NuVinci Left Side Freewheel

[MattyCiii]

...It's quite heavy compared to a freewheel, though....but given the weight of the NuVinci...

Yeah that little doughnut weighs a ton. More than the weight I'm concerned about corrosion the street environment. But hey - it's a grand experiment.

 

[mdd0127]

I've been using a csk25pp in combination with a 3/4" od roller clutch on the output of my reduction and it's worked perfectly for over 2K miles. The csk is well sealed and this is the drag that you feel. Putting the freewheel on the reduction output means that the electric drive chain is always spinning but it also reduces the jolt of taking up the chain slack significantly, which is really nice on a powerful bike.

I made a csk35pp adapter to mount to the rear hub, which was from a montessa 315r trials motorcycle, but once I saw the torque the system was actually making at the wheel, I decided to go a different route by putting the freewheel mechanism at the reduction output instead of the hub, reducing the torque that the one way bearing sees by almost 5 times.

It will be really interesting to see how the csk40 holds up! Please don't abandon the project. The novelty is cool.

 

[spinningmagnets]

I'm very happy to see someone proving this works.

There are two problems (as I see it) having a FW at the small sprocket end of a chain reduction. If a reduction uses Bike chain (#41X) there's not much of a smaller drive-sprocket that's available, but if we switch to #219 chain, the motor sprocket can be half the size, doubling the reduction. This pushes the FW to the large sprocket/pulley, which turns slower. As you have stated, a conventional FW can be loud, like the White-Industries ENO which is often recommended for the high-power builds. If we can achieve enough reduction, sometimes we can get away with achieving the goal with a single-reduction.

Thud decided to not produce these for sale due to low demand, but here's a pic of his side-by-side FW/tiny-sprocket:

For high-power builds (exceeding the long-life rating of common bicycle parts on the stock right side drive), there is a market for a left-side freewheeling chainring adapter. Since a left-side drive is still a "one-speed", I would recommend that it is best used on a high-power build (for me, anything more than 2200W). For that, I believe you are better off with the larger CSK40.

For anything under 2200W (72V X 30A), I feel its very useful to give the motor some gears, with BB-drives being one of the easiest methods to do that. I would recommend the adapter disc have holes to accept both 130-BCD bicycle chainrings and Extron #219 chainrings (if possible).

 

[MattyCiii]

I've been using a csk25pp in combination with a 3/4" od roller clutch on the output of my reduction and it's worked perfectly for over 2K miles. The csk is well sealed and this is the drag that you feel. Putting the freewheel on the reduction output means that the electric drive chain is always spinning but it also reduces the jolt of taking up the chain slack significantly, which is really nice on a powerful bike.

I made a csk35pp adapter to mount to the rear hub, which was from a montessa 315r trials motorcycle, but once I saw the torque the system was actually making at the wheel, I decided to go a different route by putting the freewheel mechanism at the reduction output instead of the hub, reducing the torque that the one way bearing sees by almost 5 times.

It will be really interesting to see how the csk40 holds up! Please don't abandon the project. The novelty is cool.

Great info, thanks for sharing. You clearly articulate the push-pull of what I'm trying to achieve.

Presently I have a 13t freewheel driving a 52t chainring. When the bike is in motion, the chain drive is always moving. This project is all about eliminating this chain motion when I'm riding in a pedal only mode (all the more reason to do it on the folding bike first; the folding bike actually works well as a pedal only bike. The A-Line... let's just say it's a bear to move under pedal power only).

But nothing comes for free. Freewheeling at the 13t cog means only having to handle 1/4 the forces of freewheeling at the 52t cog. To stop back driving the chain, I have to use something sturdy enough to withstand the peak forces. The CSK-40 is rated at 260nm... we'll see if that's enough!

 

[MattyCiii]

...I would recommend the adapter plate have holes to accept both 130-BCD bicycle chainrings and Extron #219 chainrings (if possible).

Thanks Spinningmagnets!
You hit the nail on the head right there. One of my design goals is to do just that. IIRC, #219 chain uses six bolts at 5.325" (please correct me if I'm wrong!) - which is a good match for the 130 BCD chainring spacing. The outer part for this will have bolt holes for both.

I'll initially run the system with the current chain, 13t freewheel and 52t chainring I have on there now. No harm in that, and initially fewer parts will be swapped. Then very soon later (like... a week of testing, then a week for parts to be ordered/arrive) I'll swap out driver, chainring and chain for #219.

