, thus cone width is controlled by a computer that is programmed to decide the "best" ratio given the situation.

(warning lots of physics/engineering begins here)
The nature of the CVT design is that the belt relies only on friction to connect the two cones. And at that, some of the time its sliding friction. The belt/cone interface is only static friction at a very small line contact for a brief time. Therefore, to keep the proper friction force, the hydraulic pressure has to be maintained pinching the belt as tightly as possible, but also still not too tight or noise and efficiency will go down. (pinch too tight for the situation and the pulley doesn't want to let go of the belt creating noise and excess heat) In my reading up on this, the computer does have knowledge of engine load demand and varies pressure in the hydraulics based on this.

Now the super tricky part of a CVT: CHANGING RATIOS! To change ratios, the two pulleys must move in a coordinated manner, one farther apart, and one closer together. Here's a link to a simple animation on how this looks. https://commons.wikimedia.org/wiki/File:GearBoxRotRotVar.gif#/media/File:GearBoxRotRotVar.gif

In order to smoothly change, the friction forces must be low enough that the belt can transmit power to your wheels as you're accelerating the car, AND slide "up and down" on the pulley SIMULTANEOUSLY. In order to not feel/hear "chain noises" (as stated in the owner's manual can happen) the pulleys pinch a little less.

So finally the conclusion of all of the above! When the engine torque on our Ascent is low, 0-50% or so throttle, the power output of the engine is low enough to allow the "normal" CVT feel, smooth as butter ratio changes -- and little or no engine RPM variation as acceleration happens. However, over a certain power, this would result in belt slippage therefore the computer knows this, and briefly: backs off the engine torque, changes ratios ASAP, and reinstates requested torque. It acts like a traditional manual transmission. We feel this and perceive this as the "fake" shifts.

What all of this means? I think it means the engineers would have rather been able to smooth shift the CVT throughout the power band. But given the high engine output needed to move our big Subies, and given the current design of CVT, they couldn't handle the torque any other way.

TLDR -- My theory is the "fake shifts" are due to the high engine output/high weight of the Ascent, and the engineers were forced to add "fake shifts" when we really get on the power to avoid belt slippage and improve CVT life. The engine backs off power, changes ratios, and turns the power back on in order to keep our CVTs alive.

Discuss.
maybe that is why the CVT performance is better when implementing smooth acceleration rather than stomping on the accelerator?
 

·
Registered
2019 Subaru Ascent Limited 2010 Subaru Legacy GT limited
Joined
·
976 Posts
I think it was done to fool consumers into feeling like a traditional transmission. The cvt got a bad reputation with the first variations of it especially the subaru ones, deservedly so. Better to make the consumer believe what they feel rather than what is on the window sticker. I am pretty sure there are market studies pointing this out.

 

·
Registered
Joined
·
31 Posts
Discussion Starter · #8 ·
I’ve seen this explanation bandied around, but the part that’s never made sense to me is why it only happens above certain throttle points. Give it say 35% throttle and you have a nice smooth “CVT” like acceleration. One would think if that’s the point, it would always happen. Right?

I think it was done to fool consumers into feeling like a traditional transmission. The cvt got a bad reputation with the first variations of it especially the subaru ones, deservedly so. Better to make the consumer believe what they feel rather than what is on the window sticker. I am pretty sure there are market studies pointing this out.

 

·
Super Moderator
2019 Ascent Touring (CWP)
Joined
·
5,871 Posts
I do not really believe there is any reason that the powertrain in the Ascent "needs" the fake shifts...and as @Robert.Mauro has mentioned numerous times, I believe that they exist because of the company caving to reviewers like Car and Driver who, if they had their way, would have the fake shifts even more extreme and noticeable. The fake shift points go directly against the whole benefit of having a CVT, IMHO.
 

·
Registered
2022 Ascent Onyx, Ice Silver
Joined
·
671 Posts
Yes. I saw one Youtube review/road test where the driver brake boosted the car and the 0-60 time dropped from 7 seconds to 6.2, and the faux shifts were eliminated, with the car going into pure CVT mode. This seems the opposite of the OP's theory.
Only Subaru got Shifts in all cars , not just ascent …, and if you brake boost engine it will skip fake shifts .. from 0-60+
 

·
Registered
2019 Subaru Ascent Limited 2010 Subaru Legacy GT limited
Joined
·
976 Posts
My in-law’s 2020 Outback and a loaner 2019 Impreza I’ve had do not have the fake shift points.
Under certain circumstances neither does the Ascent. The previous heavy duty transmission the 5eat had nearly imperceptible shifts until you got into the throttle. All the recent Subarus I have seen have the paddle shifters on the steering wheel, so they do have simulated gears, that doesn't mean you can always feel the "shift" or that it always behaves the same way, at times during towing our Ascent behaves like a CVT should other times it seems to struggle to eek out a fake shift point. The programming really leaves a lot to be desired.
 

