Kepler Friction drive comes of age.

[urga]

Great.Thanks for the pics.
I think I understand how you can use the existing Ardunio setup to acheive regen. Think I would miss the 2 power options though.
I'm thinking that since I've been unable to program a single Arduino with the code you sent me,nor been able to get a HM-10 bluetooth module to talk to either my Focbox or Flipsky mini,I might be better being spoon fed with your products ie purchasing a pre programmed arduino with the two power options and regen.
Does the motor still rattle over bumps with this setup? Scottish roads aren't getting any better,and my previous setup drove me crazy, hence why I went with the constant contact setup. Is the cadence setup fairly simple to get it talking to the Vesc ? I also have a Trek Domane which is similar to your Madone in that I could use https://www.trekbikes.com/gb/en_GB/equipment/cycling-accessories/bike-computers-gps/bike-computer-sensors-accessories/bontrager-duotrap-digital-speed/cadence-sensor/p/08298/
that maye could talk to the Vesc via the bluetooh module that I can't get to work
They're a bit pricey but would mitigate any more wires on the bike
Just as an aside,I managed a 83 mile ride on the Crux, using a lot of assist(4620 ft of climbing) by using my homemade bottle battery(6s2p) and a Turnigy 8Ah battery in the saddle bag. I ran the saddle bag battery feed down the seat tube and into the bottle battery and connected them in parallel to increase the range. Cells were sitting at 3.4v at the end,so imagine the range I would have if I can factor in regen and twin batteries

 

[scottorious]

I just had a sad realization. The bike i ordered to make this work has a bb386evo bottom bracket. Anybody that watches this thread know of a conversion option? The bike is a diverge carbon comp. I got a decent deal and the bigger tires seemed like a good option. Also has bottle attachments on the bottom side if the downtube that look prime for a custom battery and electronics mounting. Or, whats the possibility that those 3 attachment points could handle the torque of the motor?

 

[Kepler]

Put together this bike to use as a low cost commuter. The bike is an Aldi special hard tail that for the bargain price of $400, came surprisingly well equipped. Included hydraulic discs and Shimano SLX group-set.

Installed my standard drive bracket but added a mechanical activation lever but this time used used a cheap front derailleur activation lever.

Installed a Vesc 4.2, Ardunio interface, custom button throttle, and a 6S 6 Ahr battery hidden away in a water bottle.

Looking forward to using this a my new commuter once we get out of lockdown.

 

[MJS]

Hi Kepler

I came across your videos on YouTube which led me here.

Like you, I was looking for the cheapest, lightest, simplest ideas to give me a little help up a hill or into wind when needed. Your designs and ideas are the best I could find, thanks for sharing all your work.

I particularly liked this last build of yours, a cheap Aldi bike conversion, because another problem with Ebikes is they have become desirable to steal..Having something that doesn't look expensive for a daily runaround is a big advantage.

What motor are you using on this latest bike? And what is that tape wrapped around the centre of the can?

Best regards

Mark

 

[gianfr80]

hi i made a nice setup that work awsome on a straight road, can reach up to 50km/hrs, also on hills no more than 15% can reach up to 33km/hr with pedal help. the problem its on high angle hills, there is no way for it to work, i have to change shifts on the bike an end up peddaling all the way up with almost no help of the motor an really slow.

have a:
Flipsky FSESC Mini V4 rated 50A
TURNIGY AERODRIVE SK3 - 6374-192KV MOTOR OUTRUNNER rated 80A
An a hailong 48v 12.8ah with BMS rated as 30A continuos discharge. made of lg mh1 3200mah, 13s4p

an running on FOC mode
set 80a on motor
an 30a on battery

any idea what could a change to improve the uphill performance?
(its ugly as hell , but it works )
[youtube]https://www.youtube.com/watch?v=381T4H5lJiI[/youtube]

 

[Kepler]

I thought I would post an update on my latest friction drive design. Although the previous design made from 304 stainless has been very successful, there were elements to the design that I wanted to improve on.

Firstly, I wanted to see if I could come up with a 3D printed design that would be robust enough to handle the rigors of hard every day riding.

I also didn't want to be reliant on a commercial laser cutting company to make the required parts as I tend to get an order or 2 a week and really liked the idea of just running off the parts when an order was placed.

The core of this design is the pivoting swingarm that the motor is attached to. The previous design relied on metal to metal pivot point which when clean and lubricated, worked well. However it was important to keep the pivot clean or the pivot action became sticky and less effective.

