Calling all Belt Drive Experts

[Skaiwerd]

Can you design your bike motor to not need an idler? I'm planning to avoid them if I can.

It looks like the same 825mm belt can be used both with and without an idler. And the belt is only $10.98 at V belt supply!
Belt Length a.jpg

FZBob

Nice Cad work. I forgot about the black background. Do you have a center to center (motor pulley center to crank center) distance for the version without the tensioner? And A big thank you for the CAD file. We will have new pulley designs soon. I always like printing my own designs the most. The printing drives my designing actually.

 

[FZBob]

lets try it

It feels very strong. I had to slow down the print infill and put tape on the bed leveling screws as the printer shakes like crazy when printing the teeth on the pulley. A good challange though nonetheless. So one failed print attempt. Actually printed in 7hrs 49 min. Remember larger print layers are stronger and smaller layers look better. Some clean up/detailing at the teeth of the pulley may be required.

FZbob, PM address i'll mail it to you.

I just received the 120 tooth 3D printed pulley from Skaiwerd. Many Thanks! Also thanks to Qwiksand for the design!



I ordered a couple of belts which will be here soon. I also ordered a 12 tooth motor pulley from England, which will be here after the first of the year...

I would like to find a second (faster) source for the 12 tooth pulley. Any suggestions?

 

[FZBob]

Nice Cad work. I forgot about the black background. Do you have a center to center (motor pulley center to crank center) distance for the version without the tensioner? And A big thank you for the CAD file. We will have new pulley designs soon. I always like printing my own designs the most. The printing drives my designing actually.

Thanks! Sorry I missed this post earlier...

Center to Center is 9.020" without idler, and 8.520 with idler (depending on idler placement). With 825 mm belt.

 

[FZBob]

I received the 12 tooth pulley from Rubber Belts on Line. https://www.rubberbeltsonline.com/6-35mm-Bore-Aluminum-Timing-Pulley-5mm-Pitch-12-Te-p/6.35mm-12t-15mm-htd5m-pulley.htm I bored and reamed the shaft hole from 1/4" (6.35 mm) to 8mm. This pulley comes with two set screws and no keyway. I need to grind a couple of flats on the end of my motor shaft for the set screws. I will try running without a key on this pulley.

Also made a clean up cut on the central bore and 4 mounting holes of the 120 tooth pulley that Skaiwerd printed for me. There was about .005" to .010 of material allowance remaining for the cleanup, which was pefect for me!



View attachment 1



Test fit on the Shimano crank arm went well. However, I'm going to use the RaceFace crank arm which is a lot narrower, and may interfere with the pulley. It looks like I may need to mill a few mm off the inside of the RaceFace crank arm.

Belts should arrive in the next few days...

 

[Chalo]

Test fit on the Shimano crank arm went well. However, I'm going to use the RaceFace crank arm which is a lot narrower, and may interfere with the pulley. It looks like I may need to mill a few mm off the inside of the RaceFace crank arm.

Yikes! You know they didn't make the thing thicker and heavier than they thought was necessary, right?

 

[fechter]

Yikes! You know they didn't make the thing thicker and heavier than they thought was necessary, right?

Who actually pedals around here?

One issue you are likely to run into is belt skipping. From experience I know that a small pulley driving a large one works well up to very high power levels. But during regen, when the large pulley is driving the smaller one, the belt had a bad tendency to skip. More belt tension and a larger angle of wrap are about the only things you can do to minimize this.

 

[Chalo]

Yikes! You know they didn't make the thing thicker and heavier than they thought was necessary, right?

Who actually pedals around here?

It's true that pedaling seems to be a optional feature for some of us here. But the highest force most people put on their pedals comes when they stand up on the pedals and ride over a bump. Any of us might do that if we're not on some godforsaken wheelchair bike.

But during regen, when the large pulley is driving the smaller one, the belt had a bad tendency to skip.

Regenerative braking on a crank drive? That would require fixed gearing, no? It sounds like a terrible mistake.

 

[FZBob]

Yikes! You know they didn't make the thing thicker and heavier than they thought was necessary, right?

It's what they call an Interference Fit...

This one's on me, and was not totally unexpected. It's purely a belt width issue. There is about 14.5 mm from the inside of the crank arm (outside of the pulley) to where the inside of the pulley interferes with the bike chain. I could have gone with a narrower belt. but I want to give it every chance for success.



Not the best pic, but you can see the pulley hitting the crank up top, and not meeting up with the spider down at the bottom.

I think I'll go with the Shimano square taper crankset "For Testing Purposes Only", as it has space to spare. It's just heavy and wide. If the belt drive works well and is quiet, then I'll trim things here and there to make the RaceFace crank work...

 

[FZBob]

From experience I know that a small pulley driving a large one works well up to very high power levels. But during regen, when the large pulley is driving the smaller one, the belt had a bad tendency to skip. More belt tension and a larger angle of wrap are about the only things you can do to minimize this.

Good to hear. No regen, so hopefully no problem.

