Commuter Booster - <1kg Friction Drive

[rearengine]

Adrien! Very interesting thread, Pictures worth a 1,000 words... I've been following your build all along , with Keplers, EvTodds , and a host of others.. I'm still working on Rearengine Hill Helper, friction drive With a whole lot of help from Spinningmagnets.. Enough of that on your thread... Bill

[deVries]

Back from holiday now, was an awesome trip to Cradle Mountain in Tassie. Here is a few pics from the trip.

Thanks for the pics... beautiful.

Good find on that 150burst amp ESC... somehow I missed that one, and it should give you plenty of headroom to not fry it. Will you add caps?

P.S. @Kepler.
John, you have to admit you are a bit of a tease. Updating your avatar with unreleased products. Dropping tasty little tid bits about your throttle interface, without let people know when they may be able to get their hands on it. Come on already.

Please note, he has now de-updated the avatar putting his two speed roller design, I guess, in the top-secret cave??? ...Hmmm.

Well, just open a thread on it, and be done with the surprises in store. After all, ES is a prototype, feedback, open-source idea community.

At least get the throttle thread going soon???

[Kepler]

P.S. @Kepler.
John, you have to admit you are a bit of a tease. Updating your avatar with unreleased products. Dropping tasty little tid bits about your throttle interface, without let people know when they may be able to get their hands on it. Come on already.

At least you know I am not sitting on my hands doing nothing
Just a quick up date (I don’t want to hijack this fine thread) The dual speed setup is on the back burner at the moment. I really question if two speeds are necessary for a friction drive that is used an assist as apposed to a primary drive. However, I do have some designs drawn up for future development.
The latest design has grown a little so it can accommodate the battery (5S 5ah) as a complete package. I had been trying to make the drive as small as possible but I have now moved a little in the opposite direction as I think the 20% extra bulk is worth it seeing you no longer need some method to carry the battery. Also the battery gets well protected between metal side frames.

Interface boards are being built at the moment. The main processor is now mounted in a small box on the stem or the handle bars. Accessories are then plugged directly into this box. Lots of flexibility here including options for LCD screens, simple low cost SunDing bike computers, throttle, brake, pedlec, voltage taps up to 18V to power lights and accessories. There is a second slave board back at the drive that takes the ESC input, current sense, and volt pickups.

Enough of cluttering up Adrian's thread (you did ask though ) I will get a stand alone thread running in a few weeks time for some serious discussion on this stuff.

Adrian, offer is still open to test ride the button throttle interface.

[Hillhater]

This time I went with the $60 TURNIGY K-Force 100A Brushless ESC,
But it does apparently have USB programmability, and firmware upgrade. But I don't see no USB port on it. From reading the comments on the HobbyKing site it looks like you need to buy some seperate progammer, that also allows you to connect it to the PC via USB. Great. How handy.....

Adrian, you are aware you can reprogram all the key variables on that ESC using just the servo tester/throttle actions ??
Its explained in the handbook.

glad you found the throttle mod useful too.
HH

[damonjackson_spl]

I thought i recognised the pics of cradle mountain!! beautiful walk..

the worst thing in, cradle mountain is only a 3.5hr drive from hobart and most of us have not done that walk :O

pity as it is a great spot!!

i think the friction drive would love those winding timber tracks!!

D

[adrian_sm]

Instead of programmer you can use USB linker for 8$.
http://www.hobbyking.com/hobbyking/stor ... duct=10628

Thanks for that Hal. That is cheaper, but I guess I was just expecting a mini-USB port right there on the ESC.

Adrian, you are aware you can reprogram all the key variables on that ESC using just the servo tester/throttle actions ??
Its explained in the handbook.

Yep. I used it to originally set-up the ESC, but found it annoying. It is slow, requires you to have the manual right there with you to make sense of it all. Generally just a royal pain. That is where the Pro Programmer Box I linked too would actually be worth it. As it displays all the settings nicely on a screen, and you can then reprogram quickly and easily in the field. Great for trying out different settings quickly while out on the road.

