A friction drive folding bike

[rearengine]

Ferias77 ?http://www.leaderhobby.com/product.asp?ID=9394001220364

Does this motor have a skirt bearing ?? Your build looks really great .. should be a fun ride.. Bill

[Quent]

This is shaping up to be a nice variation on the concept. Good work.

I really like the way you have locked the pivot axle to the swing arm.

I can see mechanically there are two more areas for you to attack, adjustable swing arm travel adjust and some way of neutralising the weight of the motor.

What do you have in mind here?

I have solve my servo tester problem, just a bad solder

Now, the bike works with a sandow (see pics)
but i have some springs to try also. firsts rides will be done with the sandow.

[ferias77]

This morning, first ride with the friction drive folding bike to go to work. Like Quentin, I put skateboard grip tape on the motor and it works very well : http://www.decathlon.fr/grip-skateboard ... 11158.html
I installed my cycle analyst, waiting for the Adrian interface (I am raring to test it !!! )

Top speed : 42 km/h
Mean speed : 26 km/h
Trip distance : 11.1 km
7.9 Wh/km, pedalling only during the hills
No load power with the grip : 37W

Now I have to adjust my geometry : The motor always touches the tire. I would like it doen't touches, as Adrian. But if i do that, it doesn't stay on the tire after starting...
And I have the impress that the tire is wearing down very fast...
A video soon !

Eric

[adrian_sm]

Woo hoo. First real ride, you must be pretty happy. Efficiency stats look nice. 7.9wh/km is great.

I am surprised that the skateboard deck grip survived. I tried something similar and the motor had too much torque, and just sheared the material. I now use something with a fabric backing to handle the shear forces, when I use grip tape.

As for tire wear it could be a few different things:
1) too high power. Your tire can only take so much torque before it shears the rubber apart
2) initial contact angle of motor and tire. Needs to climb the motor and not slip on first contact. Simple test put the rear brake on, light hold the motor on the tire, then try to turn the motor. In one direction it will slip, the other should climb the tire. If it slips in both directions, then you need to have a more perpendicular contact angle.
3) amount of tire/motor engagement. Not enough slips at high torque, too much and it distorts the tires across the motor causing additional tire wear.
4) drive angular alignment. The motor shaft may not be parallel to the tires rear hub, this will cause the motor to shear tire side ways. Have a close look at the rear tire as you engage the motor under power, and make sure it is not pulling the tire to the left or right, if it is rotate the drive on the seat tube and check again.
5) acceleration rate of the motor. If it ramps up too quickly it can rip through the tire too.
6) Tire pressure may have some impact, but I have done enough testing to understand it properly yet.

As for my interface, it is so close now. It will eliminate all the non-alignment issues mentioned above.
I am really happy with how it is turning out. The firmware side is pretty settled now, but I do need to clean up the hardware side a bit. Now that I have made and tested a few I have found a few minor things that need tweaking. I totally underestimated how much time and effort it would take to make something that functions so simply on the outside.

Looking forward to the video.

- Adrian

[ferias77]

Woo hoo. First real ride, you must be pretty happy. Efficiency stats look nice. 7.9wh/km is great.

You're right, I'm pretty happy, I didn't think it would work so fast. By contrast, Quentin has a problem with his motor which becomes very hot very fast (we have the same motor, but he has 6S batteries and me 5S)...

I am surprised that the skateboard deck grip survived. I tried something similar and the motor had too much torque, and just sheared the material. I now use something with a fabric backing to handle the shear forces, when I use grip tape.

I don't know how long it will last... But at lunch time, 6 km more, and it is perfect. The tape is very hard : it is impossible to remove !

As for tire wear it could be a few different things:
1) too high power. Your tire can only take so much torque before it shears the rubber apart
2) initial contact angle of motor and tire. Needs to climb the motor and not slip on first contact. Simple test put the rear brake on, light hold the motor on the tire, then try to turn the motor. In one direction it will slip, the other should climb the tire. If it slips in both directions, then you need to have a more perpendicular contact angle.
3) amount of tire/motor engagement. Not enough slips at high torque, too much and it distorts the tires across the motor causing additional tire wear.
4) drive angular alignment. The motor shaft may not be parallel to the tires rear hub, this will cause the motor to shear tire side ways. Have a close look at the rear tire as you engage the motor under power, and make sure it is not pulling the tire to the left or right, if it is rotate the drive on the seat tube and check again.
5) acceleration rate of the motor. If it ramps up too quickly it can rip through the tire too.
6) Tire pressure may have some impact, but I have done enough testing to understand it properly yet.

