Friction drive outrunner setup. New DJ bike.

[liveforphysics]

Two ideas for friction drive in general (I've no idea if they've been suggested elsewhere already):
1. Would there be an advantage in making your drive roller capstan shaped ( waisted as it were) so that more of it touches the tyre?
2. I don't know how you "engage drive" - do you drop the spinning drive roller on the tyre, or does it have to be in contact before you fire it up? If you drop it on the tire, would it be possible to sense the rotation of the drive roller and use that message to trigger engagement of the drive?
Just my 2 pence.

I can answer question 1. Making it a non-flat surface seems very appealing from a traction perspective on the tire. Unfortunately, it also means the larger diameter parts of the drive spindle touching the tire are trying to push the tire at a faster drive ratio than the smaller diameter center area. This would result in continuous excess friction, tire wear, and heat.

 

[deecanio]

Hi EVTodd,

i really like your design, can you get some video up of it in action, like etard i don't get how it applies more pressure with more throttle either , damn fine idea though.

Cheers,

D

 

[spinningmagnets]

I don't plan on building a friction-drive, but After seeing some of the well-engineered small-motor RC systems recently, I have pondered them. Some enthusiasts live where its pretty dry, and also, regardless of where you live these could be very easy to be very stealthy.

A lithium pack and an RC motor can be worth MUCH more than the bike its on, and some bikers live in a high-crime area. If you really can't put your bike inside your apartment, you could purposefully own a beater-bike, and a compact friction-drive set-up like the Go-Ped shown could be adapted to simply "pop off" so it can be carried inside at home, and also at work.

 

[12p3phPMDC]

I may be wrong, but it seems the the drive plate slides due to forces caused the spinning of the knurled wheel.
The angle of the assembly is pointed into the wheel.
The further the slide moves the more it pushes into the wheel causing it to "bite"
harder.

 

[EVTodd]

I may be wrong, but it seems the the drive plate slides due to forces caused the spinning of the knurled wheel.
The angle of the assembly is pointed into the wheel.
The further the slide moves the more it pushes into the wheel causing it to "bite"
harder.

That pretty much sums it up. When you apply throttle the motor is pulled by the roller into the tire. When you back off the throttle the tire pushes it the other direction. It works really well. The transition between full and partial throttle is very smooth. When you're at half throttle the roller is only partially pushed into the tire. With some adjusting of how much movement the slide has you could get just about any results you want. I had almost decided against trying the sliding mount but now I'm glad I did.

I still don't think it would work very well in the wet I guess with some super abrasive roller maybe but that's not an issue for me.

I'll get a couple of pictures of it mounted on the bike tonight. I also want to get a video of the thing working. That may take me a couple of days but I dug out my firewire card so I can transfer video to my computer (anyone know the best video editing software for Ubuntu?).

That's the first time I've seen that go-ped friction drive. It actually looks pretty well built for a low power setup. They mounted the motor in almost the exact location on the rack as mine. I wonder if it pivots into the tire.

 

[12p3phPMDC]

Well, quite ingenious EVTodd....

nice work!

I wonder how easy it would be to machine a roller using knurling tool
on a lathe? Would it have enough bite? Knurling tools are fun!! Whoever invented
those was pretty slick..

How deep are the teeth that you have Todd? 1mm? 1/16"?

It would also appear that the T plate you fabbed up keeps the drive from becoming
"cocked" in the slots...

It sounds like the action is smooth!!

It kind of reminds me of the "ways" on a milling machine table.

 

[spinningmagnets]

Here's a site that will sell knurled rollers in any one of 6 diameters, and two bearings, 3 parts for $40 (plus shipping). Its for gasoline weed-wackers, but...They should get smart and add a TNC electric motor option.

edit: the prices have gone up.

http://cyclehappy.com/parts_for_friction_drive.html

 

[Storm]

A single sided cone roller with the motor mounted to a sliding mount on a rack, CVT?

Any opinions weather or not a oneway bearing would work as a roller by itself on a skinny 700c wheel? I have never touched a oneway bearing so I don't know how tall or wide they are.

 

[Grinhill]

Any opinions weather or not a oneway bearing would work as a roller by itself on a skinny 700c wheel?

The one I am using is 3/4" diameter, bore to suit 1/2" shaft, about 7/8" long. The outer casing is smooth steel, may not be that great for grip.

The other issue that you would need to address is something to stop it wandering back and forth on the shaft, preferably something that had bearings or very slippery surfaces.

The cone looks like a neat idea if it will work

 

[Miles]

I think a straight pattern knurl might be better than a diamond pattern...

 

[paultrafalgar]

... I also want to get a video of the thing working. That may take me a couple of days but I dug out my firewire card so I can transfer video to my computer (anyone know the best video editing software for Ubuntu?).

Only one I have used is Kino. It's OK, but learning how to use it needs a tutorial - I expect you can find one by googleing.

 

[liveforphysics]

I think a straight pattern knurl might be better than a diamond pattern...

Agree 100%

 

[EVTodd]

I like the cone roller idea but with a sliding mount I don't think you really need it. If you have enough movement on the mount the roller will really bite into the tire anyway. Heck, with enough movement and an outrunner you could probably crush a tire or break the rack.

Spinningmagnets, The rollers from TNC are nice but I don't think they have one way bearings in them like the EV Warrior rollers.

