Commuter Booster - <1kg Friction Drive

Update: Dual-personality

I think my drive has a dual personality, which is causing me and others confusion.
It started with personality 1, and is now fighting with personality 2.

Personality 1) The Original DIY
- light weight friction drive <1kg
- good performance >1kw
- cheap parts, see how cheap works
- able to be built with minimal tools, minimal hurdle for people to DIY
- simple
- market, ES type enthusiast

Personality 2) Commercial product
- full friction drive kit
- light weight friction drive, want the bike to still be a bike
- "legal" performance, able to be limited to local legal limits
- competitive total package cost
- initial small batch production to test market
- robust and idiot proof, self protecting
- market, anyone that doesn't think it is cheating

The problem is that these are actually quite different products, and will require different solutions. But I am trying to walk some middle ground ... and obviously failing.

For Personality 1). the mechanicals work, make some yourself and give it a go. But you will need to sort out that electrical stuff for yourself at the moment. There are a few options that should work, not elegant but work, depending on budget vs skill.

For Personality 2)........... well let's just wait and see once I make a few, get more road time, and get some feedback.

Sorry for the confusion people. Just getting ahead of myself.

- Adrian

To be fair, the 160A I mentioned should really only be neccessary for an axle-driven E-bike with an appropriate gear-reduction. There are web-retailers that sell components for noticably less than a seller that must maintain the costs of a physical building. That being said, the "suggested retail" price of the HV-160A is $320 on their home-page, for Motor-shell FDs such as Kep's and Adrians, the HV-120A is $270, and for drives like Todds, Mine, Zibby, and Rearengine (where the roller is only 1-1/4" dia) the HV-80A is $190.

Heres a web-retailer that has over 3000 sales, a 99.8% positive feedback, and regularly sells the HV-160A for around $250.
http://cgi.ebay.com/Castle-Creation...Control_Parts_Accessories&hash=item4155aa7c11

The same guy is selling the HV-80A for $130, and the HV-120A for around $210.

If you only need 5S-8S (18V-33V) plus the Castle Creations electronics and chip program for filtering and sensing the back-EMF (perhaps thats the only part of the CC "secret sauce" that we need?),the "non-HV" Phoenix-Ice 100A can be found for around $120. Plus...this model has a built-in BEC, so you dont have to buy a separate UBEC. This is the ESC that Zibby used, and he didn't report any sync issues. http://endless-sphere.com/forums/viewtopic.php?f=28&t=14403#p214997

http://toys.shop.ebay.com/i.html?_nkw=ice+100&_sacat=34056&_dmpt=Radio_Control_Parts_Accessories&_odkw=&_osacat=34056&_trksid=p3286.c0.m270.l1313

When the weather clears I will be trying out the $80 Hobbywing ESC I just got. If it doesn't work out, I may get the CC 8S/33V unit next.

I wonder how Hidden Power have dealt with the controller issues.

Thanks for the heads up on the excellent prices out there. I was under the false impression the CC set the retail price.

Spinner, your drive has the advantage of a smaller roller so 8S is certainly a good option. The CC controller will suit this very well.

adrian_sm said:

Wow. Long time lurker indeed. Glad you dropped by my little thread.

Click to expand...

Great, thanks. So the startup torque provide enough movement about the povit point to provide initial tire contact. From there the friction would allow the motor to climb until being limited by the deadstop. Sounds perfectly reasonable... as long as I'm understanding this right...

Good idea that.

... and written so much more concisely than my babbling.

Sounds like you understand it just fine. Key to my version, is the counter balance spring, making it a easily adjsutbale, and fairly light load required to lift it during motor start-up.

So when can we expect to see a creation from you gfulton?

I have created an index of sorts in one of the reserved posts at the start of the thread.
Hopefully this will be an easier way to find things in the huge thread now.

Here's the link.

I have also updated the Commuter Booster link in my signature to point to the index post.

- Adrian

Kepler said:

. A custom built sensorless 6FET setup for 6S and 50A makes the most sense to me especially considering RPM on a 200kv motor on 6S is around 4800. Most likely source would be from Lyen. Maybe he is already working on it. (I have no idea if he is) But would this guarantee no more sync issues?

Considering the whole point of going down the path of a friction drive is to be small, light weight, and cheap, I dont think we are too far off the mark with what we are doing.

Click to expand...

Lyen's recommendation to me was to go with a modified 12 FET controller ($129 ++) to be reliable.

Still wondering why no one has tried the CC Mamba Max Pro.. a 6s ESC with auto sensor /sensorless running.
http://www.castlecreations.com/products/mamba_max_pro.html
They can be bought for $110 and should have all the castle magic plus sensor capability ?
http://cgi.ebay.com.au/Castle-Creations-1-10-Mamba-Max-Pro-ESC-Smart-Sense-/380305328971?pt=Radio_Control_Parts_Accessories&hash=item588bf6cb4b

PS: Yes i do have a cheaper 150A sensored ESC ready to try, but no halls yet !!