Thanks for all this information, I will make tests to see where the problem is. I will tie my gopro HD to the bike to see exactly what happens during engagement.

As for my interface, it is so close now. It will eliminate all the non-alignment issues mentioned above.
I am really happy with how it is turning out. The firmware side is pretty settled now, but I do need to clean up the hardware side a bit. Now that I have made and tested a few I have found a few minor things that need tweaking. I totally underestimated how much time and effort it would take to make something that functions so simply on the outside.
- Adrian

So it confirms what I thought : I hope I will have the possibility to have one ! A stand alone cycle analyst for sale soon

Eric

[ferias77]

A small video of the friction drive folding bike :

Nothing extraordinary, but it works !!!

Eric

[ferias77]

Yesterday, I did an other trip to go to work : 23 km in the day. I was very often at top speed, and pedaled less than the first day for the hills. Conclusion : 11.3 Wh/km and 32.1 km/h average. I modified the sandow, and now the motor doesn't touches the tire in free wheel, nickel ! But when there are bumbs on the road, the motor rotates and comes to touch the tire.

As for tire wear it could be a few different things:
1) too high power. Your tire can only take so much torque before it shears the rubber apart

Maybe, I don't know : the max power is about 1 kW, during the hills.

2) initial contact angle of motor and tire. Needs to climb the motor and not slip on first contact. Simple test put the rear brake on, light hold the motor on the tire, then try to turn the motor. In one direction it will slip, the other should climb the tire. If it slips in both directions, then you need to have a more perpendicular contact angle.

I verified, and it is ok, the motor climbs and doesn't slip.

3) amount of tire/motor engagement. Not enough slips at high torque, too much and it distorts the tires across the motor causing additional tire wear.

I modified the stops : now the motor can engage more in the tire, and I have the impress that the tire is less wearing down than before. But I have to test it longer to see if it is really efficient. Riding for a total of 55 km, I have already lost the middle of the tire rubber thickness

4) drive angular alignment. The motor shaft may not be parallel to the tires rear hub, this will cause the motor to shear tire side ways. Have a close look at the rear tire as you engage the motor under power, and make sure it is not pulling the tire to the left or right, if it is rotate the drive on the seat tube and check again.

I will try it, but I need to use the camera, otherwise I can't see because of the frame of the bike.

5) acceleration rate of the motor. If it ramps up too quickly it can rip through the tire too.

Tha acceleration rate is slow, thanks to the Hobbycity ramp module. But there is inevitably slip when I am in freewheel and I start the motor : the tangential speed of the motor and the tangentail speed of the tire are not the same.

6) Tire pressure may have some impact, but I have done enough testing to understand it properly yet.

The tire pressure is 3 bars, do you think it is ok?

As for my interface, it is so close now. It will eliminate all the non-alignment issues mentioned above.

I don't undestand, why do you say it will eliminate the non-alignment issues. These non-alignment issues are due to "mechanical " problems, isn't it? How could you solve them with an electronic interface?

Eric

[adrian_sm]

Maybe it is a translation thing. By non-alignment issues, I meant it addresses the issues that are not related to alignment, specifically the maximum power, and the acceleration rates. Additionally by enforcing a minimum speed before the motor is allowed to start (in addition to the controlled acceleration rate) it is much more gentle on initial contact with the tire.

I think it is probably a combination of tuning the alignment/deadstops, and limiting the power. From you video it sounds like you hit the throttle at very low speed, then the motor looses sync/or is bouncing into/out of the tire. This is not good. If it is loosing sync then you asking the motor to apply too much torque at too low a speed. If it is bouncing off the tire, then either the controller is cutting power, or alignment is not right. I would wait to you are move a little faster first, maybe 6kph or so and see if you still get the problem.