I'm going to try without an idler first (because I'm lazy), and if the belt skips, I'll install an idler.

 

[fechter]

My experience comes from my old Zappy scooter. No pedals involved, but during regen the larger sprocket drives the smaller one.
I'm just pointing out that anytime a larger sprocket is driving a smaller one, skipping is more of a problem.

 

[Skaiwerd]

Nice work...

Glad that pulley is getting some dialogue going. I just picked up qs1000w mid drive motor. It came with the larger 8mm pitch pulley installed. For skipping reasons I think it needs to be this size. It’s for a 15mm wide belt. I didn’t count teeth but it’s around 2 inches in diameter. So some quick figuring for me is a 1:10 ratio would be 20 inches in diameter pulley on the rear wheel. Can’t print that big, yet. I’m still figuring what I need for the reduction ratio. Trying not to do a two stage if I can help it.

 

[fechter]

Minimum pulley size is another issue. Bigger pulleys have more teeth so can handle more load without slipping. The minimum recommended size is specified for a given belt type, but that is based on bending radius and not damaging the cords. You may need to use more teeth to get satisfactory results. There are formulas to calculate most of this stuff.

 

[Grantmac]

Nice work...

Glad that pulley is getting some dialogue going. I just picked up qs1000w mid drive motor. It came with the larger 8mm pitch pulley installed. For skipping reasons I think it needs to be this size. It’s for a 15mm wide belt. I didn’t count teeth but it’s around 2 inches in diameter. So some quick figuring for me is a 1:10 ratio would be 20 inches in diameter pulley on the rear wheel. Can’t print that big, yet. I’m still figuring what I need for the reduction ratio. Trying not to do a two stage if I can help it.

Why not use a 20" BMX rim with sections of printed cogs? Not like it would skip with that big of a pully.

 

[FZBob]

The belt has arrived! Looks cool as H^ll!



Length looks good. Do these belts "break in", ie become less stiff? It took a lot of tension just to make it look straight...

Motor shaft needs flats for the set screws, and then I think I can try it.

Just by spinning the pedals, the belt appears to not wander off, even though there is no outer retention flange on the 120T pulley. I may try it without an outer flange to see if it works. Thoughts?

 

[Chalo]

So some quick figuring for me is a 1:10 ratio would be 20 inches in diameter pulley on the rear wheel. Can’t print that big, yet. I’m still figuring what I need for the reduction ratio. Trying not to do a two stage if I can help it.

Why not use a 20" BMX rim with sections of printed cogs? Not like it would skip with that big of a pully.

A 20" pulley built that way would require a 24" bicycle rim.

I remember in the early days of e-bikes someone turning a toothed belt inside out around a sheave to make a jumbo sprocket.

 

[falcongsr]

That does look pretty sweet

I had been researching 3D Printers for some time, but this thread put me over the edge and decided to order one. I'm probably many months away from being able to print something useful for an electric drive, but it's now a goal.

 

[fechter]

The belt has arrived! Looks cool as H^ll!

Length looks good. Do these belts "break in", ie become less stiff? It took a lot of tension just to make it look straight...

Not much. They stay pretty consistent over their life. That ratio looks pretty close to what I had on my old Vego. Proper belt tension is quite a bit.



Are you going to have a freewheel between the pulley and the crank? Might be dangerous if the motor runs and the feet are not in sync.

 

[Skaiwerd]

FZBob
That does look bad ass, sick what have you. The best part is the simplicity and compactness. I was worried the belt might not engage in the teeth of the pulley that I printed for you. Glad it does! Things can grow a bit when 3d printing, if your flow rate is too high or you’re printing squished layers.

falcongsr
Get a printer! Design and print your own parts. Previously I mentioned a good recommendation, the alfawise u10. For a cheaper but smaller model the u20 is also good I’ve read.

Chalo and Grantmac
Still up in the air about the amount of reduction ratio. I’ve read of using a bike rim. One old school method actually used a bike rim and somehow a belt was used with epoxy to mold the pulley using the belt itself to form the teeth, then removed. Sounds dirty. Once you start 3d printing the old dirty way seems so troublesome.

Fechter
The non freewheel method will allow this belt setup on most bikes. The crank flywheel would require replacing the bottom bracket so the freewheel has something to thread onto. Maybe a freewheel at the motor is easier if needed.

 

[FZBob]

Not much. They stay pretty consistent over their life. That ratio looks pretty close to what I had on my old Vego. Proper belt tension is quite a bit.

Vego belt drive.jpg

Are you going to have a freewheel between the pulley and the crank? Might be dangerous if the motor runs and the feet are not in sync.

That looks like a cool toy! 5mm pitch belt? Do you happen to know how many teeth on the motor pulley? How many watts?


No freewheels for me!

Eliminating the freewheel makes the assist extremely nice to use. Some trails that I ride require constantly varying accurate power levels, both leg and motor. As my motor is always spinning, I can bring on power very smoothly and precisely. The power is just there - instantly. There is no backlash and "clack" as the freewheel takes up, which tended to upset the bike in the technical stuff. I had a freewheel on the initial build, and I would never go back!

