Another friction drive

[Kepler]

Its probably a good precaution on any RC motor but that being said, I have never glued the magnets in my 63-74 test motor (or replaced the bearings) as I wanted to see how they went. After heaps and heaps of abuse, happy to report that a magnets has never come loose nor have i had any reports of others coming loose. Bearings are getting a little noisy now but still not too bad.

 

[drifter]

Just got some parts for my new drive build and made a mock-up to take measurements. Could someone tell me how tight to tension the timing belt.

 

[spinningmagnets]

Is there a freewheel clutch bearing inside the motor pulley or roller-pulley? That is very close to the design I was going to make as a roller test-rig (but yours is better than my mental picture), as I have several different rollers I want to try out. I don't know yet about proper belt tension...

In spite of the added cost and complexity, I think having the belt-reduction opens up a whole new world of options as far as motor kV/system voltages, and also available motors. You won't need a skirt bearing, and the diameter of the motor is no longer a consideration
http://endless-sphere.com/forums/viewtopic.php?f=28&t=9652&start=285#p300012

Since the roller takes a lot of stress, I would recommend that the motor is the part that should have mounting slots to slide the motor up. I have seen an idler wheel successfully used on the slack side of the belt a few times, but I prefer the sliding motor mount-holes. Looks like a good start!

 

[drifter]

No, no freewheel clutch, the idea is to pivot the whole unit at the seat post / handle bar stem so it swings onto the tyre like Kepler's and my earlier build and others.
The pulleys are 20T and 30T and the roller is 1.25" so that gives me two possible gear ratios already.

In spite of the added cost and complexity, I think having the belt-reduction opens up a whole new world of options as far as motor kV/system voltages, and also available motors. You won't need a skirt bearing, and the diameter of the motor is no longer a consideration

Exactly what I was thinking and also might lessen the risk of loose magnets.
I agree with you about having motor adjusting slots but I don't know how tight to tension the belt.

 

[Hillhater]

I agree with you about having motor adjusting slots but I don't know how tight to tension the belt.

In industrial applications we tension them up like a drum skin !
.. but for this i would not be that tight, . No slack, maybe a few mm movement on the belt when pressed with your finger in the middle of the longest run.

 

[spinningmagnets]

I also 'suspect' that since you now have a smaller roller diameter, once you add a motor with a higher kV (now an option due to the belt-reduction) you will have fewer sync issues and a noticeable reduction in standing-start amps, including reduced overall peak amps.

I'm glad you are doing this, you have come up with some clever solutions recently and I am curious to see how this turns out. Best of luck!

 

[Kepler]

Thats a very long pivot moment you have on that setup. Might make the geometry a little difficult to setup with such a big arc. Something to keep in mind. In relation to tension, I used one of the many available belt length calculators to come up with an exact distance between shaft centres for the selected belt and pulleys used.

 

[Miles]

I agree with you about having motor adjusting slots but I don't know how tight to tension the belt.

See section10 in the attached doc.: http://www.endless-sphere.com/forums/viewtopic.php?p=126539#p126539

 

[drifter]

Thats a very long pivot moment you have on that setup. Might make the geometry a little difficult to setup with such a big arc.

Yes I wondered about that but that's the shortest I can make it on this bike unless I get an adjustable stem. Anyway I'll give it a try.
Spinner, also I have plenty of room between motor and roller to allow for different size pulleys.
Thanks Hillhater. Actually I like hills but only when going down.
Thanks Miles.

 

[GGoodrum]

You could also shorten the "moment" by putting the motor above the pivot point.

-- Gary

 

[kevo]

You could also shorten the "moment" by putting the motor above the pivot point.

