Heavy duty disc CVT

[Divitis]

I am building something big, both in scale and ambition.

The first module is a heavy duty CVT with minimal friction, capable of spinning significant mass (that cone is heavy) at high speed and... with a hole.

 

 

(21Mb gif, give it time to load)

 

So far, so good, I think. I'm really happy with the inertia it has, even when spinning it from a bare axle which is rather disadvantageous force wise.

Will try motorizing it next.

Feedback is welcome, and I'll share the LLD with anyone interested.

Edited November 17, 2025 by Divitis
Embedded links

[2GodBDGlory]

Nice, I love to see cool mechanisms like that being developed! I did a similar build years ago, but I was trying to get it compact enough to squeeze into a car model, so it was built a bit differently. Yours should have a much larger spread of ratios than mine did!

I do think this style of CVT is probably the most practical one for Technic.

Is there a larger project this is going in? That cone looks pretty specific

[Divitis]

5 hours ago, 2GodBDGlory said:

Nice, I love to see cool mechanisms like that being developed! I did a similar build years ago, but I was trying to get it compact enough to squeeze into a car model, so it was built a bit differently. Yours should have a much larger spread of ratios than mine did!

That's a very compact implementation indeed, how was the torque and reliability?
What's really working well (for my standards at least) in my design is driving the disc from both sides. This ensure good traction while eliminating the need to brace the large disc to make it spin perfectly flat, which generates lots of attrition.

One day, this will become a GBC module. 

Balls will enter the conical basket from below, that's why the opening in the base, and then exit from above once the cone spins fast enough.

I also need to time the CVT so that the cone only reaches 'escape velocity' for a couple seconds and then slows down and keeps going slowly for some time, while more balls are loaded in.

In terms of ratios, I should be able to get to 4:1, which should make the change in speed interesting enough.

This requirement dictated the two pieces design, spinning a disc with a hole and basket directly on top means that the closest to the center I can get with the driver wheel is 5 studs. And the disc would need to be 40 studs in diameter to have the 4x velocity change. A bit impractical.

Edited November 19, 2025 by Divitis

[2GodBDGlory]

6 hours ago, Divitis said:

That's a very compact implementation indeed, how was the torque and reliability?
What's really working well (for my standards at least) in my design is driving the disc from both sides. This ensure good traction while eliminating the need to brace the large disc to make it spin perfectly flat, which generates lots of attrition.

One day, this will become a GBC module. 

Balls will enter the conical basket from below, that's why the opening in the base, and then exit from above once the cone spins fast enough.

I also need to time the CVT so that the cone only reaches 'escape velocity' for a couple seconds and then slows down and keeps going slowly for some time, while more balls are loaded in.

In terms of ratios, I should be able to get to 4:1, which should make the change in speed interesting enough.

This requirement dictated the two pieces design, spinning a disc with a hole and basket directly on top means that the closest to the center I can get with the driver wheel is 5 studs. And the disc would need to be 40 studs in diameter to have the 4x velocity change. A bit impractical.

I was pretty impressed with the performance of my design! Torque wasn't enormous, but it was enough to drive a large car reliably, with planetary reduction afterwards.

Oh, I hadn't realized that you're driving the disc from both sides! That's cool, and should improve grip over my design. I had to use a 2-stud wide tire, which produced a speed differential over the different parts of the tire, causing wear.

Sounds like a cool GBC project!

[Divitis]

Some progress.

The motorized contraption spins fast enough for the balls to escape. Now for some better bracing and I can move on to motorizing the sliding of the driving gears on the disc.

I'm thinking of using a chain to be have full control over the timing - any suggestion is very welcome!

The rails with the wheel on top somewhat visible in the back are the 'ball accelerator' which will propel the balls inside the cone from below. I wanted to use gravity but the tower was going to be ridiculously tall.

Will post a video of it once it works properly, the prototype was satisfying but I must have messed up some distances when rebuilding it here.

 

[2GodBDGlory]

Cool! I'm glad to see the progress!