Can someone compare and contrast the the reliability aspect of using an rc esc like the ice hv 160 over a 6 FET ebike controller of around the same size? if my friction drive fails I would like to use a hub motor, and if the CAstle does not work well with a hub motor I'll most likely go with an ebike controller...

You can't buy an ebike type controller off the shelf that will drive these motors at these voltages.

So I would just recommend going for an RC ESC that are designed for these motors.

The issue is that these RC-ESCs are not typically designed for low speed, high load scenarios that we may expose them to. So you have to be careful.

Update: Turnigy Servo Speed Regulator

In the quest for cheap off the shelf solutions to some of the perculiarities of friction drives I found this.

The $8 Turnigy 3 Channel Servo Speed/Direction Regulator


What it does is take over the throttle signal, and sets an adjustable maximum rate of change for both accelerating, and decelerating the motor.
So when accelerating, it makes sure the motor doesn't spin up so fast that it slams into the tyre.
And when letting go of the throttle it can slow things down enough to avoid the motor getting thrown into the deadstop.

This means the rider doesn't need to be gentle with the throttle, as this little guy will stop anything stupid happening to the drive. Come to think of it, for a low power system this might make an on/off button throttle possible, as it would soften the whole transition from 0 to 100% throttle. Hmmm will have to try that.

All in all very happy with this little find.

- Adrian

adrian_sm said:

... and written so much more concisely than my babbling.

Sounds like you understand it just fine. Key to my version, is the counter balance spring, making it a easily adjsutbale, and fairly light load required to lift it during motor start-up.

So when can we expect to see a creation from you gfulton?

Click to expand...

Well, I've got the frame from a larger, old MTB... I'm about 6'1" 205 so I need something a bit larger. Thankfully the space provides for a motor without any hassle... only problem is rust and wear from the elements. I need to refinish the frame, replace the chain, check bearings etc.

All the electrical stuff is pretty much done, but I need someone with a metal shop (garage or otherwise) to fab some parts for me. I'd say by this spring I'll have something on the go. I may not document the build, since there are so many friction drives being done right now... but I'll definitely post some pics when I'm finished.

Thanks to all for your help and excellent ideas.

Adrian, thanks for the link on the 'throttle control'.

adrian_sm said:

Update: Turnigy Servo Speed Regulator

In the quest for cheap off the shelf solutions to some of the perculiarities of friction drives I found this.

The $8 Turnigy 3 Channel Servo Speed/Direction Regulator

Click to expand...

Adrian,
you may want to review Olaf's efforts with this unit from last year..
http://endless-sphere.com/forums/viewtopic.php?f=28&t=21552&hilit=ramp+control

Thanks for that link, I must have missed that first time around. But it sounds like Olaf had different plans for the same device. He was trying to use it to to protect his electronics, where as I am using it to protect my mechanical parts.

Just to clarify, this ramp controll will not allow you to get away with a smaller cheaper ESC (see the discussion in the above thread), but it does appear to help smooth the engagement, and disengagement of the drive with the tyre in my application.

On a related note the ESC soft-start option is not an alternative. It does something very similar but in a very unuseable way for our needs, so should always be set to OFF.

It does provide a slow initial ramp up of the throttle, but only does it when first accelerating, then it will have instant response after that, until the throttle sits at zero for a period of time (~3secs for the Turnigy K-Force 100A according to the manual). You typically only have two available ramp rates to select from too, when you really need to have some fine control to make it work. For example the Turnigy 85A has options of 1.5sec, 1.0sec or off. The Turnigy K-Force has 8sec, 3sec, and off. For their Very soft, Soft and Off settings.

- Adrian

P.S. Kepler, I noticed in another thread that you recommended soft or very soft for this setting, I would be interested to know why, as it really didn't work for me.

Kepler's Interface

Dropped in on Kepler this morning for a bit of a mutual show'n'tell. I got to try out his little interface hooked up on his beauiful flat bar road bike, with an eboost on the back of course. And he got a try the 6374 Commuter Booster on my cheap hardtail MTB.

I was really impressed with Kepler's new interface. It is much more than just a throttle interface though, it really is or could be an all inclusive electronics solution. It provides all the throttle logic, protects the motor, ESC, and batteries, stops the "Is this thing on?" events, makes the system legal (or not), and from what I hear :wink: will be very reasonablly priced.

It has really made me have second thoughts about trying to develop anything that performs a similar function for my drive, as he and his colleagues have all ready put in all the hard yards. I heard my self describing how I could hack away using the watt meter as a hardward platform to do a similar thing. But at the end of the day it might be worth it just for the challenge, but it would never be worth the time and effort long term when Keplers little device is just around the corner.