It also had a tendency to rattle when not engaged. This didn't effect its performance but it was a common complaint. I had built a fully ball raced porotype that I still use on my road bike now. This design always pivots perfectly and due to its smooth action, didn't need a gravity spring. Also there was no rattle from the drive when not engaged.

However building a ball raced version of the previous design added complication and would double the cost of the bracket so I stayed away from this path as I wanted to stick with the $100 USD price point.

So enter the new 3D printed prototype and its new design features.

The positives:
Firstly, the swingarm has dual 8mm sealed bearings fitted providing a super smooth rattle free and maintenance free pivot.

No gravity spring is needed to assist with engagement due to swingarm moving so freely. Added a cam style endstop adjustment making setup much easier.

Added a built-in cable guide for installations that want to be able to lock the drive in the fully engaged position from the handlebars.

Lowered the parts count to 4 x 3D printed parts, 2 bearings, 2 washers, and 7 off the shelf fasteners.
Halved the weight of the bracket. (from 200g to 100g)

Still working though a few things. So far I have been making the bracket from PLA. I need to do some warm weather testing to see how the PLA stands up. I may need to switch to PETG or some sort of composite filament in the future.

I am fairly new to the world of 3D printing and so far my prints in PETG have not been great. As anyone who 3D prints would know, its a hobby in its own right.

And now the negatives.
The attachment on the bottom bracket now moves the Q factor out by 3mm. The 304 stainless design moved it by 2.5mm. I will need to see if this will be an issue.

There is more bulk to main support bracket which may effect clearances on some bikes. So far on the bikes I have tested, it has been fine but it may be problematic on carbon frames with bulkier bottom bracket areas.

Anyway, I am really pleased with the design in its current state and think it has a heap of potential. Below are some renders and a photo of an assembled drive.

 

[SlowCo]

That's a very nice "new" design. It seems the stainless steel version was at the end of its development where this 3D printed design opens up new possibilities for further evolution using this technique :thumb:

 

[LewTwo]

Still working though a few things. So far I have been making the bracket from PLA. I need to do some warm weather testing to see how the PLA stands up. I may need to switch to PETG or some sort of composite filament in the future. I am fairly new to the world of 3D printing and so far my prints in PETG have not been great. As anyone who 3D prints would know, it's a hobby in its own right.

PLA is not a good choice for anything that will be used outside. It was always "intended" to be a biodegradable material. Regardless of what you do to it (coatings, annealing, etc) it will eventually start to break down. It is however fairly strong for such an inexpensive material.

ABS was the original 3D printed filament and still a good choice for parts exposed to the weather. ASA is the modern replacement for ABS. I believe that Nylon is still "King" among the "engineering grade" filaments. PEEK is far stronger but hugely expensive.

Most "engineering grade" filaments need both an enclosed printer and a high temperature hot end (160 degrees Celsius or better). Some specialty composite materials (especially carbon fiber filled filaments) need a steel nozzle rather than a brass nozzle.

If you have not already found it, then see CNC Kitchen on youtube for real world practical information on various filament materials.
https://www.youtube.com/channel/UCiczXOhGpvoQGhOL16EZiTg

Your design looks fairly compact. What is the build volume (x,y,z) needed for the largest part?

 

[ahagge]

Hi Kepler,

That looks like a design with lots of possibilities, though I agree with LewTwo that aside from prototypes, you'll probably need to work with a material other than PLA. I have ABS, but have never used it due to what I perceive as the complexities of the printing process. I've also printed in PET-G, which I think would be a better choice than PLA, assuming you can get your printer dialed in.

A question or two about the new design:

• Do you really think that a gravity spring won't be necessary with this design?
• Is this design limited to using 50mm outrunners? Seems like the larger, heavier 63mm ones might have issues. What's your preferred motor these days?
• Any issues with squishing the bottom bracket mounting ring? Seems like the amount of torque necessary to keep it from rotating might deform/crack the material.
• Any chance you'll consider selling just the STL files for those of us who have 3D printers?

I'll be interested to hear follow-up posts about how well this holds up after hundreds (& thousands) of km. Good luck!

 

[chilau2000]

Looks awesome.
I have a PETG 3D printed bracket on my sons friction drive bike, the motor mount is in carbon fibre.
The PETG part is holding up really well, my experience of PLA is that it will be to brittle for this application.
I do notice that my sons mount does have some flex in the mount even though the motor mount plate is 4mm carbon, wonder how PETG would cope with the flexing.
My son will need to move to a bigger bike next year and a 3D printed holder with bearings sounds great for his next bike.

