19 Subaru Impreza - CVT failed

[Cujet]

Fact of the matter is, what AT these days doesn't have early failures?

The ZF 8HP?

That's what is in my Jaguar F-Type and it's downright EPIC good. Fast shifts up and down, and does not do unwanted things. If a specific gear is selected, it will stay in that gear, bumping off the rev-limiter all day. Not much in the way of nanny programming.

OP, did you see my Subaru CVT failure video? Car had 2400 miles on it and it failed with cruise control on, going up a long highway hill. RPM went up and up and up, until the CVT let go.

 

[benjy]

good luck! thats why i ONLY but MANUALS. usually one clutch job during ownership about a thou compared to 5 thou plus or minus for slushbox work depending on your area $$$$$

 

[andrew_j]

Sorry to hear. Seems like a manufacturer defect if only 20k miles. It says little about Subaru or CVTs in general except a small subset fail just like automatic transmissions.

 

[Pablo]

Fact of the matter is, what AT these days doesn't have early failures?

That's what is in my Jaguar F-Type and it's downright EPIC good. Fast shifts up and down, and does not do unwanted things. If a specific gear is selected, it will stay in that gear, bumping off the rev-limiter all day. Not much in the way of nanny programming.

OP, did you see my Subaru CVT failure video? Car had 2400 miles on it and it failed with cruise control on, going up a long highway hill. RPM went up and up and up, until the CVT let go.

That sounds like it's down to one rod!

 

[Pablo]

good luck! thats why i ONLY but MANUALS.

Good with that. But yeah I'm with you.

 

[M1Accord]

I think thats where CVTs belong. In small scooters and ATVs. I won't ever buy a car with one, unless i have no choice.

My feeling too. And ant car w/out ATF dipstick or an easy way to supplement one with an oil dipstick. So next car is likely a Honda again as they removed the dipstick, which saved them $3, but you can use the oil dipstick to make new mark indicators for cold and hot fluid conditions as a mean to check; don’t leave the dipstick in there as I don’t know if that gets in a way of anything.

 

[14Accent]

Subaru's and Nissan's CVT transmissions are both made by JATCO.

There are only three manufacturers of CVT transmissions in the world. Toyota, Honda, and JATCO.

If it has a CVT and it's not a Toyota or a Honda then the transmission was made by JATCO.

What about the VAG Multitronic, made by LUK?

 

[ARCOgraphite]

At Lafayette? That's Subaru's plant, not Toyota. And they stopped making Camrys there several years ago.

Yes. SIA. Just seeing who is paying attention.

My Scottish side of the family landed in Lafayette around the turn of the Century.

 

[ARCOgraphite]

Can someone explain to me how CVT belts are described as “pushers?“ if the belt were “pushed,” would it not need a “keeper” to guide it and press it against the driven pulley? In my experience, if I push a flexible object like a chain or belt, it will just flop. It requires tension, which is a fulling force, to have any grip against the pulley surfaces. How is the CVT any different?

It is NOT not a belt per se, it is a stack of shims held together with metal band(s)

 

[demarpaint]

My father had a Subaru with a CVT transmission. The car was fantastic in snow, their AWD system is great imo. I hated the transmission and as a result would never own one.

 

[JTK]

The ZF 8HP?

True. I owned two of them and loved them. Having followed them in Ram 1500's since 2017, there was only a few odd ball failures on the forum I followed. I've heard it said that used ZF8's are cheap because no one needs them.

Can someone explain to me how CVT belts are described as “pushers?“ if the belt were “pushed,” would it not need a “keeper” to guide it and press it against the driven pulley? In my experience, if I push a flexible object like a chain or belt, it will just flop. It requires tension, which is a fulling force, to have any grip against the pulley surfaces. How is the CVT any different?

The CVT belt/chain push vs pull thing has always cracked me up as well. One side is always going to do the "pushing" wile the other "pulls" I see no relevance in arguing either way. They all have to be held in tension, with the clamping force being on the sides.

