recommendations for torque sensing hub motor system

elbiot

10 µW
Joined
Sep 1, 2014
Messages
6
Hi!

I am a bicycle commuter of 10+ years, considering a new much longer daily commute (15 miles each way plus 100+ degree weather). I can do this without assistance, but to be reliable daily for years, I'd like some electrical assist that doesn't lull me into laziness. From reading here and elsewhere, I'm thinking I'd like a small (<= 250 watts) geared brushless hub motor with torque sensing assist. I have seen xofo motors that fit this. But I'm uncertain about what controller and other things I would need. I would buy a lithium iron phosphate battery, probably around 9 A/hr. I could build a "new" bike for this. Currently I have a univega grand premio and specialized sequoia (touring/cargo). Also, $1000 is a lot of money for me, I'd really like to stay well below that number if possible.

Suggestions welcome.

Thanks,
Elliot
 
Welcome to Endless Sphere Elliot! It sounds like you've done your homework, well done. Thank you so much for asking your question and asking it so well! It sounds like you are asking the right questions, to me, and that you know what you are looking for.

There are some critical pieces of information missing from your post. It's pretty important that your battery pack voltage is compatible with the controller you use, if the voltage is too high from your battery pack, it can damage/destroy a controller that can't handle the voltage. I am guessing the rim you will use on your bike is 622mm, if that is so, I would probably suggest the Q100h 201rpm for your hub motor, I'd use a higher RPM motor if you are using something like a 406mm rim diameter or smaller. There are two other hub motors I might suggest, Q85 and Q75. The Q85 and Q75 are both smaller but comes in a front variety only.

http://www.bmsbattery.com/ebike-parts/570-pas-pulse-padel-assistant-sensor.html
http://www.bmsbattery.com/controller/552-s06-250w-imitation-torque-square-wave-controller.html
http://www.bmsbattery.com/controller/549-s-led770-e-bike-led-meter.html And/or http://www.bmsbattery.com/controller/670-s-led770-e-bike-led-meter.html *1
Q100h: http://www.bmsbattery.com/hub-motor/631-250w350w-q-85100sx-motor-e-bike-kit.html <Rear or Front>http://www.bmsbattery.com/front-driving/630-250w350w-q-85100sx-motor-e-bike-kit.html
If you are using a front hub, I'd probably suggest using a Q75 instead of a Q100h.

This setup seems to match up with your criteria very well, and comes in at a pretty decent price.

*1: I don't know if you can use one or the other to adjust the level of the pedal assist.

I've never used any of these parts, but I've looking over their website plenty, read a lot of things about these motors and it sounds like a pretty decent setup. I haven't found any geared hub motors that are as small or light, but if you do, I'd love to hear about them. Please be sure to very my suggestions, I could have made a mistake or am simply unaware of some important piece of information.

The common wisdom around these parts is when setting up an order from some place far away, that it is wise to order multiples of things. It would be a downer to have to an additional single component in the future with shipping for just one thing being so high. I also recommend looking into wiring options. Wiring is kind of a complicated subject, but I would guess you would want to trim the wires to appropriate sizes and put new connectors on their ends. Think of this as setting up brake and derrailuer cables of the correct lengths, a bad job will probably just look bad at best, and at worst may have functionality issues.

I hope others come by with great suggestions for you.
 
Thanks! 700C actually, but it doesn't look like it matters. So, the Q100 will take a cassette and give me multiple pedaling gears? Do I need to check my fork width?

Would an ungeared motor be lighter (I had the opposite impression)? I have no problem pedaling to maintain some rpm range if it means a lighter or more efficient motor.

Is that a cadence sensor? I'd really like an actual torque sensor as it would feel the most natural. Incorporated into the hub seems ideal. Could a front hub incorporate a torque sensor? I'd assume not but thought I'd ask.

I'll poke around that site some more tomorrow. Thanks again.
 
The Xofo motor with torque sensor is a very nice motor for low power. It comes with its own special controller, which has very good algorithms. It has some drawbacks though. When you pedal hard, like up a steep hill, your torque is greater than the springs in the sensor, so the sensor bottoms out, which gives a sort of knocking sensation, like when you have a loose crank; however, I guess you'd soon get used to it. It's not a problem. It just feels like one. Also, the torque is fairly weak compared with other motors of that size.

Modern cadence sensor systems are very good for general commuting, touring, etc. You can use current control on the S series controllers from BMSBattery if you want that authentic cycling feel, but IMHO, speed control is better for general cycling as long as the controller has several levels. The S-series controllers can switch between the two modes.
 
700c, 28" and 29" means 622mm rim size. I doubt I understand how torque sensors work, http://www.bmsbattery.com/91-torque-sensor but there they are on this website. I don't think it matters where the motor is on the bike relating to a torque vs cadence setup.

