Bikeon, "cassette drive" torque sensing motor

Another question: to use a dedicated bike path (not part of a road) in California, an e-bike has to be class 1, which means no throttle and assist cutting out at 20 mph. Does that mean one has to disconnect from the BikeOn app, or is there a way to keep the app active and have the class 1 limit in place?
 
I believe we touched on this before but don't remember a definite answer... have you tested BikeOn with fully charged 52V batteries (up to 58.8V) and can these be used without damaging the motor or controller?
 
Gruesome said:
Does that mean one has to disconnect from the BikeOn app, or is there a way to keep the app active and have the class 1 limit in place?
Click to expand...
For now you will have to disconnect the mile app, if you want to keep it as class 1 device. It is on our to do list to have a different way of switching between class 1 and 3.
 
raylo32 said:
have you tested BikeOn with fully charged 52V batteries (up to 58.8V) and can these be used without damaging the motor or controller?
Click to expand...
That was not tested, but all electronic components are rated for well over 60V. A 48V battery was tested and that is the maximum we approve to use at the moment.
 
With some calculations i've determined that if you had an unrestricted controller ( let's say you don't care about warranty and run a third party controller ), the top speed in a 20" wheel on 52v would be around 31mph.

For all intents and purposes you should consider this motor to be a geared hub ( switching the sprocket size doesn't change the gearing of the motor )

FYI!
 
Does it really depend on the wheel size? Running the cycling simulator yields 600W wheel power (=750W input power with 80% efficiency) needed for 31 mph, dominated by air resistance.
cycling-600W-30mph.png
Under the same circumstances (flat ground, no head or tail wind, 220 lbs rider & bicycle) only 190W are needed for 20 mph, which I find to be about my top speed (without motor), give or take a few mph.

Also interesting: 55 Nm with a 700cx30mm wheel (2.15 m circumference) gives a tractive force of 55 Nm/0.34 m = 160 N, or just enough to push that 220 lbs bike & rider (100 kg) up a 16% grade. With a 20"x1.5" tire (1.49m circumference), I get 230 N, or enough for going up a 23% grade with vanishing human supplied torque. Of course with a pure pedal assist motor that doesn't work, you have to supply some torque yourself. It would be interesting to know the torque (or chain force?) multiplier the BikeOn uses for 100% assist.
 
Last edited:
Well if you have an unrestricted controller then you could push the ~1000w needed to maintain that speed, given that the motor is capable of higher continuous watts on higher voltage ( usually the case, but not always )

The stock watt limit would work for a semi recumbent just fine at that speed.. :)
 
Another question I have is how the BikeOn torque sensing deals with or recalibrates to changing gears, both on the front chainrings, and on the cassette.
I'm figuring you are dealing with a roughly 30mm chain height change in front, for example going from a 50T to a 34T chainring, and about 20mm in the back, going from say a 23 teeth sprocket to the smallest 11 teeth sprocket. That's a 20 to 25 mm variation in vertical chain position that the torque sensor has to deal with. Otherwise you'd get lots of support on the large sprocket in the back, and much less when you are on the fastest sprocket.
 
Maybe it's just always pushing down on the chain like a chain tensioner? It looks quite wide and smooth, so that shouldn't stop the derailleur below from pulling the chain to either side:
Screenshot_20240625-162950.png
 
Gruesome said:
Another question I have is how the BikeOn torque sensing deals with or recalibrates to changing gears, both on the front chainrings, and on the cassette.
Click to expand...
If I understand the question correctly, the answer is this: if the user switches to the smaller sprocket, the force applied to the torque sensor roller decreases (the angle between the incoming and outgoing portions of the chain is larger). This makes sense: if the user chooses the smaller sprocket, then speed is prioritized over torque.

If the user switches to the larger sprocket, the force on the sensor roller increases (the angle between the incoming and outgoing portions of the chain is smaller). Consequently, the sensor moves to a larger degree, resulting in more current flowing into the motor.
It is a bit more convoluted because these changes also affect the direction of the resultant force vector. I omitted it for the sake of simplicity.
lnanek said:
Maybe it's just always pushing down on the chain like a chain tensioner?
Click to expand...

The sensor arm is spring loaded indeed.
 
Last edited:
spaceman said:
If I understand the question correctly, the answer is this: if the user switches to the smaller sprocket, the force applied to the torque sensor roller decreases (the angle between the incoming and outgoing portions of the chain is larger). This makes sense: if the user chooses the smaller sprocket, then speed is prioritized over torque.
Click to expand...
Except when you switch to the smaller chainring in front. Then you typically want more torque and lower speed. But maybe I'm worrying about nothing.
spaceman said:
If the user switches to the larger sprocket, the force on the sensor roller increases (the angle between the incoming and outgoing portions of the chain is smaller). Consequently, the sensor moves to a larger degree, resulting in more current flowing into the motor.
It is a bit more convoluted because these changes also affect the direction of the resultant force vector. I omitted it for the sake of simplicity.


The sensor arm is spring loaded indeed.
Click to expand...
I should probably just wait and see how it works when I receive my kit!
 
Gruesome said:
I should probably just wait and see how it works when I receive my kit!
Click to expand...

Oh wow, someone finally signed up to be the test hamster for the forum 😅

Look forward to seeing your experience, i'm still going back and forth on ordering one myself!
 