-- Gary

Gary (and Kepler),
This is where I feel ignorant of the auto engage friction drive. Below the pivot point, centrifugal force (or is it centripetal?) http://en.wikipedia.org/wiki/Centrifugal_force causes the motor to engage to the wheel. Wouldn't force go in the opposite direction above the pivot point?
Sorry feeling ignorant,
Kevo

 

[spinningmagnets]

As I mentioned before, I had planned on making a similar set-up, though the one big difference was that I would leave the roller in contact with the tire, with just a tiny amount of rocking-arm tension (weighted so that it wants to rub against the tire gently). Once the motor engaged, the drive would pull itself into the tread (into a certain angle, held back from engaging further by a stop-cable/arm) and at that point the drive wouldn't bounce around anymore.

As Todd established, to keep the roller in constant contact with the tire tread, a free-wheel must be employed in some part of the drive-line. The clutch-bearings I am using have a 3/4" OD and slip over a 1/2" shaft. I have found they fit snugly inside the SDP-SI pulleys with a 5mm-HTD tooth. They are cast plastic with a metal center. (thanks, Grinhill!)

I suspect they would hold up at lower powers with the factory interference fit, but if it proved to be slightly too loose, you could epoxy them in ($10 bearing, and $20 pulley), or perhaps apply solder to the bearing OD to snug-up the interference-fit a little more. This would make your shaft pulley freewheel, and the belt/motor should remain fairly idle when just pedaling.

Consider the $6 polyurethane boat-trailer keel-rollers with a 1/2" bore...I've found them locally that are 3-1/2" wide, so they fit inside my 4" wide housing (pulley and freewheel on the outside of the sidewall)

 

[drifter]

G'day Gary, like Kevo I wondered which way the motor would "kick" if the motor was above the pivot and my brain began to hurt so I made a mock-up and it does indeed go the correct way to push the roller against the tyre. Now I might rethink my design again. I will have to ditch the handle bar stem and try something else.
Bugger, my brain is starting to hurt again. Only joking, I am getting so much enjoyment out of this project. Thanks all for the continued input.
Robbo

 

[spinningmagnets]

"...I will have to ditch the handle bar stem and try something else.

WHAT!!!...(grumble grumble...) Your stem-mount is what put my current prototype on hold!...I WAS hoping that YOU would work out the bugs. Big box from China arrived. Some parts arrived separate in small boxes, and the big box flew the midnight planes from China to New York, then back to Kansas!

Drifter, consider triangular side plates. The stem-mount pivot could be the top point of the triangle, the front point is the motor, and the rear is the roller. If the rear edge is vertical, the weight of the motor would gently push the swinging-icity (engineering term) into the tire, but the tires forward motion would hold it off.

If the triangular side plates were tall, you could have a line of optional pivot mounting-holes, so others could adjust the height of the drive to match their frame, then they could trim off the excess for a more compact fit. Such a drive would not interfere with the addition of a very useful rear cargo rack. I might also add an additional channel around the roller to remove any flex and twist.

Edit: I have settled on a certain adjustable handlebar stem (well,...at least for today) that uses a 1-1/8" MTB steer-tube quill-clamp, (same as many seat-post diameters, with many available adapter-bushings to other seat-post diameters), and a 31.8mm bar clamp (which allows me to use a very available and common size of MTB seat tube for my cross pivot-bar)

 

[drifter]

Now I am really undecided, another possibility.
Your Idea sounds good, I will soon have a room full of mock-ups.
I am waiting for a new motor to arrive since I broke one of the magnets while cleaning off the old epoxy.
Maybe the magnet will be ok if I just put it back together and glue it back in place.

 

[drifter]

Spinner, I just thought I should say that the stem idea credit goes to Hillhater.
And don't hold up your build waiting for me because I tend to change my mind daily.
Of course I will give as much help as I can from down under but remember everything is upside down here.

 

[Hillhater]

Drifter, either you have a very tall bike frame, or a very short stem on that seat post. !
Try finding a longer stem with more "rise" , then mount it upside down to drop the bar clamp nearer to the tyre. :wink:

 

[kevo]

G'day Gary, like Kevo I wondered which way the motor would "kick" if the motor was above the pivot and my brain began to hurt so I made a mock-up and it does indeed go the correct way to push the roller against the tyre. Now I might rethink my design again. I will have to ditch the handle bar stem and try something else.
Bugger, my brain is starting to hurt again. Only joking, I am getting so much enjoyment out of this project. Thanks all for the continued input.
Robbo

Robbo,
Did the mock-up include a wheel and if so was the wheel rolling in the forward direction? I
Thx,
Kevo

 

[kevo]

Spinningmagnets,
I like the keel roller idea and look forward to hearing how well it performs.

