Commuter Booster - <1kg Friction Drive

In my attempts to eliminate RF interference to the HRM, I experimented with ESC switching frequency. With 16kHz I couldn't get rid of the emissions, but I came to conclusion that 16 kHz can possibly improve the overall efficiency. Sounds strange, right? :) The reason is, the motor runs quieter at 16 kHz, that's because it runs smoother. Previously, I got some slippage at low speeds using high power. With smooth running motor, it almost never slips now. So, less tyre engagement can be used, which leads to improved efficiency.
 
How useful is this setup? I skipped this stage altogether and went straight to current limiting circuit, and then to Arduino... I presume you need to adjust the throttle as speed increases, right?
 
We all have motor axles hanging on one side completely unused... How about fitting a small flywheel there? The motor will run smoother, and there will be less slippage, if any, and less tyre wear, be it with or without grip material. Did anyone try fitting flywheels?
 
Now that this topic seems to have re-ignited a bit, I have a question for anyone who might be able to answer:

Would it be possible to 3D print the clamp/pivot bearing that Adrian made out of acetal? Would either ABS or PLA be strong enough to handle the forces that would be involved? Would anyone with an engineering background be willing to try to make something up?

Also - Adrian - a page or two back you'd commented on tire pressures that you'd tested - what pressure do you run in the back tire on your road bike that had the Commuter Booster attached? I'm a bit of a Clydesdale and have to run 120-130 PSI, and I agree with others that the (700x23c) tire is so rock-hard that it's difficult for the motor to get enough "bite", and to get the bracket adjusted properly.

Lastly - Windtrader - your "Core" items (motor mount and pivot/stop mechanism mount) would certainly be of interest to me, as I have the motor, batteries and most of the rest of the gear from prior experimentation. Here's hoping you can bring something to fruition!
 
I have a couple of questions about how this setup would work for my specific needs:

1) Use only at quite low speeds going up hill
2) Use with a 20" rear tire (recumbent). I have plenty of room to mount and not worried about that part of it.

So I need this to work at a pretty low speed/high torque.

Only looking for around 300 watts of boost.
 
How steep are your hills, and how heavy are you together with your recumbent?
From my experience, going up 10% hills requires 700W to cycle at 20+kph. Me plus my bike weigh 80kg. With lower power at this kind of slope, the speed drops dramatically. Below 10kph, I believe, the efficiency of the motor drops, and it really starts struggling. That’s why I prefer high power for up hills. The motor will have less chance of overheating at higher speeds, despite higher power feed.
For hills of gentler slope, 500W is enough for me. And 300W is good for flats only. Remember, 80-150W will be lost for deforming the tyre.
 
P.S. The best setup for low power, low speed climbing is mid-drive. That’s because you can select the appropriate gear for the motor, so it will run efficiently no matter the speed.
 
Folken said:
Remember, 80-150W will be lost for deforming the tyre.

Where did that number come from? That would depend on what kind of setup you have and tons of other factors. So by that theory if you have a 150w friction drive it will add zero assist. I don't think so. With a pivoting or sliding mount, once you are past the initial acceleration the roller isn't deforming the tire at all. In fact, it would take a crapload of power to "deform" a tire with friction drive in the first place. I'm running well over 1000 watts with my friction drive (since 2009) and I've never experienced anything like that.

As for hill climbing. I tried a 36v Bafang mid drive last year for a bit. My 36v friction drive will hands down embarrass a 36v Bafang any day. Up hill, on flats, whatever. Well, except in rain. Friction drive sucks in the rain. I was so disappointed with the Bafang that I sold it right away. Best advice is to use a smaller roller instead of the motor can if you want great hill climbing ability with friction drive.
 
This power is measured with the wheel off the ground, while holding the pivot arm in engaged position. Depending on speed, the motor is using 80-150W (upper number is for ~40kph). Free spinning motor completely off the tyre consumes 45W already at the same 40kph.
I have to admit though, I had to mount the motor pretty high on the seat tube to reach full engagement at about 150W. This amount of engagement is OK for up to 700W, then slippage begins. If I mount it lower, it sometimes wants to disengage because my tyre is not round enough.
 
badboybike said:
EVTodd said:
Folken said:
Remember, 80-150W will be lost for deforming the tyre.



As for hill climbing. I tried a 36v Bafang mid drive last year for a bit. My 36v friction drive will hands down embarrass a 36v Bafang any day. Up hill, on flats, whatever.

oh come on, if power is the same you can't make such statement...a friction better than a middrive..Honestly..this is mechanic and physics not opinions..

The power isn't the same at all. It's the same voltage but much different watt output. My drive also has a top speed of 28-30 mph. What's a Bafang bbs01? Maybe 20? And yup, My drive KILLS a bafang bbs01 up hills. Throw mine in a low gear and pedal with it just like a Bafang and it pulls me up any large hill in this area. If you're ever in Illinois let me know and I'll let you try it for yourself.

In fact, here's a friendly offer... I'll put my bike up against any other 36v ebike that's out there. I have a nice hill in my neighborhood were we can compare the two and post the results here. The only drive that would beat it that I can think of would be an rc powered mid drive. Nothing else has enough power at 36 volts.
 
Folken said:
How steep are your hills, and how heavy are you together with your recumbent?
From my experience, going up 10% hills requires 700W to cycle at 20+kph. Me plus my bike weigh 80kg. With lower power at this kind of slope, the speed drops dramatically. Below 10kph, I believe, the efficiency of the motor drops, and it really starts struggling. That’s why I prefer high power for up hills. The motor will have less chance of overheating at higher speeds, despite higher power feed.
For hills of gentler slope, 500W is enough for me. And 300W is good for flats only. Remember, 80-150W will be lost for deforming the tyre.