Some from the saddle impressions of Kepler's interface.
1) Button Throttle
- it works, I liked it
- If you hold down the button, the drive engages and picks up the speed a bit to let you know its there
- momentary release and press again, and you get a litle surge of power and go quicker
- let go, and after a very brief delay ~0.5 seconds the drive disengages
- Too easy

2) The other features
- most of these you don't notice, they occur in the background
- which is great, they don't get in the way of enjoying the bike

3) The Display
- Yep, looks like a display is an option
- Kepler showed me the prototype debug display, and it has all the info you could ever want to get distracted by
- I am keen to see what the final polished version looks like

Anyway, as long as I don't abduct one of his kids, or make crop circles in his front lawn, I might get a chance at bolting one of these onto my ride for some serious road testing with the Commuter Booster in the not too distant future.

Thanks again John, for letting me drop by, shot the breeze, and dream of the future of light weight electric bikes.

Cheers,
Adrian

Good to catch up Adrian. I'm glad you liked the way the interface operated. A lot of work has gone into getting it to this stage and its nice to now see light at the end of the tunnel. I will get an early release to you as soon as I can.

It was really great to get a first hand look at your handy work and giving it a good test ride.

My impressions:
Firstly I really doubted that the single pivot arm concept would work. I was wrong .Plenty of rigidity and a much reduced parts count (compared to my design). I like it a lot.

End stop and tension spring arrangement. Again a nice design that makes drive adjustment a breeze.

All in all, great work. I am now re thinking some of my design and will certainly look at incorporating some of your ideas. into my drive.

OK, no mucking about here. Started work on a single arm version to bolt onto my drive. Thought I would try a different approach in relation to the return spring / damper spring. This design uses a spring steel wire to carry out damping duties rather then winding up a coiled spring like Adrian does. This spring wire goes through the stop pin which is just a roll pin and anchors to the bearing block. Disadvantage is that its not adjustable like Adrian's setup. Time will tell if this is a problem. I can swap out different diam spring steel rods to fine tune.

:lol:

Great stuff!! I am very eager to see how this works out, I had a similar idea with a leaf spring and some form of adjustment screws to increase/decrease tension, but I like this even better! Simple is usually best, and this looks very simple and if it gets some form of adjustment for fine-tuning, I would think it would be ideal.

Glad to have the two of you to get this stuff figured out for the rest of us.

The main advantage I see of Kepler's version is not having to drill a little hole on the shaft clamp on the end of the pivot axle that comes with the motor.
I tried similar things in the past, but found the advantages of the easy of adjustment of my design, far outweighed the effort to drill a single hole. I also found it more compact, easy to assembly etc. And torsion springs are cheap and repeatable to manufacture in volume.

But I love seeing others take another spin on the design.

Some people asked in the past whether a single arm design was a good idea, strong enough etc. I have convinced myself, and Kepler seems to be pretty keen on the idea now...

adrian_sm said:

... Some people asked in the past whether a single arm design was a good idea, strong enough etc. I have convinced myself, and Kepler seems to be pretty keen on the idea now...

Click to expand...

My gut feel is that you are approaching the limitations of the design putting 2kW through this, but ... you are both MEs and presumably have access to the software so do a FEA on it.

[Edit]On my custom frame, layman's FEA took the form of some blocks of 4x2 and a 14 pound sledgehammer

Cheers,
GT

I am making a trike using similar turnigy 50-65d 270kv motor
http://endless-sphere.com/forums/viewtopic.php?f=28&t=24572
.just wanted to know the performance of motor
1)maximum loaded current
2)maximum efficiency current
3)loaded rpm (no load will be approx 6750 @ 25v )
4)power output
5)any magnet loosing or bearing issues or heating issues

these are general specs ,if you have measured anything just let me know

thanks
pankaj

Hey there pankaj,

Let's see what I can tell you.

1)maximum loaded current
- I have seen 120Amp peaks when running at 6S, but I would not recommend this, and am now using 5s or ~20V

2)maximum efficiency current
- not sure what you mean. But I have no way of measure motor efficiency unfortunately.

3)loaded rpm (no load will be approx 6750 @ 25v )
- This really depends on the load. But the motor will take the usual parabolic power output, when plotted agaisnt RPM.
- So it shouldn't be too hard to plot out the peak power at half max RPM, then work out what sort of load your bike, hill etc will put on it, using a table like this.


4)power output
- Seen over 2600W, but again would not recommend this. The motor doens't have enough surface area to disipate the waste heat.

5)any magnet loosing or bearing issues or heating issues
- no loose magnets
- bearings are tiny, open it up and have a look. I haven'y killed them yet.
- heating issues. Yep. Burnt out the motor going up hill at 40kph, on 6s battery.
- Without limitting the power to the motor it will kill it self at these votlages.

Essentially this motor is too small on its own @ ~24V, and needs to be protected somehow.

So I would recommend either:
1) limitting power to the motor
- there are a few of us working on this, with different strategies at the moment
2) Using a lower voltage
- again to limit the power, and therefore waste heat
3) Going for a bigger motor
- that can take the abuse better.

Good Luck.

- Adrian