The cheaper USB adapter thing Hal linked too, would be great if I had a computer in the shed where I park the bike.

I'll probably stick with the throttle adjustment of settings, as it is not something you change very often. But thanks for everyone's input.

- Adrian

[adrian_sm]

Adrien! Very interesting thread, Pictures worth a 1,000 words... I've been following your build all along , with Keplers, EvTodds , and a host of others.. I'm still working on Rearengine Hill Helper, friction drive With a whole lot of help from Spinningmagnets.. Enough of that on your thread... Bill

......
Enough of cluttering up Adrian's thread (you did ask though ) I will get a stand alone thread running in a few weeks time for some serious discussion on this stuff.

Adrian, offer is still open to test ride the button throttle interface.

Thanks guys. Don't feel like you are cluttering the thread. It is great to have others drop by. Especially those working on similar set-ups, and new innovations.

Keep up the good work.

- Adrian

P.S. John, I will take you up on that offer one day, but I am struggling to find time on weekends, and the few spare minutes I do get I want to focus on getting my drive up and running.

Then I'll drop around, and we can compare. Or more likely you can help me fix it, when I burn something else up.

[adrian_sm]

Major Update: Commuter Booster - Proto #2

Amazing what you can get done, when you find a spare 4 hours to work in the shed.

Behold Commuter Booster - Prototype #2
- Turnigy 6374-200
- single swing arm design
- easily adjustable (spring preload, engage deadstop, and disengage deadstop)
- suits 34.9mm seat-tube.
- weighs in at 1.2kg, 0.9kg of which is the motor itself. Not bad since I have made no effort to shave weight down.

1.2 kg. Not bad since the motor weighs 0.9kg

DSC_0838.JPG (131.75 KiB) Viewed 1223 times

Left Side

DSC_0839.JPG (124.93 KiB) Viewed 1229 times

Right Side

DSC_0840.JPG (132.49 KiB) Viewed 1223 times

Really happy with how this one has come together.
The single swing arm is defintely stiff enough, and the motor has a skirt bearing so no stress issues.
The deadstops are adjust via two M4 screws. So it is ultra easy to adjust how far the motor sits from the tyre, and limit how much it will rotate in to the tyre as well. So it will be really easy to tweak the amount of engagement to minimise losses. I will need some method of locking them in place once adjustment is done. i.e. lock nut. And I will probably add some rubber to soften the blow when it hits the deadstops.

The dead stop arrangement. Note the dodgy work with a dremel because I don't have a mill.

DSC_0846.JPG (102.58 KiB) Viewed 1229 times

I have also made the spring that takes the weight of the motor easily adjsutable too. So this will allow for more subtle tuning of the drive engagement. All you have to do is rotate that little collar that retains one end of the spring, then tighten the grub screw. (Note that little collar comes with the motor)

The motor counter balance spring.

DSC_0849.JPG (101.96 KiB) Viewed 1223 times

Here is a pic showing all the parts that make up the drive. On the left is the fasteners, in the centre the custom parts, and on the right the parts required from hobbyking. A total of 7 parts (4 custom, one modified), plus 10 fasteners (6 new, 4 came with the motor). Definitely happy with that. Can't see any way to reduce that while keeping the level of adjustment I have.

The Commuter Booster Drive in bits.

DSC_0863.JPG (150.38 KiB) Viewed 1223 times

The custom parts:
- The main clamp/pivot block. Made from Acetal with a drill press, hole saw, various drill bits, M4,M5 & M6 taps, and a dremel. Size and shape is still totally un-optimised. I just used the blocks that I had laying around.
- The swing arm. Made from aluminium with an angle grinder to cut to size, drill bits, & M4 tap. (Note the centre clearance hole is too big as this was originally a right hand swing arm that took a bearing.)
- A spring. Repurposed from an extension spring using a dremel to cut to size, then lots of swearing to bend it into the shape I needed.