Next I would look at how quickly the throttle is ramping up. I think I had it going from 0 to full rpm in about 2 seconds when I sued the hobbycity ramp module.

- Adrian

[ferias77]

You're right, it is a translation problem, I understand better now. I will try applying power moving over 6 km/h. For the ramping up, I verified this morning, and it's ok : about 2 s from 0 to full thottle.

Eric

[ferias77]

My folding bike has now made 77 km with the friction drive. It works, but there are some problems to solve :
- the motor doesn't engage enough in the tire : it disengages and re-engages when there are bumbs on the road. It wears down the tire and the grip of the motor.
- a problem I didn't see : Adrian, your swing arm is almost vertical, it isn't the case for mine and it increases the problem of disengagment when there are bumbs.

I knew it wouldn't be perfect at the first trial. I designed the swing arm to have the possibility to drill other holes to modify the geometry, so I will do it asap. And I will see if it is possible the swing arm to be more vertical...

Eric

[adrian_sm]

I am surprised it disengages when under power. The motor torque should be enough to ensure it doesn't pop out.

My design has a torsion spring that you can adjust to take the weight of the motor. So once adjusted it only takes a light force to lift the motor. Do you have something similar?

[EDIT] I just checked previous posts for images, and noticed your elastic rope used to keep the drive engaged, and no apart deadstops. So I thought I should mention I also have the deadstops that I adjust so that the motor just doesn't touch the tire (checked with a full rotation of the tire, as some are not round). This also limits the possibility of the motor bouncing around too much.

- Adrian

[ferias77]

Yes, it disengages under power, so it confirms that the motor has to engage more in the tire. I will make the swing arm shorter to do it.
Your torsion spring is replaced by the sandow to take the weight of the motor. I use 2 pieces of string to make the stops, but it is too flexible. I will make deadstops with aluminium plate, it should be better.

Eric

[ferias77]

I modified my motor mount to try different geometries, I put it on the bike asap :


I hope it will solve the problem of wearing down the tire. I made 220 km with the first tire, and it is completely destroyed...
I changed it, I will see with the new geometry.

[ferias77]

Yesterday, I tested the bike with the modified swing arm. The motor is now more close to the pivot and it works better
But at the evening, I made 200 m to go to home and the motor has made a very strange noise... Impossible to start it again...
I have tested all the components today : the controler and the servo tester are dead...
When I had the problem, I was about 30 km/h and 400W, so not too much. Any idea of the reason why?

Eric

[adrian_sm]

I have only killed a controller as a result of a seriously overheated motor that had shorted windings. But I am always going above 6kph (more often more than 12 kph) before I use the drive.

If the motor made a strange noise, it may have lost sync, and that may have helped the controller kill itself. Low speed, high load is the worst scenario for the ESCs.

Can you tell us any more of what you were doing before the controller died?

[ferias77]

Thanks for your reply Adrian !
I started pedalling and when 10 km/h reached, I began to use the drive. I limited the power to about 400W (manually, with the throttle). I was very surprised bacause I already made really harder starts...

Eric

[ferias77]

Hi,

My friction drive folding bike is alive ! I changed the controler and the servo tester, modified the controler parameters following the commuter booster website and installed the cycle analyst on the saddle tube to put the speed sensor on the rear wheel. So I made a test to see the no load speed : 59 km/h !!! The motor is given for 200 kv, it looks to me more 250 kv...
Could it explain why I killed the controler, riding too long at partial throttle?
Last week, 29 km under the rain to go to work and no problem.
Yesterday, I made a 17 km ride with 2 5S 4900 mAh Rhino batteries : average speed = 27.2 km/h, top speed = 40.1 km/h and 9.5 Wh/km, pedalling only from 0 to 10 km/h, no pedalling after.
I have to ride with it now, but I have the impress that the tire doesn't wear down like before with this new geometry !





Eric

[adrian_sm]

Sounds like it is all working really well now. Glad to hear it.

You should be able to lengthen the speed sensor cable on the Cycle Analyst to allow you to install it up front on the bars.