Alright this is my newest version of my Torque converter. The gear box is about a 20:1 and the cones are 1:2 to 2:1 reduction/gain. The atwood is mounted seperately from the drive plate to make removal easier. Im thinking that a peice of the brecoflex toothed timing belt would be sufficient in making driving the cones. This system was designed for a 3.5 inch cog/track set(whats on our robot). Please comment.
i dont know how the timing belt is going to work… does it stretch? idont think it does… which means its probably not going to slide up or down those cones… because the bigger side of the cone would have a bigger gap between each “timing ridge” than on the smaller side.
A few thoughts:
If you’re planning on using some sort of flat timing belt, it appears that there will be significant twisting occuring, since the cones are pointed in opposite directions. The belt could have a difficult time fitting around either cone edge. You may end up with only an edge of contact, which could reduce the life span of your timing belt significantly.
A few “deep” thoughts… and some questions that you may want to ask yourself:
Why do you want to build this? Is this for a hobby or for a competitive advantage? If this is for competition, what advantages does this sort of design carry over dual speed, on-the-fly shifting transmissions?
Benchmarks to compare:
- Complexity / Ease of Manufacturing
- Ease of Repair
- Reliability / Consistancy
Before you get too far into this design… these are some questions you may wish to ask yourself. Initially… I think that you’re mostly looking to improve performance, since you’ll have a “perfect” gear ratio every time.
**Do you really need a “perfect” gear ratio? **
Some other things to think about_:
CVTs are often used in motors that are running in one direction for (reasonably) long periods of time with moderate changes in torque needs. How does this design handle start and stop (and reverse motion!) that are so common in FIRST competitions?
Just some questions to ask yourself… good luck!
Ohh man Matt!
Did you have to go and bring some actual engineering into this?
What were you thinking?
subliminal Everyone should rep Matt right now, because he’s awesome, and we need more posts like this one. subliminal
I’ve seen mucho designs like this, I’ve even made one with chain (don’t ask how, it’ll take forever to explain), but you generally want to offset the cones so that the belt is always contacting at 90 degrees with the cone. What I mean is for you to offset the cones enough so that the belt runs at a diagonal to the frame, but at 90 degrees to the cones. This is done so that the belt isn’t running crooked on the cones. Also, you want to keep the belt as narrow as possible, since as you go up or down the cones, no matter how little, the ratio changes. Theoretically you want one contact point, or an infinately narrow belt, but that doesn’t work. If the belt is too wide, the edges of it will be consantly slipping and will get worn out really fast. This design works and all, but you also need to put tension on that belt or it will slip. Other than that, it seems like a very cool design, very makable, and it should work.
My question is, how are you going to move the belt up and down the cones? If you had 2 such transmissions on tank drive, it might be hard to make the thing go straight forward since they probably won’t be in exactly the same gear.
Tell you what, if you want to see the one I designed, I could show you some pics and describe how it works, maybe we can help each other out. Just email me at email@example.com.
You don’t need to belt.
They way those cones are set up, a gear with widely spaced, narrow teeth would work.
One tooth for each grove.
However the shaft the gear slides on would need to be angled to maintaint a constant distance between the cones. And that would mean a weird free-rotation thing for the gear.
Argh, I spent 3 hours designing a CVT in inventor only to realize "wait a minute, this won’t work . . . . . "
Yeah, the gear is out because of teeth. You need a sticky wheel in between them, but not sticky enough to prevent ot from sliding up and down. Argh. Well, at least I learned something.