Subaru Ascent Forum banner
1 - 20 of 36 Posts

·
Registered
Joined
·
561 Posts
Discussion Starter · #1 ·
Interesting info about torque and the Subaru CVTs. I guess the ECU can limit the about of torque that can be sent to the CVT.
This is probably why Fast Lane Car was having issues off roading with the Outback.



 

·
Registered
Joined
·
223 Posts
Interesting that nobody else has posted on this thread. I find this topic very interesting as the Ascent's CVT's is apparently different and better than many other brands' CVT transmissions. But many are not sure why other than a belt/chain difference. Some of the videos information I've been reading, Subaru CVT does have trouble in certain situations, which can be frustrating ie the video above with no power being put to the wheels with the accelerator matted.

Does anyone have any technical articles or websites with credible information about CVT torque especially with respect to towing or off-roading in low traction mud/snow/incline?
 

·
Administrator
Joined
·
10,968 Posts
Interesting that nobody else has posted on this thread. I find this topic very interesting as the Ascent's CVT's is apparently different and better than many other brands' CVT transmissions. But many are not sure why other than a belt/chain difference. Some of the videos information I've been reading, Subaru CVT does have trouble in certain situations, which can be frustrating ie the video above with no power being put to the wheels with the accelerator matted.

Does anyone have any technical articles or websites with credible information about CVT torque especially with respect to towing or off-roading in low traction mud/snow/incline?
I've posted a ton of info on it. And, no, the Ascent doesn't limit torque. It sends torque info to the CVT so the CVT can regulate chain tension.

I can go into more detail later if you can't find my posts.
 

·
Administrator
Joined
·
10,968 Posts
@gmatt1000 the Ascent Lineartronic TR690 is used to a ridiculously small percent of its torque capabilities.
  • Note "Lineartronic SOP Specification" which is standard operating spec for the TR690. Note that the max spec is waaaay at the other end. Audi and Subaru are the key players who use big LuK chains.
2756


It has a big, multi-layer drive chain.
2757


  • JATCO CVT's (Nissan and other brands) and virtually every other CVT model out there (except Audi, for one) use a steel band pushbelt that cannot handle much torque.

    Note that there's a thin metal band inside of those "teeth" that must continuously flex and unflex around the pulley cones.
2758



Seriously, try to imagine flexing those bands and sliding teeth above as much as this (much less thousands of times a minute) while under heavy load with high torque being ran through it.
2764




MORE STUFF
The TR690 is capable of handling a massive amount of power. It is configured for a torque range and load band per each application (eg: Outback 3.6R vs WRX vs Forester XT vs Ascent) - and then a certain safety buffer is added (and, I know how much, btw).

The Ascent, being the absolute heaviest Subaru ever to roll out of the factory, and with more power than many stock WRX's and more than the Outback 3.6R, is also designed to tow 5,000 pounds, and it does so easily.

So, that's a total of 11,000 pounds our TR690 is configured to move. The Outback 3.6R is the next closest, at a total of 7,550 pounds.

  • Ascent TR690: 11,000 pounds
  • Outback 3.6R TR690: 7,550 pounds max load
  • WRX TR690: 4,400 pounds
That's a MASSIVE difference in TR690 configurations.

Don't confuse a lower torque, NO tow rating WRX's TR690 with the one that's in my Ascent. While the insides are pretty much the same, the valve bodies are configured drastically differently, primary and final reduction gears are different, and, the clamping pressure on that big chain is MUCH higher on the Ascent.

It's that clamping pressure, btw, that determines the load the TR690 can handle (the load on both sides: engine torque and what weight it has to move).

In the video below, the travel trailer is full... grey tank, black tank, and potable tank - plus full of gear. It's at its 4,395 pound GVWR. The Ascent can still pull it with ease, even with my Ascent at 5,600 pounds of car, gear and people.

???
That's about 10,000 pounds, or, THREE WRX's (curb weight average about 3,350 pounds).
???

Yes, the TR690 is THAT capable, and, even mine isn't anywhere near the design limits (even with the safety margin included). This little test would obliterate a pushbelt Jatco.