The Custom Parts

DSC_0865.JPG (138.27 KiB) Viewed 1229 times

The fasteners:
2x M6 screws to clamp it to the seat tube
2x M5 screws for the dead stop adjustment
1x M4 screw which is what the adjustment screws act on
1x M4 grub screw to lock the shaft to the swing arm (should probably have two)

The fasteners

DSC_0866.JPG (130.11 KiB) Viewed 1229 times

The OEM parts
1x $60 Turnigy 6374-200
1x $8 Turnigy AerodriveXp 63-74 SK Series Shaft

The OEM parts

DSC_0867.JPG (149.84 KiB) Viewed 1229 times

Now I just need a chain ring, and brakes on the Avanti then I can mount it up and give it a run.

Mounted on the bike

IMG_1432.JPG (155.6 KiB) Viewed 1090 times

- Adrian

[recumpence]

The parts look well made!

My only concern is the use of the Acetal (Delrin) as the frame mount clamp. It may turn easily. Also, Acetal breaks easily when under tension.

I may be wrong, however. It is just a thought.

Matt

[adrian_sm]

Thanks Matt. I really didn't put much thought into materials just grabbing what was laying around, so your comments are welcome.

For the final model I'll probably just make them in ally, and use some Nyliner bushes. That should do the trick.

Cheers, Adrian

[recumpence]

Cool.

One simple method of reducing the tendency of Delrin from twisting when used as a clamp like that is to use a thin piece of rubber (inner tube?) as a buffer inside of the clamp. That rubber will protect the paint as well as reduce the tendency of the drive to twist on the frame because it has a "Tacky" surface.

Matt

[jbond]

This is turning out really sweet. One suggestion. Now you've gone to a single sided swing arm, you migh think about putting the pivot on the other side of the seat tube. Having a longer pivot arm will make the arc that the motor moves in shallower. This might make the end stops less critical as the motor con't cut into the tyre at such a sharp angle.

[Kepler]

Looking good Adrian. I have built a few prototypes out of Acetal. Works really well as long as you give it plenty of meat around any stress points. I think you might have some issues with the threaded stop screws breaking out but time will tell. The Acetal is great for proof of concept.

I presume you are putting a second swing arm on the other side. Looking forward to seeing the road test

[adrian_sm]

Looking good Adrian. I have built a few prototypes out of Acetal. Works really well as long as you give it plenty of meat around any stress points. I think you might have some issues with the threaded stop screws breaking out but time will tell. The Acetal is great for proof of concept.

Yep, the threaded screw stops are definitely the weakest link in this design, especially being in plastic. The location doesn't help much either. If I want it to last I will need to add some rubber to soften the blow.

It was all a bit of a by-product of only limited time to make this one up, so I just re-used the swing arm I had which stopped me from putting a bigger 'lever arm' on the dead stop.

I presume you are putting a second swing arm on the other side. Looking forward to seeing the road test

I was originally planning on this too. But once I put this together it seems pretty damn strong. It has a pretty good light press/sliding fit between the pivot arm and swing arm, so no slop, then the grub screw gets done up to take the fairly minor spring torque load. It also makes the assembly really straight forward, and as a real bonus provided a super simple method to adjust the spring tension. Previously I was going to put another screw in to adjust that force, as you wouldn't have had access to the end of the spring like you do on this single swing arm design.

I really like the single swing arm design. I'll be keeping it unless things go drastically wrong in testing.

This is turning out really sweet. One suggestion. Now you've gone to a single sided swing arm, you migh think about putting the pivot on the other side of the seat tube.

What advantage do you see with that? Geometry is pretty similar either way. Clearances to pedals is the same. Motor can is fixed in position with radial screws so no issue there. With the spring I have used it is currently winding it up (i.e. reducing it's size in diameter when under load) which is a better then unwinding in my experience. Apart from that it would be easy to flip over. I just don't see the advantage at the moment.