 

[ARCOgraphite]

good luck! thats why i ONLY but MANUALS. usually one clutch job during ownership about a thou compared to 5 thou plus or minus for slushbox work depending on your area $$$$$

Subaru dropped the manual for 2024 in the Impreza and Crosstrek.

So that leaves: VW Jetta, Kia Forte, Hyundia Elantra, Honda Civic and Acura Integra, then Mazda 3
in "AFFORDABLE" cars with a manual. Other than the Jetta and the Mazda good luck finding one for sale.

I may be in the same hunt.

 

[JTK]

The new Nissan Versa S can be had with a 5MT.

 

[ARCOgraphite]

The new Nissan Versa S can be had with a 5MT.

Maybe I have to try another, but worst quality car I have ever driven and I am a cheap ecobox kinda guy.

Checked under the hood before the drive and couldn't even get the hood to close, lol. Wonky transmission, screwey engine management - just not put together.

I do recall years ago testing Golf turbodiesels and driving a few to pick a good one that steered straight and ran well. Almost found one, but when reversing the exhaust hit the under-body and buzzed loudly. After three hours of this nonsense, I said forget it.

This is from a guy (me) that loved his Yugo GV, a Brazilian VW Fox, Geo Metro 3 cyl (Suzuki) and tried to own a '14 Versa Note MT - the affair only lasted two weeks.


FolksVaagen Fox

 

[JTK]

Maybe I have to try another, but worst quality car I have ever driven and I am a cheap ecobox kinda guy.

The latest generation has quite the following for those who've been able to locate/purchase an S model. They can be had for $17405 (which includes "shipping/handling" as nissan calls it now).

I hear you on the previous gen. I just drove my 2015 S-model sedan w/ 4AT today as it has been sitting unused in my driveway for a month or so. It's like taking a step back in time.

 

[ARCOgraphite]

The CVT belt/chain push vs pull thing has always cracked me up as well. One side is always going to do the "pushing" wile the other "pulls" I see no relevance in arguing either way. They all have to be held in tension, with the clamping force being on the sides.

NO argument to be had - other than on these pages

The designer of the unit states the force is transferred buy "pushing"

The links are not connected to one another or to the assembly band(s) - so how do you pull a line formation of dominoes? Regardless, it a very odd design I can't even imagine coming up with such a power transfer concept.

 

[KGMtech]

I'm waiting on the update from OP. The lower torque Jatco CVT with the 432 pieces 'belt' is less capable than the chain style, I would suppose that because the 432 shim pieces are stacked by a multi layer steel piece that has no mechanical or welded connection to these shims, this is going to be a failure point. However I've heard these belts mostly fail because the hydraulic pressure on the 2 big cones is lost, so the 'belt failure' is the result and not the cause. Take apart the failed CVT and it will show an explosion of shim pieces, and most conclude the belt left the chat first. No no.

 

[andrew_j]

I'm waiting on the update from OP. The lower torque Jatco CVT with the 432 pieces 'belt' is less capable than the chain style, I would suppose that because the 432 shim pieces are stacked by a multi layer steel piece that has no mechanical or welded connection to these shims, this is going to be a failure point. However I've heard these belts mostly fail because the hydraulic pressure on the 2 big cones is lost, so the 'belt failure' is the result and not the cause. Take apart the failed CVT and it will show an explosion of shim pieces, and most conclude the belt left the chat first. No no.

Warranty won’t repair they’ll just replace the entire unit with remanufactured one.

 

[JTK]

...However I've heard these belts mostly fail because the hydraulic pressure on the 2 big cones is lost, so the 'belt failure' is the result and not the cause. Take apart the failed CVT and it will show an explosion of shim pieces, and most conclude the belt left the chat first. No no.

Exactly.

I don't think the root cause has ever been anything different than a system problem causing not enough clamping pressure. Loss of clamping pressure = slippage, cone wear due to slippage, now metal debris in the oil and that's all she wrote.

 

[4WD]

Have only had CVT’s in rentals and will keep it that way …