ASCENT TOWING 4,395 POUNDS 0-60 TEST:





2762



Other Details:
  • The ECU sends torque data to the TCU. The TR690 comes configured for the vehicle model it is going in. Clamping pressure is lower when the vehicle is in park, neutral or not doing anything in drive/reverse.
  • The TR690 is configured to handle maximum torque and load with a safety margin built in. When torque increases, the TCU increases chain clamping pressure to the necessary level plus the safety margin. That's instantaneous.
  • Some people confuse torque converter lock (or non-lock under a certain RPM) with power being cut to the CVT. That's not what is happening. The torque converter simply hasn't locked yet. It happens to ALL automatics that don't have stall converters or some other performance converter.
This flexible band with teeth does not compare to a Subaru/LuK chain.
2759


TR690 Chain:
It can turn very tight radiuses without risk of snapping, just like chains do.
2760




EXTRA TRIVIA:
OUR
TR690 CVT chain looks different than this one that has uniform teeth for a reason. The chain below supposedly makes a horrible droning sound. LuK spent a lot of time calculating the quantity and locations of the chain plates on the pins to make our chain make their much more pleasing power-steering-ish sound.
2761
 

·
Administrator
Joined
·
10,968 Posts
Nice information, but...
"...and then a certain safety buffer..."
If you know how much... spill the beans!
I do, and I even have the base math for the torque vs clamping pressure calculations, and an outline of the math for the chain plate placement (my CVT library is pretty big), but I can't divulge that stuff.

But, let's just say that some tuning apparatus company likely found the limit during testing, then backed off to the number they settled on for max torque (max 21% increase)... ;)
2766

The number is NOT 21% - that still leaves a safety margin. As they said, they backed off a certain amount to come up with 21%.

And remember, that's not the TR690's limits... that's the Ascent configuration limit. ;)
 

·
Registered
Joined
·
1,164 Posts
...Don't confuse a lower torque, NO tow rating WRX's TR690 with the one that's in my Ascent. While the insides are pretty much the same, the valve bodies are configured drastically differently, primary and final reduction gears are different, and, the clamping pressure on that big chain is MUCH higher on the Ascent
You've been pretty adamant in asserting the transmission has gone 10 years completely unchanged, something I've remained skeptical of. This sounds like it has changed.

Also, the torque the transmission is subject to is generated by the engine irrespective of the weight of the vehicle. With a heavier vehicle, the tranmsission will be subject to that torque for a greater percentage of time, but it's the same max torque.
 

·
Administrator
Joined
·
10,968 Posts
You've been pretty adamant in asserting the transmission has gone 10 years completely unchanged, something I've remained skeptical of. This sounds like it has changed.
No, that is not an accurate representation of my posts on the matter....

I have repeatedly said, the TR690 is a ten year old design. BUT, I have also repeatedly said that it is configured for the vehicle in question. The valve body is ALWAYS configured to the car and the year's design specs. The TCM is car specific. I've said that repeatedly, including pointing out that the TR690 did NOT come with faux gearing valve bodies, and that when the faux gearing was introduced, the faux gearing changed from (none to) 6 to 8 faux gears. I've also repeatedly pointed out that primary and final reduction gears are determined by the application, in conjunction with the center differential and rear differential gearing (the differentials are not the CVT). I've pointed out the differences in gear ratios between other popular TR690's, such as in the Forester XT.

Here's one of dozens of examples:
The pulleys, chain, ATS system, etc, are a ten year old tried and tested design. The TR690, in every application, is a ten year old tried and tested design. The load bearing parts are mostly unchanged in ten years.

NONE of those are anything related to the "CVT" problem experienced in an Ascent. The harness is not the transmission. I've also pointed out, numerous times, including having to get part numbers for a forum member who thought otherwise, that the harnesses are different in each Subie model.

Also, the torque the transmission is subject to is generated by the engine irrespective of the weight of the vehicle. With a heavier vehicle, the tranmsission will be subject to that torque for a greater percentage of time, but it's the same max torque.
Torque input is "read" from the ECU via the torque sensors and fed to the TCM. Regardless of that, on the other side of the pulley is the load. That load is absolutely positively relevant to clamping pressure. Torque alone is entirely irrelevant.
  • On one side of the system (on the drive pulley cone) is torque.
  • On the other side of the system (on the driven pulley cone) is load.
I hope that explains that better. Both sides of the system are vitally important in figuring out clamping pressure.