Having a longer pivot arm will make the arc that the motor moves in shallower. This might make the end stops less critical as the motor con't cut into the tyre at such a sharp angle.

You might be right, I haven't put a lot of effort in tuning this. Kepler has. I am running on the lower limits of the arm lengths that he found were good. The first prototype I made had the same pivot arm length and the engagement was fine once tuned, and dead-stops were not that critical. In fact thinking about John/Kepler's eBoost design I am not sure if it has this type of adjustment, and people seem to be able to set them up just fine.

The reason I have been keeping it short are to make the drive suit a larger range of bikes. The longer the pivot arm, the bigger a space you will need in this region of your bike. Obviously not a problem on my Avanti frame in the picture above, but try and put it on a modern road bike, and the is not much room in there, especially with this big motor. I think it will only just fit my road bike with this motor, as I have a large frame size. For smaller frames with less space, you would need to stick with the smaller motor size.

- Adrian

[Hillhater]

really neat, clean design Adrian.
..but i think you really need to thoroughly test it with that single arm mount as it is now relying entirely on the internal motor bearing locations to resist the "bending" forces and hold it together,.... and i have had instances of those not holding up too well !
I hope it does hold up, but the second support arm sure gives it much more support for only a few grams more weight.

[Kepler]

My thoughts also Hillhater. Still worth testing as it is though.

[adrian_sm]

This is turning out really sweet. One suggestion. Now you've gone to a single sided swing arm, you migh think about putting the pivot on the other side of the seat tube.

What advantage do you see with that? Geometry is pretty similar either way. Clearances to pedals is the same. Motor can is fixed in position with radial screws so no issue there. With the spring I have used it is currently winding it up (i.e. reducing it's size in diameter when under load) which is a better then unwinding in my experience. Apart from that it would be easy to flip over. I just don't see the advantage at the moment.

Hang on. You meant in front of the seat-tube didn't you. Ah. I thought you meant left versus right.

If I get good results with this geometry I'll keep it, as it is more compact. Initial test results look promising. We will see.

[adrian_sm]

really neat, clean design Adrian.
..but i think you really need to thoroughly test it with that single arm mount as it is now relying entirely on the internal motor bearing locations to resist the "bending" forces and hold it together,.... and i have had instances of those not holding up too well !
I hope it does hold up, but the second support arm sure gives it much more support for only a few grams more weight.

Hillhater. What motor/support scenario was that. Which bearing died? The skirt bearing? Have you got pics, links, details. Would love to be able to learn from your pain rather than my own for a change.

- Adrian

[Hillhater]

Adrian, sorry no photos.
It was a while ago on one of my earlier designs with a similar 63mm turigy and the two arms , but that did not have them rigidly connected.
After a few test runs , i noticed the motor "working itself" apart.
Basicly, the motor bearings were "wriggling" in their mounts, with only the circlip on the shaft preventing it from falling apart.

[adrian_sm]

Thanks for that HH. I know what to keep an eye on now.

Update:

Spent some time getting the Avanti test mule ready to run.
- put new front wheel on (thanks full-throttle)
- mounted Avid BB7 brakes to the front
- adjusted rear rim brakes to suit narrow rim
- mounted Commuter Booster Proto #2
- very crudely mounted controller and servo-tester throttle interface, with thumb throttle
- mounted battery in a bottle, just because I could

Still needs a chain ring, and chain.

Bit worried that the resolution of the thumb throttle won't be good enough. I can't seem to make fine adjsutments. I am not sure if this is beacuse of the stiction causing a bit of a jump in position, or something else. Road testing will sort it out.

I also need to work out the permenent solutions for mounting controller and throttle interface. For this large prototype there might be enough space to mount it to the pivot block directly which would be neat. For the smaller motor version, it will need to be mounted seperately.

Here are some pics, before I fixed the brakes.