As for the straight knurl instead of a diamond pattern. I actually tried that on my Kollmorgen drive (I made sleeves to make the ev warrior roller a bigger diameter). It worked fine too but the diamond pattern tends to track better and keep itself lined up perfect on the tire.

12p3phPMDC, thanks for the compliments. I wish I could take credit for the idea but Rkosiorek is really the one that showed us the moving mount concept. As far as it looking like the ways of a mill. That's exactly where I got that idea. I was sitting in front of one of my desktop CNC mills when I was trying to come up with a simple sliding mechanism.

 

[deecanio]

Hi EVTodd,

video would be great thanks.
It really is a corking idea, im very impressed at the more throttle more bite workings, super slick.
Looking forward to seeing it in action, for the next bike (if i do one) i'd like to keep it very simple direct drive, both yours and grinhills are looking like suitable ways to keep the setup simple and diy'able

Cheers,

Andy

 

[EVTodd]

Ok, I dug out my video camera and got Kino working. Here's a little video of the mount in action. I had to simulate a load by applying the brakes to make the mount move. If you watch closely you can see the mount move when the brakes are applied and slide back when I let go of the brakes.

Edit. I screwed up an deleted the video by mistake. I'll repost soon.

 

[12p3phPMDC]

The action of the sliding mechanism is awesome to watch...
The position is totally proportional to the load *edit* ...

This setup should get an award for the simplest design...

especially compared to the Rube Goldberg I've got

Great video and nice throttle...

 

[paultrafalgar]

While looking for an alternative source for drive rollers, I came across this:
http://www.motoredbikes.com/attachment.php?attachmentid=47&d=1185755395
I was particularly chuffed because this gent uses a capstan-shaped (waisted) drive roller which I suggested (and was pooh-poohed above :wink: ). You notice he states that you can use it in the wet!

"Thanks for the questions about the Zipcycle. As for the drive roller, we made extensive tests of every type of roller from knurled aluminum (too abrasive on the rear tire) to modified skateboard wheels. We sent a skateboard wheel off to a chemical lab, and found it was primarily polyurethane. We discovered that polyurethane can be compounded from any Shore Number (an industry measure of hardness) from rock-hard to gummy soft. We tested the Shore hardness of a street tire, and ordered polyurethane rollers that were just slightly harder than the tire compound (this was so that the majority of wear would be on the tire). The result was a roller that gave an excellent grip, but was long wearing. We found we could ride in the rain or morning dew without slipping.
The driver roller was suspended at both ends by sealed ball bearings. With the 25cc. Homelite string trimmer engine, most bikes would run from 23-25 mph. We experimented with different gearing and found that although we could run a higher gear and obtain slightly higher speeds, the bike had no reserve and the speed fell quickly on hills. So we compromised for a slightly slower speed and moderate hill climbing ability. Basically, the bike was rev-limited at the top end. We advertised a top speed of only 20 mph (that way no one complained). Tom"
Edit: quote from:
http://www.motoredbikes.com/showthread.php?t=6045
-----
Here's a pic:

That roller looks like the exhaust pipe (muffler?) hanger that the use to hold exhaust pipes to the chassis of cars. You could fix a roller bearing to each side of one of those and fabricate EVTodds mount to hold it, perhaps.

 

[paultrafalgar]

... and the magic phrase is "Polyurethane Suspension Bushes". For example, in the UK, I found:
http://www.bonaprene.co.uk/shop/4_Polyurethane_Suspension_Bushes
...maybe they have waisted ones?

 

[etard]

I definately give this a huge thumbs up for design of the year!! That sliding action is awesome, gonna have to copy this Todd, you know what they say about imitation!

And Paul, maybe Physics can explain why the outer surface would be traveling at a different speed than the interior on the conical roller It seems to me that its a ratio the stays the same just like shifting gears and changing cogs.

I do like the CVT idea, I wonder if we can refine this and have it tested somehow...... Todd, are you ready for a new project?

 

[E-Ride]

Really nice build you have. Can you show more pictures about your throttle?
It does turn out really nice on the video you posted. But what parts etc did you use?

Edit: I guess I'm after a EVTodd throttle howto

Cheers E-Ride

 

[paultrafalgar]

This is no way as good as EVTodd's mount, but some of you who are useless at engineering (like me) might like to adapt this mount:
http://bikemotorparts.com/

 

[EVTodd]

This is no way as good as EVTodd's mount, but some of you who are useless at engineering (like me) might like to adapt this mount:
http://bikemotorparts.com/

I see that place even sells the EV Warrior roller:

http://bikemotorparts.com/BEARINGROLLER.jpg

They must have been one of the places buying all the surplus rollers.

I'm currently at work on my phone but I'll get some more pics of the throttle up soon.

 

[E-Ride]

They have also a polyurethane rubber rolller 1.50" diameter for 36$
Better on rainy days maybe...

 

[spinningmagnets]

I would guess that polyurethane rollers would be easy to turn on a home-made simple lathe (to achieve the waisted shape to engage more of rounded tire-tread cross-section). I wouldn't hesitate to try it, but does anybody know for sure? (Pic shown is a ships capstan)

PS, EVTodd, you stated already you don't want to produce these, but would you be willing to post a simple blueprint? I suspect there are a suprising number of people who would buy these.

 

[deecanio]

Todd,

please re-post trhe vid, i missed it

D