My sons friction drive
https://endless-sphere.com/forums/viewtopic.php?f=28&t=96581&p=1656493#p1656493

 

[Cowardlyduck]

PETG is the way to go. I find that a spool of PETG needs to be used pretty quickly after opening or it will start to absorb moisture and then it's all over unless you want to print spaghetti.

I would be interested in any improvements on the design that might work well with knobbly tires. I have a few projects in mind that could use this kind of setup if it works ok with more off-road type tires.

Cheers

 

[Andre]

Hi all!

I'm back with another version of the friction drive...




Video here:
https://youtu.be/8FUkQn2LDEk

I kinda went all out this time

Features:
-integrated into a handlebar bag
-compact drive unit attaches to a regular Nitto rack (convert the bike to electric and back in less than 30sec)
-automatic deployment/retract (with an elegant lockout feature when retracted)
-easy to attach and compact handlebar switch with plug-in wiring
-drive unit powers the front light, too
-bluetooth connectivity
-lots of room for different battery options + plenty of space left for transporting stuff

 

[Ecyclist]

This is a fantastic finished product. You should start you own thread.

 

[Lurkin]

Golly people still use this forum? all the others I used to frequent have long since died with the advent of social media.
Hope you are well John, this has certainly come a long way from mini Arduino discussions!
MarkForged3d were the apple of 3d printing last time I looked into it for automotive part purposes, capable of printing metal now and with a combination of printing and sintering to seal/finish/temper. Aluminium was definitely out (to hard for them to sinter correctly at the time, this may have changed by now).
Their demo units are located at Monash university (amongst other crazy machines - full engine block casting printings out and more) and they are using them to print titanium bike parts, cant remember the manufacturer they are for. Whilst titanium might be too expensive certainly printing steel or an alloy may be an option.
Could these be cut with a small CNC? Could be an affordable in house option without the need for redesign.
Look forward to see how you go with the PLA - I struggled with strength and longevity for my purposes. NB can also print carbon now - perhaps this is an option.
Anyway sometimes inspiration can come from seeing other new ideas/other ways of doing things - worth a look.
Could a nylon or plastic washer/thin spacer be used to dull the rattling now?
Guess its a solution to a problem you will no longer have shifting to 3d printing.

 

[LewTwo]

Golly people still use this forum? all the others I used to frequent have long since died with the advent of social media.

social media ... what is that

 

[calab]

Social media is the downfall of society with some outspoken crazy people on it.

 

[tomjasz]

Social media is the downfall of society with some outspoken crazy people on it.

No crazies here? Too funny.

 

[massiveb]

Such a beautiful build

Hi all!

I'm back with another version of the friction drive...

Screen Shot 2022-01-03 at 6.29.58 PM.png
Screen Shot 2022-01-04 at 1.56.21 PM.png

Video here:
https://youtu.be/8FUkQn2LDEk

I kinda went all out this time

Features:
-integrated into a handlebar bag
-compact drive unit attaches to a regular Nitto rack (convert the bike to electric and back in less than 30sec)
-automatic deployment/retract (with an elegant lockout feature when retracted)
-easy to attach and compact handlebar switch with plug-in wiring
-drive unit powers the front light, too
-bluetooth connectivity
-lots of room for different battery options + plenty of space left for transporting stuff

 

[falcongsr]

It's been almost 6 years since I bought the Kepler unit for my commuter bike. I enjoyed it for about 9 months before I moved states, changed jobs, went through COVID, and changed jobs once again. The new job is close enough to finally resume bicycle commuting. I've just put the friction drive back on my bike, but I haven't been able to get the motor to engage the tire reliably. I decided to lock it in place to always make good contact with the tire. For commuting I think this is going to work well. I also turned on regen as described on the previous page and I actually like it so far. I might change the Arduino code so I can have off/low/high assist again instead of brake/off/high assist.

Does anyone have the Arduino/Visuino workspace files that John used to set up the throttle button?

I hope John's OK. I'm surprised he quit coming here as he was active and supportive of so many bike builds for so long.

 

[falcongsr]

Well it's been a great summer with the friction drive on my road bike. I am just passing 600mi/1000km biking to work this season. I've left the motor permanently pressing against the wheel, and I've come to love having regen on the downhill runs to control my speed. As I've gotten into better shape I'm using the motor so little that I hardly charge the battery, I just keep it around 70% charged and end up either charging +10Wh (regen downhill and not using assist uphill) or drawing 10-20Wh if I want to get there quicker or have a headwind. I was using up to 60Wh in the beginning of the season. I love having extra power there at the press of a button to maximize safety around cars or get through an intersection before the light changes.