IMG_1455.JPG (143.12 KiB) Viewed 1036 times

IMG_1457.JPG (137.7 KiB) Viewed 1036 times

IMG_1459.JPG (126.4 KiB) Viewed 1030 times

- Adrian

[adrian_sm]

Test ride went well. New adjustment methods work really well. Was able to tweak the geometry and engagement quickly, and easily.
New thumb throttle makes a big difference in useability. It softened the engagement purely due to the time required to move the throttle. Also much easier to tweak the throttle to the amount of assist you want.

Managed to pull 80-100 amps according to the watt-meter. So 1600-2000 watts .
After burning off about 3Ah from the battery, playing around on local streets the motor was warm to touch, controller had no obvious heat.
I think long term I would want to be able to restrict the max current, to make sure I don't overheat the motor lugging up a hill, in the lower inefficient range of motor speeds. Top speed seemed low, but this may due to me being used to my 80v hub motor ebike, that I usually cruise at 50kph on.

On a negative note, this is not a stealth bike, it has the typically RC scream. Not enough to scare pedestrians, but enough to turn heads. Guess I am spoiled with my direct drive hub motor, which has no really sound at all when cruising along. Also going over bumps causes the motor to bounce and touch the tyre, not a big deal but thought it worth of mentioning.

Really need to go for a decent length ride, and give it a proper shake down.

Getting pretty excited, this is definetely maturing into a really useable system. Meeting all the objectives I had. Light weight, cheap, good performance, minimal parts, easy to make, simple to use.

I also jsut weighed the bikes.

The road bike with the first prototype weighed in at 11.8kg (26lbs), and my heavier hardtail MTB, with prototype 2, with the heavier motor, weighed in at 15.8kg, or 35lbs. It is such a different experience riding a light efficient bike again, but still being able to get that 1000+w push.

- Adrian

[adrian_sm]

. Also going over bumps causes the motor to bounce and touch the tyre, not a big deal but thought it worth of mentioning.

Thinking about this a bit more, Kepler's eboost geometry will suffer from this issue much less than my design, since the pivot point for the motor is nearly directly above the motor. Where as mine in cantilevered out.

[adrian_sm]

I quickly modelled up the current prototype 2 design with Google Sketchup, and found out that with a plugin in I could create a 3D pdf file.
The model is pretty rough but representative, even with the weird sized pivot and clamp blocks. So if you are interested here is a 3D model you can open up, spin around, hide parts, and generally get a better look at the design.

- Adrian

[deVries]

I quickly modelled up the current prototype 2 design with Google Sketchup, and found out that with a plugin in I could create a 3D pdf file.
The model is pretty rough but representative, even with the weird sized pivot and clamp blocks. So if you are interested here is a 3D model you can open up, spin around, hide parts, and generally get a better look at the design.

Pretty darn kool... Nice design on 3D. Look forward to your RW tests & performance...

[adrian_sm]

Pretty darn kool... Nice design on 3D. Look forward to your RW tests & performance...

Thanks. I just wish it would stop bloody raining here in Melbourne, so I can get some decent road time. I am heading interstate on Tuesday, and I won't be back for a couple of weeks. So if I don't get a ride in tomorrow, it won't be until next year.

This time of year is just crazy. Between work, kids, and the usual festive social engagements, I really struggle to get any time to play with this. It is so frustrating now that I have the whole thing up and running.

But from all the recent rides, it looks like it just works. I need to pretty it up and package it a bit better (hence the CAD models), but the mechanicals are all fully functional. There will be things to tweak as I get more road time. But the developement focus will now shift to sorting out the electronics side of things, such as throttles for a road bike, packaging the ESC and throttle interface better, maybe finally getting around to trying out Jonas's excellent work on making the watt-meter a throttle interface, or getting my hands on Kepler's if he ever releases it to reduce low speed amps, and mounting multiple batteries securely and elegently.

I really need to find a good solution to the road bike handle bar throttle problems, because the drive really suits this type of bike. I have some linear sliding pots on their way, that I am keen to try. But this throttle is likely to be just as complicated as the whole of the friction drive. At least it gives me something to sketch out and design while on holiday.

- Adrian