Regarding the torque sensor: I measured a bit more, this time on the latest BikeOn advertisement pics in my mailbox, max-chain-height.png
and I think the maximum ( or minimum?? See below) chain height (measured as the distance of the chain from the rear-axle-to-pedal-axle center line) allowed by the torque sensor is actually about 38mm, which is very close to the minimum chain height (~35mm) you can get at that position with the smallest sprocket (11T) and chainring (34T), at least on my bike. So there is still the variable ratio between chain tension and torque sensor force due to the chain angle varying with sprocket size, but the offset in chain height screwing up the torque measurement I worried about earlier seems to be taken care of, except maybe to a very small extent (few mm at the highest torque) when using the very smallest sprocket. The variable chain angle should lead to relatively larger force on the torque sensor (at the same chain tension) for larger sprockets, so at the same chain tension you might get more motor support when on a larger sprocket, which is not entirely unreasonable.
Update: it's impossible to deduce from the picture whether what is shown is the maximum chain height, or the minimum chain height, or some value in between. If it were the maximum chain height, then that would take one systematic error out of the torque estimation, as written above. But if it is the minimum chain height (i.e. the torque sensor arm is pulling a slack chain all the way down), that would actually preclude measuring torque at all when using my smallest sprocket and chainring. Plenty to figure out once I receive my motor kit, hopefully next week.
 
Last edited:
A few notes from the inaugural run with the BikeOn on my Canyon Roadlite 6 street/gravel bike:
  • it works on my bike
  • set to 20% support, it already met two of my goals:
    • allow 20 mph on flat ground while bringing heart rate into the 135 - 145 bpm range (from 150-160)
    • compensate for headwind: going south, there was a 6-7 mph headwind, and the system dealt with that easily at the 20% setting
  • at 45% support, it effortlessly propels me at 25 mph+, which in my humble opinion is a bit fast without protective clothing; this might depend on how fast your skin heals...

I find it hard to judge how proportional (to human power input) the system is, but at the 20% setting the startup is very smooth, without any feeling that the bike or motor is getting away from you (I've experienced harsh starts on mid-motor e-mountain bikes, especially when trying to start uphill at a higher support setting).
The acceleration to 20 mph works very smooth and feels natural, with just the right amount of motor support as you increase the cadence and switch through the gears.
Going at a continuous speed of 20 mph feels also very smooth. This while I was putting in the power needed without motor for ~15 mph. On the way back north I checked that setting the support to 'None' (which keeps enough torque on the motor to overcome the braking torque of motor & belt) gives about the right speed, of 20 mph in this case with about 6-7 mph tailwind, maybe just a smidgen faster than without the system.

The 15 mile ride used about 2V or 13% of the 700Wh battery capacity, so about 6 Wh/mile.

I still have to test the uphill support capabilities.

A few pics:
1720326445597.png
1720326490811.png
It arrived well packaged.
1720326556065.png
All the components and spare parts it came with:
1720326612796.png

Installed on my bike, at the halfway point of a 15 mile ride:
1720326758523.png

I'll comment later on the installation process. This might have been the first 11 speed street cassette this system has been installed on.
Aram was extremely supportive, he made an installation video on one particular aspect just for me, and we exchanged several emails and spoke on the phone over the last two days. 👍👍👍
 
Last edited:
That is quite an interesting way to put a motor on a bike. I do have a pedal bike that I could foresee putting something like this motor on, from time to time.
How many watts does the motor draw when you're doing 20 mph on the level? You sound like you're putting in a fair amount at the cranks, but what are those watts, and what is the maximum wattage that the will motor draw?
 
Those are all good questions for Aram!
I do not know. From the battery drawdown and my speed I'm guessing the motor consumed about 120 W on average.
The 6 Wh/mile is pretty accurate. This is for a mix of 45-60% support and 15/20% support (I was adjusting the 'tour' and the 'sport' support levels during the ride).

I could see how somebody who wants to commute without having to shower after arrival at work would use the 45% (or higher) 'sport' level, which really feels quite leisurely at 20-22 mph; it's actually difficult not to accelerate further (for me). But I'd definitely want a better helmet and abrasion & fall resistant clothing for that, especially if commuting in traffic. But that is not my intended use.
 
Last edited:
Good report. My use case would be similar, an 11-speed gravel bike. Did you try it on any climbs?

As for speed I guess I am used to higher speeds with just lycra since we ride in an area with many rolling hills (and some bigger climbs) so some downhills are easily 30-40 mph on every ride. So 25-30 mph on the flats doesn't bother me... although I did go down once at that speed on a road bike crash back in the day and lost some skin. Do you live in a flat area?

What battery did you use? You said 700Wh, but what voltage? Your battery I assume since the BikeOn offering is smaller than 700Wh I believe.
 
You must log in or register to reply here.

Similar threads

neptronix
    • Like
Wife bike 2.0 ( Hiryuu Lady )
2
Replies
29
Views
1,210
neptronix
neptronix
M
  • Question
Bafang M635 Torque Sensing Motor Pedal problem
Replies
1
Views
126
Mongo
Mongo
ClimberBill
Felt F60 + 250w Front + Torque sensing BB
Replies
0
Views
235
ClimberBill
ClimberBill
jonsully
Torque-sensing AND Cadence-sensing — an idea
Replies
13
Views
781
jonsully
jonsully
L
Rear hub motor on aluminum frame
Replies
10
Views
550
99t4
9
Back
Top