Any recommendations on #219 suppliers? I know there are a ton out there - the problem is too much choice, not too little!
1. I know I'll have to machine a shaft collar adapter from the 1/2" shaft I have now to the #219 comon 3/4" bore.
2. Chain tool... any brand I should favor, or avoid? I am very comfortable breaking/making bike chain with common cheap bike store chain tools - I imagine the skill transfers fine to #219
3. It appears there are plenty of driver/sprocket combibnations available where I can keep my 1:4 ratio or thereabouts. I'll probably go for at least a ~16t driver, reducing wear and noise.
4. Any thoughts on vendors?

 

[mdd0127]

I got some of my #219 sprocket blanks from Tony at Asuza engineering. http://www.azusaeng.com He was really easy to work with and the prices were reasonable. I also used a slightly modified tooth pattern generated by sprocketeer2.0 and cut some sprockets on my mini cnc mill and have had excellent wear and noise results. #219 o-ring chain can be purchased from go kart suppliers but it's not cheap and requires a special chain tool. It sure is awesome though. After 2k+ miles, there is no visible wear in my drivetrain, even after doing a lot of riding in the dirt. I run o-ring chains dry and this really seems to help them last longer because larger particles don't stick to the lube.

 

[drewjet]

This is my thread when I was in need of 219 sprockets
http://endless-sphere.com/forums/viewtopic.php?f=28&t=12677&hilit=219+sprocket

And this is who I have used with good results
http://www.gokartgalaxy.com/engine_sprockets1.htm

 

[spinningmagnets]

deecanios build is the one that made me want to start saving build links, the Extron (composite plastic) chainring runs quieter than metal. the chainring is only slightly larger than the standard size, but #219 allows a much smaller drive sprocket for a dramatic improvement in reduction. I don't have any experience with #219, but I saved some info as a future option. http://endless-sphere.com/forums/viewtopic.php?f=16&t=46630&p=681768#p681768

13T steel sprocket with an 89T composite chainring (6.8:1 reduction in a single stage)

 

[bobc]

Yeah - I'd like to know how the sprag performs too......miles' drag figures look so much lower than what I was seeing, & my sprags were on the layshaft (5x revs of back wheel) with one whizzing around backwards at near double speed, so the drag was a bit more serious for me.....!!

 

[MattyCiii]

...all in the name of science!

 

[EvoBikeShop]

Do you think your design could work or be adapted to a nexus hub?

 

[MattyCiii]

Do you think your design could work or be adapted to a nexus hub?

I'm guessing no, based on looking at some Google Images of the nexus.

What makes this possible with the NuVinci is how much of the 135mm dropout width I gain by removing the modular disc brake mount:

And that I get a lot of clear radius around the axle

As it stands, the CSK40 barely fits widh-wise. A CSK-35 is narrower, but can handle a little over half the torque so it might not be up to the job.

That siad, this project has only two custom parts: the inner part interfaces the NuVinci to the inside of the CSK40, and the outer part interfaces the outside of the CSK40 to the chainring. Maybe all that's needed is a custom inner part that interfaces the nexus to the CSK. The inner dimensions of the CSK40 are 22mm wide by 40mm diameter - if you don't have at least 22mm between the nexus and the left chainstay of the bike frame, this is a non-starter.

 

[spinningmagnets]

IIRC, #219 chain uses six bolts at 5.325" (pleasecorrect me if I'm wrong!)

I found a couple references that Extron Karting sprockets are 5-1/4/5.25-inch 133.35mm Bolt-Circle-Diameter (BCD), six holes 5/16-inch diameter...I have never owned one so far.

 

[MattyCiii]

New life for an old project!

The NuVinci Left Side Freewheel project is now an important component in the "common power module", the centerpiece of my third build. Thanks to the hard work and most excellent workshop of 900steve, this project is moving forward again:

Here we see a NuVinci N360 with a 3D printed "inner piece" attached. On the table is a CSk-40 sprag clutch, wrapped in the "outer piece". This variant of the outer piece has a 104mm 4-bolt BCD pattern.

Here are all three mated together (friction fit), mounted on the N360.

Here's the assembly being spun by hand. It feels really smooth, and is whisper quiet.
[youtube]a4DXNK6ElMs[/youtube]

Now with a used 32t chainring attached:

And again the assembly mounted:

The chainring bolts are a bit too long:

Lots of clearance. I can probably get about 3-4mm closer to the NuVinci if I want to:

The plan moving forward is to build one of these in metal and get it mounted on my folding bike as soon as possible. This way we can start road/torture testing it immediately. The folding bike is the better choice than the A-Line for this task, because it does not use rear disc brakes. A disc brake variant of this freewheel adapter is possible, but more complex.

Once on the folding bike, we'll give it some acute torture tests then settle in for long term riding tests. We'll either break that CSK-40 or we won't, and that will of course be a crucial piece of information moving forward to other projects.