·
Administrator
Joined
·
12,749 Posts
My theory is the fake shift points in our transmissions are a necessity of the engine power. Lets go back to how a typical transmission works. There are gears, of varying size, one side attached to the torque converter/clutch/engine, and the other side to the wheels. The gears are clutched/banded into each other, with positive engagement 99% of the time. This means everything is so locked down that there is no slippage in normal operation. Clutches and bands in normal transmissions are sized to not slip at maximum engine torque on a stock car. (Yes, when we up the torque of an engine, we have to sometimes upsize clutches, pressures in auto trannys etc.) When gears switch in a "normal" transmission, manual or automatic, the engine power backs off. In a manual transmission, we let off the gas a bit and declutch when shifting, in an (modern) automatic trans, the engine backs off the spark and fuel for a brief few milliseconds to enable smooth shifts.
No.

The fake shift points are in our transmissions as a necessity to make Car & Driver stop crapping on every Subaru with a CVT. And yet, still, they complain it doesn't shift harsh enough or frequently enough (seriously, 2021 review).

You can torque launch your Ascent, and the CVT will go into "torque launch" mode, bypass the fake shift programming and operate entirely as a CVT for the full acceleration, while applying maximum torque from ZERO miles per hour to (in the Ascent's case) the 80's or so.
 

·
Administrator
Joined
·
12,749 Posts
My in-law’s 2020 Outback and a loaner 2019 Impreza I’ve had do not have the fake shift points.
Yes, they absolutely do. BUT, they are TR580's that shift "smoother" and shift under medium load (and shift less or not at all under high load).

ALL TR580's ever built have fake shifting programmed into them.

ALL TR690's from 2013ish have fake shifting programmed into them.
 

·
Registered
Joined
·
31 Posts
Discussion Starter · #16 ·
I totally respect your opinion about C&D Robert and will accept that seems to be they’re programmed in for a reason. However by bowing to these journalist’s opinion, they’ve completely destroyed my wife’s opinion of the car. She wants to get rid of it purely because the transmission “feels cheap when it clunks”.

She wants her 5speed auto from her acura back…. Which, because I still daily drive it, can tell you clunks much more than the fake shifts in the ascent.

No.

The fake shift points are in our transmissions as a necessity to make Car & Driver stop crapping on every Subaru with a CVT. And yet, still, they complain it doesn't shift harsh enough or frequently enough (seriously, 2021 review).

You can torque launch your Ascent, and the CVT will go into "torque launch" more, bypass the fake shift programming and operate entirely as a CVT for the full acceleration, while applying maximum torque from ZERO miles per hour to (in the Ascent's case) the 80's or so.
 

·
Registered
Joined
·
31 Posts
Discussion Starter · #17 ·
Not my experience driving them at all. Sure using the manual mode, but never in “D”. And I’ve tried to reproduce it just to test my theory.

Where are you getting this data?

2019 Impreza is 7 shift points
2020 Outback 7 and 8 depending on the engine and CVT
but depending on how you drive you can skip it .. so is Ascent
 

·
Registered
Just ordered a '23 Ascent Touring
Joined
·
66 Posts
I don't normally post on auto forums, and especially don't start threads, mainly because of the propensity for discussions to get way off in the weeds. But have done lots of thinking about the CVT on the Ascent especially regarding the "fake shift points", and want to see what all you think. This may be really long, so hang on.

First, some background, definitely not to intimidate or think I'm smarter than any of you, but to lend some insight into where these theories are coming from. I'm a 20 year practicing mechanical design engineer, that has been working on cars for 35 or so. I have lots of background in control systems, and design of power transmission devices, but NOT automotive. So take my insight with that grain of salt.

My theory is the fake shift points in our transmissions are a necessity of the engine power. Lets go back to how a typical transmission works. There are gears, of varying size, one side attached to the torque converter/clutch/engine, and the other side to the wheels. The gears are clutched/banded into each other, with positive engagement 99% of the time. This means everything is so locked down that there is no slippage in normal operation. Clutches and bands in normal transmissions are sized to not slip at maximum engine torque on a stock car. (Yes, when we up the torque of an engine, we have to sometimes upsize clutches, pressures in auto trannys etc.) When gears switch in a "normal" transmission, manual or automatic, the engine power backs off. In a manual transmission, we let off the gas a bit and declutch when shifting, in an (modern) automatic trans, the engine backs off the spark and fuel for a brief few milliseconds to enable smooth shifts.

Now along comes the CVT with it's "infinite" ratios. This is achieved by pushing, or pulling (ours) a metal chain/belt hybrid between two cones. These cones vary in width by hydraulic force to change "gear" ratios. Smaller (farther apart) on the input pulley, and larger (closer) on the output pulley gives a higher (taller or numerically lower) ratio. Hydraulic pressure, thus cone width is controlled by a computer that is programmed to decide the "best" ratio given the situation.