 

[drifter]

Kevo, I double checked just now and it turns the right way.
Hillhater, yes I have a very tall frame.

 

[Kepler]

You could also shorten the "moment" by putting the motor above the pivot point.

-- Gary

Gary (and Kepler),
This is where I feel ignorant of the auto engage friction drive. Below the pivot point, centrifugal force (or is it centripetal?) http://en.wikipedia.org/wiki/Centrifugal_force causes the motor to engage to the wheel. Wouldn't force go in the opposite direction above the pivot point?
Sorry feeling ignorant,
Kevo

Kevo,

The problem that can be encountered with too large an arc through the pivot is that the roller ends up not being able to climb the tire effectively. The roller movement needs to be towards the tire but also follow the tire in an upwards direction. If the arc is too large, the movement ends up being horizontally biased and can cause the roller to bounce off the tire. I spent a lot of time getting this pivot distance correct. Between 50mm and 60mm I found worked best.

 

[kevo]

Drifter, spin and Kepler,
Appreciate all the help and feedback. My interest in the motor above the pivot would be to allow a rack as spin points out or to place the motor in the seat/wheel tube triangle and the unlikely possibility of placing the roller between the seat tube and wheel or even move everything down in place of kickstand.

 

[GGoodrum]

Kevo, I double checked just now and it turns the right way.
Hillhater, yes I have a very tall frame.

Yes, that is what I meant. Put the pivot point between the motor and the roller. That way you can get the 50-60mm optimum pivot distance. Why would this not still work with the stem?

-- Gary

 

[Hillhater]

Yes, that is what I meant. Put the pivot point between the motor and the roller. That way you can get the 50-60mm optimum pivot distance. Why would this not still work with the stem?

-- Gary

I think his problem is a tall frame, and finding a stem that will place the pivot point close enough to the tyre. Hence my suggestion of using a long stem ,with a big "rise" , but mounted upside down.

 

[spinningmagnets]

This is the variation I was imagining a few posts back after seeing this great idea you showed. The roller would sit inside a squared channel, and the triangular side plates would also act as the bearing retainer plates.

I have one of the SDP-SI "shaft-loc" expanding bushings. The roller shaft (1/2" dia) would have the clutch-bearing (outside the sideplate) and the motor shaft would have an adapter sleeve to beef up the shaft dia (10mm or 8mm) to 1/2" so the shaft-loc could attach the motor shaft to the 5mm-HTD pulley.

http://www.endless-sphere.com/forums/viewtopic.php?f=28&t=8491&start=15#p301680

In spite of the obvious added complexity and cost, it allows the use of the higher kV 8mm-shaft motors (300kv-400kV), and that is somewhat of a sweet-spot for using a single 6s LiPo brick (22V, 5800-mAh, 30C).

Once thats in place, there are over twenty 5mm-HTD pulley diameters that will work, allowing anyone to choose exactly the top speed they want. or, to later change the top speed (quite easily) by simply purchasing a third pulley to swap in. These motors are small, but the direct-drive I am working on has the motor poking out to the side, and your idea puts the motor squarely in the center of the drives mass.

The stems (used as the mount) can be angled up or down, and can be bought long or short. If a customer wanted to change the drive to a different bike-frame, there are 38 pages of bike stems at Niagra to allow a rapid, cheap, and easy re-fit.

If you decide to make something similar to this drawing, perhaps make the belt-side plate wide enough all around so that later you can add a belt-cover? then, at that point you can trim off the excess.

It may even be possible to use an aluminum channel (4" wide, 0.250" wall thickness), with the part that connnects the two side-plates being on the back of the drive?