Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

CVTs for Electric Vehicles

This site may earn commission on affiliate links.
So I’ve read through the earlier forum discussions on gearboxes for EVs, but was hoping we could start one specific to Continuously Variable Transmissions (CVT). It seems to me that there could be some serious benefit to engineering an effective marriage of EV powertrain to a CVT, but I’m interested to hear what the brain trust has to say.

Here’s some evidence:
1) This article validated the improvement in torque off the line and vehicle top speed by modifying a two-wheeled EV with a CVT. Note the negligible range impact in Table 1.
2) The upcoming Lito Sora has successfully applied this concept to a high-performance E-Bike
3) The 2012 Nissan Maxima (290 hp / 261 ft-lb) has specs similar to the Tesla Roadster, and utilizes CVT technology. The 2012 Toyota Highlander Hybrid (280 hp / 215 ft-lb) isn’t quite as similar, but incorporates electric motors and a CVT.
4) This article talks about a 2-speed gearbox can reduce battery consumption 5-10% by keeping the operating range nearer to its sweet spot. A properly-tuned CVT should have further benefit.

Just a quick list of pros/cons I’ve come up with:

Pros:
Faster acceleration
Higher top speed
Longer range

Cons:
Cost
Weight
Complexity
Reliability
Traction Response
Regen (?)

Disclaimer: I understand that the direct drive of Tesla’s current and proposed models are more than adequate for typical drivers… this is more of a “what if” to show CVT possibilities in the utility, commercial, and racing vehicle markets. As such, please don’t bomb the thread with “why do you want…” comments or miscellaneous widgets (battery/motor/wheels, etc) that may also improve performance.

Thx in advance & have fun!
 
Have you owned a car with CVT? (not just driven once or twice)
I had a Nissan Murano for several years. Compared to a normal AT, I loved CVT. But it still has a huge drawback -- response time. For me, the pros of the direct-drive single-speed Roadster far outweighs any of the cons.
 
Cinergi is correct. the eCVT in the Prius is far better than any automatic (and far more reliable) but there is lag time while things adjust to a new speed. I vote for fewer parts.
 
Cinergi is correct. the eCVT in the Prius is far better than any automatic (and far more reliable) but there is lag time while things adjust to a new speed. I vote for fewer parts.


I've been surprised on how much I love the single drive in the Roadster. Just a no brainer and go!
I just got an email to finalize my order on the Brammo e-bike. Interestingly, it has a 6 speed gear box.
 
No one has figured out how to push a lot of torque through a CVT yet. EVs own torque, so it's not clear that CVTs and EVs are a happy combination.

It would probably be less problematic to make two stage motor. Then you'd only have to deal with the electronics of the motor and windings rather than a transmission
 
No CVT EVER. I hate the way they drive and had poor experience when related to reliability.

You are correct for the normal CVT. The Prius' eCVT has an almost perfect record for reliability (as it really isn't a CVT, it's more like a differential, but they had to call it something or the DMV would be confused). Other than the slight lag I really like the way it drives.
 
The power split device (sometimes incorrectly referred to as a CVT) in the Prius is an elegant solution to the problem of combining a gasoline engine with two motor-generators in a way that allows the car to go whether the gas engine is running or not, and allows the engine to run whether the car is going or not, and allows the car's computer to charge the battery or draw from it, as well as allowing the battery to be charged from the gas engine or from regenerative braking.

But a CVT is ill-suited to an EV because an electric motor typically has a very wide torque curve and therefore has an extremely broad "sweet spot" and does not need to be finely matched to vehicle speed. A gas engine, with its narrow sweet spot, can benefit from a very fine adjustment in gear ratio, and continuously variable is as fine as you get. This need does not exist with an EV, so you can avoid the reliability issues, torque-handling limits, and latency of a CVT by having discreet gears (including the Roadster's solution of a single-speed transmission).

Tesla tried to put a two-speed transmission in the Roadster at first, but they could not find one that could handle the torque. One-speed is IMHO a better choice.

DC motors I guess have a much narrower torque curve than AC motors, but I think still much broader than gas engines, so again, a CVT is addressing a problem that does not exist for an EV.
 
daniel, you're right, but electric motors can be improved even further by adjusting the voltage/current ratio or by manipulating the magnetic fields. The net result is that electric motors can have a near perfect CVT built in, and Model S seems to already have one.

The ideal motor delivers constant power from zero to infinite speed. In practice the best possible performance is to deliver as much torque as the tires can handle until max power is reached, and then stay at max power up to the RPM limit. Looking at the MSP torque curve, what it's actually doing seems to be very close to the ideal.

I tried to explain how this works in the MSP vs BMW M5 thread here and here, without being aware of this thread.

The fourth article linked by the OP is interesting, though - it shows that while electric motors don't have a power sweet spot like the ICE, they do have an efficiency sweet spot, and a transmission can be added to improve efficiency. Simply choosing a longer gear ratio to increase top speed will also move the efficiency sweet spot to a higher speed.

I'm not sure it makes sense in practice, because the efficiency is actually very high even well outside the sweet spot, and I think efficiency can also be improved by increasing the system voltage instead. But I have no idea how all the tradeoffs work out.
 
Last edited: