Kepler Friction drive comes of age.

[Kepler]

Most of the parts are now in for the new batch of drives. Just waiting on a bulk order of 304 Stainless steel fasteners.



And of course a new drive deserves a new bike. Went with a Cyclocross bike this time so i could fit wider tires to for a bit more comfort an grip during high speed commuting.



Should have the drives ready for shipping by the end of next week. Stay tuned.

 

[Slopes]

Need a bit more information.

Are you using a Hall throttle or a resistive throttle like a Magura?

...

Thanks for your suggestions, Kepler. Yes it is a hall-effect throttle.

I didn't have any luck following your wiring scheme... pic below (sorry for poor quality - the green wire from throttle going to the centre pin of where the pot used to be).

I firstly hooked up the servo tester - with its pot still attached - to the +5V and ground from the Vesc and that all worked fine (LEDs functioning). I then removed the pot and tacked on the rest of the wiring (inc. 'UART and PPM' in the BLDC Tool) following your advice. But when I power up the Vesc again everything's dead.

I could go back to the running the throttle through the ADC port but having bluetooth with the 'Vesc Monitor' app providing live feedback would be a great addition to the bike.

 

[Kepler]

It might be just the type of servo tester you are using. The wiring looks correct.

I would try eliminating the Vesc and see if you can drive a servo.

Disconnect the yellow signal wire going to the vesc but keep the black and red connected from the vesc to power servo tester.
Connect a servo to the servo output of the servo tester and see if it dives the servo when the throttle is twisted.



If it doesn't drive, you know the issue is with the servo tester.
Personally I would give the servo tester a miss and go to Arduino. Driving a servo with an analog signal is Arduino 101.

 

[Kepler]

If you plan to have a go at using an Arduino to drive the vesc with a hall throttle, here is a simple guide to get there.

1- Purchase an Arduino Nano. The Nano has a USB connection built in for programming making connection straight forward.
2- Download the free Arduino software from here https://www.arduino.cc/
3- Wire the Nano as shown below.

View attachment 1

4- Connect the Nano to your laptop through a USB conncetion
5- Start the Arduino software and set it to the correct port. (Go to Tools then Port)
6- Go to File - Examples - Servo - knob, and open the sketch. The code will now be displayed within the Arduino Sketchbook.
This code is for a potentiometer and as such the code expects a 0V to 5V sweep. We need a 1V to 4V sweep to suit a hall throttle and
as such the scaling in the code needs to be changed. instead of an input of 0 to 1023, we re scale this to 200 to 800.
7- Once you have made the change, compile the code and upload to the Arduino by clicking on the arrow pointing right in the the top
left hand corner pf the program.



This will now output a PPM signal to your Vesc that will need to be scaled in your Vesc Tool.

I hope that helps.

 

[Slopes]

Thanks for this Kepler - it's very useful. I've given up on the servo-tester hack and have an Arduino Nano board on order.

 

[Kepler]

Good to hear. You wont regret going down the Arduino path. I have really enjoyed learning some basic programming and the opportunities it opens up.

 

[ScooterMan101]

What places sell the Arduino Nano ?

one place I found , Mouser sells it for $ 28.86 , but I see a different place Gearbest , that list's one for $ 4.22

 

[Slopes]

What places sell the Arduino Nano ?

one place I found , Mouser sells it for $ 28.86 , but I see a different place Gearbest , that list's one for $ 4.22

I found ebay had a lot of them - for around the $4 mark.

 

[Kepler]

The standard drive bracket suits probably %70 of bike frames however making the main plate in two pieces opens up opportunities to mount the drive on practically any type of frame you can imagine. This is achieved by either making your own custom Upper Main Plate that say attaches to your kickstand mount or you can use the recently designed optional extension Lower Main Plate that extends the drive mounting point up to 130mm (distance between the centre of bottom bracket and the face of the tire)





It should be noted that the current design using the split Main Plate design with the Anti Rotation Plate has reduced the overall adjustment of the standard drive to a range of 65mm to 95mm (BB to tire face). Its unfortunately a compromise I needed to make but I think was worth the trade off.

 

[Aeron]

Has someone though of using the HK "Sk8" series motors ? Like this 6354 or 5045 for example. I was wondering whether the ribbed can would be a problem.

 

[Kepler]

Lots of requests for pedal assist. I can see merit in using pedal assist but I like the idea of still using the single button for 2 speeds and kill but only operational if pedaling. Anyway working on the code at the moment. thought it would be piece of cake lol. Wrong!!!

Anyway, got my litte test jig built and trying a few ideas now. give me a week or two

 

[Kepler]

I am currently in the process of putting together some YouTube Tutorials for various aspects of the drive.

The video below is just to demonstrate the button action and usability in real world riding conditions.

The next series of videos will be be on setup of the Vesc to use in conjunction with the drive. As there have been quite a few of the Arduino button throttle interfaces released, I thought i would make the next video tutorial on calibrating the button throttle within the Vesc tool. Stay tuned for that one.

Also plan to do some drive assembly and fitment videos in the near future. Let me know if you have have any specific requests.

[youtube]bkKbNOXiPgY[/youtube]

 

[Kepler]

Here is a Video tutorial on scaling the button controller in the Vesc. Hope it helps.

[youtube]TI6oENg0BiI[/youtube]

 

[Ebuilt]

Holding on to our seats for the video of arduino onterface calibration! This unit is so cool and tiny in person! Thanks John!

 

[Aeron]

Hello there,

I've received my drive mount and started its installation. I don't want to denigrate this excellent work, but a little thing bothers me and I wanted to ask here if I had missed something.

Here is the thing: by mounting between the BB and the BB shell on the crank side of the bike, the chainline is moved outwards from 2mm (and it's kind of a problem on my bike). Is there any particular reason for that ? I thought about mounting the thing on the opposite side, but then the gravity spring does not fit.

So, two questions: is there a particular reason to mount the drive on the crank side? If you think the drive could be mounted on the non-crank side, would it be possible to buy a "reverse" gravity spring (one with the locking pin on the other side)?

 

[Kepler]

the 2.5mm change in chain-line hasn't been a problem for me so far but if it doesn't work on your bike, by all means, mount the drive on the non drive side of BB.

I will place a reverse gravity spring in the post tomorrow for you. No charge of course

 

[Aeron]

Great, thanks a lot !
I have a 1x10 drivetrain, I've found these sometimes sensitive to chainline adjustments.

 

[Kepler]

Spring posted yesterday.

 

[Aeron]

I forgot to let you know I received it. This is great, thanks !

 

[Kepler]

Lots of requests for pedal assist. I can see merit in using pedal assist but I like the idea of still using the single button for 2 speeds and kill but only operational if pedaling. Anyway working on the code at the moment. thought it would be piece of cake lol. Wrong!!!

Anyway, got my litte test jig built and trying a few ideas now. give me a week or two

Been distracted with my Mountain Bike friction drive setup over the past month or 2 https://www.endless-sphere.com/forums/viewtopic.php?f=28&t=96581 but with that done and dusted, I have been back working on an Arduino based PAS system for the drive.

Took me a little while to figure out how I wanted to do this but finally came up with a solution that I liked. Being a minimalistic drive system, I wanted to make sure the PAS system was also minimalistic and as such came up with a solution that uses a single PAS sensor and a single magnet.

The PAS works in conjunction with a a single button throttle.
After the battery is plugged in, the drive stays switch off until the button is held down for half a second.
With the PAS now active, the sensor needs the magnet to pass twice within a a given time period for the drive to activate.
This time period is equal to a cadence of around 40 to 50 RPM.
Once activated by the PAS sensor, the drive will output 60% throttle.
If you stop peddling or drop below the cadence threshold, the drive will deactivate.
The button throttle also acts as a boost button. Holding the button down will ramp the throttle to 100%
Releasing the throttle button will return the drive to cadence control.
A short push of the button will deactivate the assist completely.

Below is a short video of PAS and button throttle operation. Still to road test the system but I think it is looking promising.

[youtube]jAMsSE3ZWLA[/youtube]

 

[cbr009]

Great work John, look forward to seeing the road test results.
Mick

 

[Kepler]

Road testing of the Arduino based PAS control is now complete. Settings that I came up with on the bench translated well to the road with the assist starting in a very predictable manor. Although theoretically two turns of the crank should start the drive, in real life it takes three or four turns to start the drive. This is because the cranks need to be up to speed before the assist requirement is registered. I quite like this as it means you can slow pedal when maneuvering say through a slow tight bend without the drive starting on you.

Stopping pedaling sees the drive hold for about a second before dropping off. This also has proven to be a nice length of time with no overrun experienced. A really quick stop might see an overrun issue but the Vesc can handle this ok.

And of course you can start the drive anytime by holding the boost button or stop the drive with a quick short press of the button. Again this work as designed.

60% pedal assist and 100% boost also works as hoped and as such see no reason to adjust this.

A slight change to hardware. The standard PAS sensor that is used on most ebikes was not as sensitive as I liked. So I used a OH090U. - H Uni-Polar Hall effect sensor instead. This sensor worked perfectly and allows the magnet to be a good 10mm away rom the sensor and still get a reliable trigger.

 

[juicy77]

Has anyone tested the kepler drive on dirt/gravel/mud? How does it fare? Is there excessive slippage? I am considering doing a tour of South America e-assisted and weighing my options either Kepler v Hub driven.

 

[Kepler]

I have built a custom friction drive for my full suspension mountain bike which works well in most weather. Haven't done much mud testing but for road and gravel, the setup work great.

I don't think I would recommend it for extended touring around South America though.
https://www.endless-sphere.com/forums/viewtopic.php?f=28&t=96581

 

[cal3thousand]

Will these be back in production soon? Please?