You've given me an idea for what I think will be a great vid!!! Thanks!
 

·
Registered
Joined
·
266 Posts
Torque input is "read" from the ECU via the torque sensors and fed to the TCM. Regardless of that, on the other side of the pulley is the load. That load is absolutely positively relevant to clamping pressure. Torque alone is entirely irrelevant.
  • On one side of the system (on the drive pulley cone) is torque.
  • On the other side of the system (on the driven pulley cone) is load.
I hope that explains that better. Both sides of the system are vitally important in figuring out clamping pressure.
I'm with @Merope on this one. Yes you have the load but if the load exceeds the engine torque it simply decelerates the engine, so the transmission only sees the max. torque at engine plus some peaks of dynamic force, let's say when you pull somebody with a rigid strap and receive the initial pull, it causes a peak that will exceed the max. torque of the engine by a fraction (which will make the engine decelerate rapidly). These peaks might be temporary but can easily reach 1.5 to 2 times the torque.
Then on top of that you need a safety blanket to make sure it doesn't slip. I'm sure the 20% increase is eating out into this safety margin, but as long as you aren't too rough those dynamic peaks can be kept reasonable and you still have decent safety margin.
So, yeah, @Merope has a good point, load doesn't really matter much when you know the max. torque of the engine. You only need to know the forces at one side of the transmission, you can pick which side if you want but mind you, the load side is affected by the gear ratio, so it's much easier to know the other end.
 

·
Administrator
Joined
·
10,968 Posts
Sorry, that's not correct. I've got the math. LuK's TCU programming math disagrees. ;) Torque alone is irrelevant. I will demonstrate in a week when I get the parts in.

Load is always changing. Up a hill, down a hill. No load requires minimal clamping pressure.

Test that for yourself with two pulleys and a drill. Set up a pulley system with a belt, and fire off the drill. No slip. Put a load on the "output" pulley and watch your "chain" slip. Increase the chain tension and watch the load move with no slip. Increase the load, and the chain begins to slip again, even with the same torque from your drill.

Heck, it's the first thing we learn when we're doing air conditioner work on the big split units with belt driven compressors we used to use in server rooms. The pulley gets tensioned for the combination of the torque of the motor, and the compressor-turn-on load. Tensioning it just to the point the belt doesn't slip with the compressor off means it will slip when the compressor kicks on. Torque didn't change.

Same goes for car ac compressors. Engine torque is still in the same range, but load increases and the belt slips.


I'm with @Merope on this one. Yes you have the load but if the load exceeds the engine torque it simply decelerates the engine, so the transmission only sees the max. torque at engine plus some peaks of dynamic force, let's say when you pull somebody with a rigid strap and receive the initial pull, it causes a peak that will exceed the max. torque of the engine by a fraction (which will make the engine decelerate rapidly). These peaks might be temporary but can easily reach 1.5 to 2 times the torque.
Then on top of that you need a safety blanket to make sure it doesn't slip. I'm sure the 20% increase is eating out into this safety margin, but as long as you aren't too rough those dynamic peaks can be kept reasonable and you still have decent safety margin.
So, yeah, @Merope has a good point, load doesn't really matter much when you know the max. torque of the engine. You only need to know the forces at one side of the transmission, you can pick which side if you want but mind you, the load side is affected by the gear ratio, so it's much easier to know the other end.
 

·
Registered
Joined
·
5 Posts
Can someone please tell me why my engine rpm's increase between 1000 - 1400 rpm when I'm going up a slight hill, and all the time my throttle pressure and vehicle speed remain constant?

If it was a manual transmission I'd swear the clutch was slipping.
 

·
Administrator
Joined
·
10,968 Posts
Can someone please tell me why my engine rpm's increase between 1000 - 1400 rpm when I'm going up a slight hill, and all the time my throttle pressure and vehicle speed remain constant?

If it was a manual transmission I'd swear the clutch was slipping.
Because it's an infinitely variable CVT, and the computer will adjust the ratio as necessary for a balance between power, fuel economy and power requested by the driver.