(warning lots of physics/engineering begins here)
The nature of the CVT design is that the belt relies only on friction to connect the two cones. And at that, some of the time its sliding friction. The belt/cone interface is only static friction at a very small line contact for a brief time. Therefore, to keep the proper friction force, the hydraulic pressure has to be maintained pinching the belt as tightly as possible, but also still not too tight or noise and efficiency will go down. (pinch too tight for the situation and the pulley doesn't want to let go of the belt creating noise and excess heat) In my reading up on this, the computer does have knowledge of engine load demand and varies pressure in the hydraulics based on this.

Now the super tricky part of a CVT: CHANGING RATIOS! To change ratios, the two pulleys must move in a coordinated manner, one farther apart, and one closer together. Here's a link to a simple animation on how this looks. https://commons.wikimedia.org/wiki/File:GearBoxRotRotVar.gif#/media/File:GearBoxRotRotVar.gif

In order to smoothly change, the friction forces must be low enough that the belt can transmit power to your wheels as you're accelerating the car, AND slide "up and down" on the pulley SIMULTANEOUSLY. In order to not feel/hear "chain noises" (as stated in the owner's manual can happen) the pulleys pinch a little less.

So finally the conclusion of all of the above! When the engine torque on our Ascent is low, 0-50% or so throttle, the power output of the engine is low enough to allow the "normal" CVT feel, smooth as butter ratio changes -- and little or no engine RPM variation as acceleration happens. However, over a certain power, this would result in belt slippage therefore the computer knows this, and briefly: backs off the engine torque, changes ratios ASAP, and reinstates requested torque. It acts like a traditional manual transmission. We feel this and perceive this as the "fake" shifts.

What all of this means? I think it means the engineers would have rather been able to smooth shift the CVT throughout the power band. But given the high engine output needed to move our big Subies, and given the current design of CVT, they couldn't handle the torque any other way.

TLDR -- My theory is the "fake shifts" are due to the high engine output/high weight of the Ascent, and the engineers were forced to add "fake shifts" when we really get on the power to avoid belt slippage and improve CVT life. The engine backs off power, changes ratios, and turns the power back on in order to keep our CVTs alive.

Discuss.
The General Motors Dynaflow was completely stepless but, the Buick I-8 and V8 had gobs more power than even a turbo Soob, and moving a lot more weight. The Dynaflow even went into a few Cadillacs in about 1955 or so when the HydraMatic plant had a serious fire. The Caddie V8s of the time were well into mid 300 horsepower range. Yet, they stayed in one piece,
I do not really believe there is any reason that the powertrain in the Ascent "needs" the fake shifts...and as @Robert.Mauro has mentioned numerous times, I believe that they exist because of the company caving to reviewers like Car and Driver who, if they had their way, would have the fake shifts even more extreme and noticeable. The fake shift points go directly against the whole benefit of having a CVT, IMHO.
and given the technology of the time, worked really well. So, I tend to think the fake shift points are there to give Americans a sense of what they are used to. I single out Americans because to this day, automatics remain the exception, rather than the rule, in Europe and Asia.
No.

The fake shift points are in our transmissions as a necessity to make Car & Driver stop crapping on every Subaru with a CVT. And yet, still, they complain it doesn't shift harsh enough or frequently enough (seriously, 2021 review).

You can torque launch your Ascent, and the CVT will go into "torque launch" more, bypass the fake shift programming and operate entirely as a CVT for the full acceleration, while applying maximum torque from ZERO miles per hour to (in the Ascent's case) the 80's or so.
Car & Driver is still populated with motor heads who equate a hard shift with "rugged". That is false thinking because even the beefiest > 1 ton truck automatics have gotten crazy smooth. My past several vehicles have been MoPars of various sorts, and the shifting is imperceptible. If anything, the "shifts" in a Subaru are felt more, and, to me, an old antique, they are just not necessary.
 

·
Registered
2022 Ascent Onyx, Ice Silver
Joined
·
671 Posts
No.

The fake shift points are in our transmissions as a necessity to make Car & Driver stop crapping on every Subaru with a CVT. And yet, still, they complain it doesn't shift harsh enough or frequently enough (seriously, 2021 review).

You can torque launch your Ascent, and the CVT will go into "torque launch" more, bypass the fake shift programming and operate entirely as a CVT for the full acceleration, while applying maximum torque from ZERO miles per hour to (in the Ascent's case) the 80's or so.
As I stated in another thread, Subaru should incorporate user selectable TCM options.....
"Normal", "ECO", "Tow/Haul", and "Car & Driver" (incorporating harsh faux shifts). ;)
 

·
Registered
2020 Ascent Touring / 2020 Forester Touring
Joined
·
880 Posts
i have access to 4 subarus and i drive 2 of them myself (2020 Forester and 2020 Ascent ) and im on the 6th subaru..
also access to service manual
also you can just go to subaru outback forum and see how 2019 -2022 owners talking about bypassing shifts using cruise control ( same can be done in Ascent )

Not my experience driving them at all. Sure using the manual mode, but never in “D”. And I’ve tried to reproduce it just to test my theory.

Where are you getting this data?
 
1 - 20 of 71 Posts
Top