Also, parts which are not there don't break or add weight. The motor drag when pedaling is only about 6 watts!

I have about 500 miles of technical mountain biking with the chain drive running direct, and no real problems. A couple of instances where I was clocking the pedals to get through rock gardens, got a little uncoordinated, and I could feel the pedals push a bit, but my legs can easily stall the motor. But that is out of hundreds of chances... That being said, I would NOT run direct with any more power than I have.

I am very cautious when the bike is in the work stand with the battery installed, as I'm sure the pedals could break a hand or crack a skull. Also, when I'm walking up steep loose hills and using a bit of power to help, I have smacked my shins a few times - derp...

Time for the obligatory Internet disclaimer - Don't try this at home! You could bruise a shin, get tossed over the handlebars, loose control and hit a car, burst into flames, etc. It's the end of the world as we know it... and i feel fine...! (REM) :lol:

 

[Chalo]

One old school method actually used a bike rim and somehow a belt was used with epoxy to mold the pulley using the belt itself to form the teeth, then removed. Sounds dirty. Once you start 3d printing the old dirty way seems so troublesome.

I've been making things out of metal for so long that I don't think I could settle on using filament-fed plastic for anything structural. I think I'd be happier in most cases with laser cut birch plywood. Or Durham's Rock Hard Water Putty. Whatever's stronger and stiffer than squiggled-together PLA or nylon.

It's like if I made a wish to a genie, and it was to be able to make anything I want, in any shape. And the genie says, sure-- but it can only be as big as half a loaf of bread, and it can only be made out of tree sap. I'd think, well... um, okay, better than nothing, I guess....

 

[Skaiwerd]

Chalo

With regards to 3d printing...

...“it’s just another tool in the toolbox”.

I’m looking forward to wheels and frames being printed on larger faster printers. Prototype, finished product, molds, jigs for welding, etc. It’s why we’re here to try new things and push limits. Any method that works is a good one.

 

[fechter]

If you wanted to pedal with no assist, there would be a lot of drag from the motor. That's not a big deal as long as your battery has juice. What's more of a problem is if your feet come off the pedals while it's powered, it might hurt. A torque sensing PAS system could possibly solve that issue, similar to the Bosch system.

 

[Qwiksand]

@Skaiwerd:

Awesome work on that pulley, very cool of you to send that out!

@FZBob:

Another idea for a cheap/easy belt idler without having to modify your existing motor mount, just need to pick-up 3 624 Bearings (4mm Id, 13mm OD, 5mm width)and use one of your existing motor mounting bolts like this:



Now if we bring this design to it's logical conclusion, I think we need to eliminate the belt/chain completely. This design below is a 10:1 reduction using off the shelf 12t 20dp pinion gear from Vex Robotics and a 3DP 120t 20dp crank gear. Design is shown with a 36t chainring:




Now, of course, this exposed gear system will likely not last long in a true MTB environment. Enter the integrated sprocket/internal ring gear concept. This design would be much easier to enclose, but real world fabrication will be challenging:

 

[Skaiwerd]

Qwiksand

Nice work especially the 2nd one where the gear teeth are on the inside/back of the chain wheel.

With the integrated chainwheel metal would be best. Could 3d print the dust/weather shielding to keep costs down. I’ve been following this guys thread and this amazing swingarm for his Vespa that he sent out to 3dhubs to get Cnc’d for $450, it’s a piece of art.

https://endless-sphere.com/forums/viewtopic.php?f=10&t=96356

Go to the end of the thread...

I want to upload some files to check prices myself.

 

[FZBob]

If you wanted to pedal with no assist, there would be a lot of drag from the motor. That's not a big deal as long as your battery has juice.

Umm... Not so much...

I'm far more of a biker than an E-biker. (If I had to choose between my new 7" dropper post and the E-Assist, the motor would be gone in 5 minutes!). I have 500 miles on this drive now, and 50,000 feet of climbing. I pedal 95% of flats, slight uphills, some technical uphills, and all downhills without assist. If spinning the motor with no assist was draggy, I would NOT do it. Just to calibrate my sensitivity to drag - when I bought the bike new, it came with a Kenda Nevegal rear tire. Very draggy in back. I got rid of it fast. You can find data showing the Kenda "Nevergo" is about 20 watts more draggy than the low drag back tires. For me 20 watts of drag is very obvious and totally unacceptable. My measurements indicate the motor has about 6 watts of drag. I really can't feel it! I was shocked too!

I hope that didn't come off too rant-y...

If you look at the first picture in my build thread, https://endless-sphere.com/forums/viewtopic.php?f=28&t=96769 you can see I initially had a freewheel on the left side of the crank. I too believed the motor would be draggy, so I spent a lot of money and effort to freewheel the motor. Then I made an accidental discovery, and ditched the freewheel. The drive integrates so much better now in terms of throttle response...

As always, YMMV...