I've been using a 6S 8Ah lipo (177Wh) battery, but I just ordered a 6S 2.5Ah 18650 pack (60Wh) to save weight. (1110g vs 300g)

I never did change the Arduino code and just got used to it.

With winter coming I won't be able to ride with this setup so it will get put away once the temperature goes below freezing.

 

[Pranav.ptl]

Well it's been a great summer with the friction drive on my road bike. I am just passing 600mi/1000km biking to work this season. I've left the motor permanently pressing against the wheel, and I've come to love having regen on the downhill runs to control my speed. As I've gotten into better shape I'm using the motor so little that I hardly charge the battery, I just keep it around 70% charged and end up either charging +10Wh (regen downhill and not using assist uphill) or drawing 10-20Wh if I want to get there quicker or have a headwind. I was using up to 60Wh in the beginning of the season. I love having extra power there at the press of a button to maximize safety around cars or get through an intersection before the light changes.

I've been using a 6S 8Ah lipo (177Wh) battery, but I just ordered a 6S 2.5Ah 18650 pack (60Wh) to save weight. (1110g vs 300g)

I never did change the Arduino code and just got used to it.

With winter coming I won't be able to ride with this setup so it will get put away once the temperature goes below freezing.

hi, can u share where u bought your 18650 battery and how do u mount it? bag under seat or cut up water bottle?

 

[Pranav.ptl]

Well it's been a great summer with the friction drive on my road bike. I am just passing 600mi/1000km biking to work this season. I've left the motor permanently pressing against the wheel, and I've come to love having regen on the downhill runs to control my speed. As I've gotten into better shape I'm using the motor so little that I hardly charge the battery, I just keep it around 70% charged and end up either charging +10Wh (regen downhill and not using assist uphill) or drawing 10-20Wh if I want to get there quicker or have a headwind. I was using up to 60Wh in the beginning of the season. I love having extra power there at the press of a button to maximize safety around cars or get through an intersection before the light changes.

I've been using a 6S 8Ah lipo (177Wh) battery, but I just ordered a 6S 2.5Ah 18650 pack (60Wh) to save weight. (1110g vs 300g)

I never did change the Arduino code and just got used to it.

With winter coming I won't be able to ride with this setup so it will get put away once the temperature goes below freezing.

i have the PAS visuino file.. don know how helpful they will be.. pm me if you want them

 

[falcongsr]

hi, can u share where u bought your 18650 battery and how do u mount it? bag under seat or cut up water bottle?

Hi Pranav, my original battery was a Turnigy High Capacity 8000mAh 6S 12C w/XT90 battery from HobbyKing. This is a LiPo pouch battery so it is not as safe as 18650 cells. I mount it under the seat with the large seat bag that Kepler sold with the friction drive.

The 18650 battery is an Auline VTC5A 2600mAh 6S 22.2v 35A XT60. One thing that was really noticeable with this battery is the IR drop under load. The higher IR drop means the charge current is lower than a lipo. The battery will get warm when charging anywhere near 1C. It worked fine and was lightweight, but for regen I had to limit the charge current to meet the specifications of the Sony VTC5A cells. Regen is used so infrequently that it didn't really heat up the battery, but still I limited it to 4A (1.5C). I prefer the Lipo which can be charged safely at 1C without problem and I don't have to think about the battery level because it has so much extra capacity.

Auline makes a 21700 4000mAh 6S battery now. I wish it was available when I was looking for batteries.

As far as the Visuino file, I used an oscilloscope to measure how Kepler's push-button design generates the PWM signal to the VESC, and I definitely want to review it and change it. I'll send you a PM.

Meanwhile it is the depths of winter here and I finally retired my friction drive setup in December. The motor worked fine, but when I removed it from the bike, it was very difficult to turn by hand and it felt like some of the bearings were jamming. Maybe I'll take it apart some time. I got about 1250miles/2000km out of the motor. I think having it in contact with the wheel 100% of the time for regen and power is what accelerated the wear. Not bad for the price.

Due to freezing temperatures I have to replace my slick tires with winter tires. I decided to get a Bafang G310 hub motor from Grin Tech. I selected the Fast winding so I could use the Kepler throttle, VESC, and 6S battery. I made a cable adapter to go from the VESC to the Z910 motor connector so I only needed to replace the wheel/motor. So far it is working well for my commute. I had to use an old version of VESC Tool 0.95 to connect to the VESC and recalibrate the motor settings for the Bafang.

edit - removed comment about VESC issue - i was mis-remembering how it worked.