The hill grade will vary but in general want to use the boost only on very steep short pitches. Right now I occaisonally get down to about 3 mph in spots. I don't need the power for anything but these very slow bits. I'd just like to boost that 3 to 5 mph so I'm going 6 to 8 mph instead of 3.

I am also considering a very light weight crank drive, but the recumbent already weighs 30 pounds and I don't really want to add much more weight since I'll only use the power for very steep hills, but the weight is there for every hill. I'm 170 lbs myself.

A mid drive would work easily also as I have just one gear in front. The rear is a SRAM 3 x 7 hub/cassette.

If I wanted to use a lower gear ratio by using a small roller instead of the motor, I'd need to mount the motor outboard, right? I'll read through some of the other posts about that type of setup.
 
Hi guys,

Have you all been doing this by hand? or do you have CAD files for it? I'm thinking of trying out E-machineshop, and this looks like a great project to start with. if anyone has cad files they'd be willing to share that would be awesome.
 
mclark999 said:
If I wanted to use a lower gear ratio by using a small roller instead of the motor, I'd need to mount the motor outboard, right? I'll read through some of the other posts about that type of setup.

Yup, and that's the key to getting good hill climbing ability with a friction drive. I use a 1.25" roller on mine.
 
EVTodd said:
In fact, here's a friendly offer... I'll put my bike up against any other 36v ebike that's out there. I have a nice hill in my neighborhood were we can compare the two and post the results here. The only drive that would beat it that I can think of would be an rc powered mid drive. Nothing else has enough power at 36 volts.
It's useless to talk Volts, let's talk Watts instead. My 22V bike is the fastest among ALL electric bikes in my neighborhood, so does it prove that friction drive is the fastest? Of course it doesn't. :)
 
Folken said:
EVTodd said:
In fact, here's a friendly offer... I'll put my bike up against any other 36v ebike that's out there. I have a nice hill in my neighborhood were we can compare the two and post the results here. The only drive that would beat it that I can think of would be an rc powered mid drive. Nothing else has enough power at 36 volts.
It's useless to talk Volts, let's talk Watts instead. My 22V bike is the fastest among ALL electric bikes in my neighborhood, so does it prove that friction drive is the fastest? Of course it doesn't. :)

The point isn't that it's the fastest. My point is that it's the best bang for the buck. Like I always say on here. No one wants to believe it because they've spent so much money on their bike. How much would you have to spend to match what I have if you use a hub motor? A hell of a lot more than I did. The Bafang bbs01 I got last year was more than double what I spent on my total friction drive setup (including battery) and that's before you get a battery for it.

And I disagree, voltage is important. I run fewer batteries which saves weight and money yet still get great top speed and hill climbing ability, for me anyway.
 
@Folken

I'd like to build a friction drive with the very same electronic parts you used : arduino, hall sensor for current control, ... I have very basic electronics knowledge and I'm not able to design your circuit like you did. Looking at your pictures (the one with the box opened and your arduino code), I understood the general wiring but I would not want to miss something as an electronics noob. There has to be some caps, diodes, here and there I don't see in picture :D.

I barely see the complete schematics you have on paper under your box picture so if you would not mind to share it, I'd be very grateful ! I'm afraid I could not mimic your drive system without all the detail you designed :( .
 
I'm sorry to admit there was never a schematic drawn for this one. I've added a couple of 1uF ceramics to the pins reading the voltage and current. No diodes, transistors, etc. - Arduino does everything for you. :)

Also, I wasn't very lucky with the quality of the 5V supply voltage coming from the ESC. A week ago, the circuit started behaving strangely - random reboots, incorrect voltage and current readings, etc. This has been fixed by adding a 5V linear regulator capable of 25V input. This wastes a couple of watts, so a better approach would be to use a switchmode regulator. These are readily available for Arduinos and are easy to find online. I strongly recommend not to use the 5V coming from ESC's BEC. You never know what they put in there or how it will behave at 90+ degrees which ESCs can easily reach.

PS. The schematic in my pictures is the Arduino Nano schematic from the official website. And in the end I didn't even need it, because all the pins have their names in silkscreen on the actual board.
 
Thanks for your advices.

I would have directly used the 5v from DLUX ESC :).

So if I try to sum up the components I have to use .

Resistors :
- pull down resistor from A3 input to GND
- 2 resistors to make voltage divider to read Batt+ on A2 input.

Capacitors :
- between A2 and GND to read battery volage
- between A6 and GND to read hall current sensor

Regulator :
- to power arduino with true 5v from battery
 
Masure said:
Thanks for your advices.

I would have directly used the 5v from DLUX ESC :).

So if I try to sum up the components I have to use .

Resistors :
- pull down resistor from A3 input to GND
- 2 resistors to make voltage divider to read Batt+ on A2 input.

Capacitors :
- between A2 and GND to read battery volage
- between A6 and GND to read hall current sensor

Regulator :
- to power arduino with true 5v from battery

That's exactly right! A3 is then used for throttle (Hall or button to 5V), A4 and A5 are I2C bus lines for the LCD.
 
I ordered this regulator.

I'm a bit worry to have more connections to the battery leads because I must be very clean with my assembly to prevent any shorts :). Did you make any insulation around the hall connections ?

I have to buy some capacitors but no idea which values to take. I will google around.
 
3A looks pretty good! You may be able to power your lights from it as well. :) My Arduino+LCD (incl. backlight) consume 110mA.

Isolation requirements depend on your mechanical design. I didn't have to isolate the Hall sensor terminals because they are fixed in the case and not floating around.

Any caps in the microfarad range will be OK. Sometimes I take caps and other components off old boards which doesn't cost me a cent. I took that linear regulator from an old sound card. :D
 
Back
Top