In simpler terms, it's quite normal and exactly how a CVT is supposed to operate.
 

·
Registered
Joined
·
6 Posts
I do, and I even have the base math for the torque vs clamping pressure calculations, and an outline of the math for the chain plate placement (my CVT library is pretty big), but I can't divulge that stuff.

But, let's just say that some tuning apparatus company likely found the limit during testing, then backed off to the number they settled on for max torque (max 21% increase)... ;)
View attachment 2766
The number is NOT 21% - that still leaves a safety margin. As they said, they backed off a certain amount to come up with 21%.

And remember, that's not the TR690's limits... that's the Ascent configuration limit. ;)
Why can you not share this information? Are you under a non-disclosure agreement? The point of these forums is to share information.
 

·
Administrator
Joined
·
10,968 Posts
Why can you not share this information? Are you under a non-disclosure agreement? The point of these forums is to share information.
I don't betray other's trust. I hope you can understand and accept that reason.

Besides, the point of the discussion isn't what the number is, but that there is a decent safety margin, and the basics of how the system works.
 

·
Registered
Joined
·
820 Posts
This is quite the education on CVT’s & the Ascent model. I read all of it and it just puts my mind at ease in understanding alittle more the drivetrain in this vehicle.
Although this CVT stuff is beyond my comprehension one thing I’m keen on is how Subaru put together this combination of CVT,Turbo,Symetrical AWD and running on 87octane is incredible .
I have not owned a vehicle in over 30+ yrs with horsepower over 200 not to mention a turbo’d vehicle.When I get a chance to stomp it I just smile and hope that I don’t get pulled over ?
 

·
Registered
Joined
·
6 Posts
I don't betray other's trust. I hope you can understand and accept that reason.

Besides, the point of the discussion isn't what the number is, but that there is a decent safety margin, and the basics of how the system works.
Thanks. I agree the Subaru CVT is more robust than published. I have a Forester XT with a CVT. That was tuned by Ray at TurboTek. My forester produces more than 280 lb of torque at the wheel. The car has 80k miles and is running great. I was told the CVT could do so much more.
 

·
Registered
Joined
·
731 Posts
.. one thing I’m keen on is how Subaru put together this combination of CVT,Turbo,Symetrical AWD and running on 87octane is incredible .
I have not owned a vehicle in over 30+ yrs with horsepower over 200 not to mention a turbo’d vehicle.When I get a chance to stomp it I just smile and hope that I don’t get pulled over ?
That’s one of the biggest reasons I keep coming back to Subaru. Turbos and driving dynamics.
Naturally aspirated cars are turds @ 6,800ft - 30% power loss up here. Turbos don’t have that problem, maybe 5% loss, worse case. Subaru dropped the turbo in the Forester - guess what, no more Forester for me. I don’t do naturally aspirated 4-bangers at altitude.
I want a BRZ badly, but that car is an absolute torqueless turd up here.
So, I’ve had a long history or every turbo”sports car” Subaru has made.
I’m waiting for Subaru to drop the FA24DIT into the fun cars.
 

·
Registered
Joined
·
820 Posts
That’s one of the biggest reasons I keep coming back to Subaru. Turbos and driving dynamics.
Naturally aspirated cars are turds @ 6,800ft - 30% power loss up here. Turbos don’t have that problem, maybe 5% loss, worse case. Subaru dropped the turbo in the Forester - guess what, no more Forester for me. I don’t do naturally aspirated 4-bangers at altitude.
I want a BRZ badly, but that car is an absolute torqueless turd up here.
So, I’ve had a long history or every turbo”sports car” Subaru has made.
I’m waiting for Subaru to drop the FA24DIT into the fun cars.
Imagine a turbo in the BRZ,build houses won’t be able to keep up with demands.
 

·
Registered
Joined
·
820 Posts
I’ve been imagining for a long time. And hoping.
It’s a cool looking roadster but definitely the power plant is a big let down.
A lot of manufacturers are doing away with manual cars,I get it but don’t get it,but at least give us some power to play with .When you look at cars like the BRZ or Toyota 86 they were meant,my opinion,to have the performance just by the appearance alone.

A “Baby Supra” if i may ??
 
1 - 20 of 36 Posts
Top