Just so were on the same page, http://sheldonbrown.com/free-k7.html < This explains the difference between a cassette and a freewheel. The Q100h that I linked to uses a freewheel, there is a Q100 (not the same as Q100h, but similar) that uses a casette(known as Q100c). http://www.bmsbattery.com/ebike-parts/48-14-28-teeth-shimano-freewheel.html I am pretty sure you can use a Q100h and 7 speed freewheel together between a 135mm dropout width. It probably wouldn't hurt to check the dropout spacing, but I'd guess your bikes have 100mm for the front and 135mm for the rear.

These Q75(1.2kg/2.65lbs, weights according to their website), Q85, and Q100h(1.9kg/4.2lbs) motors, as I said, are some of the lightest and smallest I have found. To my knowledge, geared hub motors are smaller and lighter and the Q75 is one of the lightest hub motors I have ever come across.

I think the lightest and smallest hub I've found is the Q75.
 
Hmmm, The price is right for that torque sensing BB, $109. The only other option I've seen is a $240 BB plus a $150 CycleAnalyst V 3. http://www.ebikes.ca/shop/ebike-parts/torque-sensors.html?SID=58f94d351078b0d8ff94758312779677

I wonder if there is a huge quality difference. Either way, I may have to build a bike especially for the BB. I haven't measured. The more expensive options have more dimensional diversity, so less chance of needing a specific bike.

I'm not sure about the cadence sensor plus torque simulating controller. I think that isn't really what I want.

I like the in hub option the best except for the knocking issue d8veh. I guess it is a bionix patented technology and costs $$$, but there are cheaper chinese rip offs like the XOFO XFL. http://www.aliexpress.com/store/323339/search?SearchText=torque&minPrice=&maxPrice=&filterAllSearch=false&freeShipCountry=

I'll have to think about this some more. The Q100h looks like a great motor for a great price.

More advice/experience still welcome. Thanks so far.
 
Have you ridden a bike with a cadence sensor? Most of the advice on here is from experience of high-powered motors with very crude single level power algorithms in the controllers. It's completely different with the modern controllers with LCDs. Torque multiplication systems are great for sporting riding, but not for general cycling. When you legs get tired going up a steep hill, you can't pedal hard enough to get the power you need, so you end up slowing down and stalling out. With a cadence sensor and speed control system, power increases as you slow down.
 
d8veh,

I haven't ridden with a cadence sensor, but your description here makes me still think I'd prefer a torque multiplier. Though, I'm not certain I totally understand what the cadence sensor and speed control system would feel like. Does it provide an amount of power proportional to your pedal speed, so that if I'm pedaling at y strokes per second, the motor is putting out x watts, regardless of what gear I am in? So it would assist a lot in low gears and much less at high gears?

What I don't want is a system that encourages or accommodates bad habits. I don't have a concern about my legs getting so tired going up a steep hill that I have to quit. Power increasing as I slow down sounds like it encourages trying less in difficult situations.

Thanks for your feedback
 
elbiot said:
d8veh,

I haven't ridden with a cadence sensor, but your description here makes me still think I'd prefer a torque multiplier. Though, I'm not certain I totally understand what the cadence sensor and speed control system would feel like. Does it provide an amount of power proportional to your pedal speed, so that if I'm pedaling at y strokes per second, the motor is putting out x watts, regardless of what gear I am in? So it would assist a lot in low gears and much less at high gears?

What I don't want is a system that encourages or accommodates bad habits. I don't have a concern about my legs getting so tired going up a steep hill that I have to quit. Power increasing as I slow down sounds like it encourages trying less in difficult situations.

Thanks for your feedback

I imagine you would be best served in buying a setup with this cadence system, installing and using it, while still reading and researching the idea of using a torque sensor so you have experience and knowledge to compare the two ideas. I've never used either, and I suspect the only way I could ever fully understand both these ideas is to use them personally. Which seems like what you will probably end up doing. I wouldn't rule out the idea of using a throttle either, I have a power limit (set through a cycle analyst) and a three speed switch that allows me to set things to a degree where the motor I use is to the level that I prefer (I like lower power than what most seem to).
 
I tried cadence PAS a few years ago and hated it, but that was a primitive "all or nothing" deal. As others have said that basic PAS system has evolved and now can offer various speed levels. I'm going to try one of these systems on my next build but I gotta tell ya I am very happy right now using the cruise control function of the controllers. The term "cruise control" however is a bit of a misnomer, it's actually more like speed control. In practice it works like this; I pedal along and bring up the throttle to just ease the effort slightly then let auto-cruise engage (a momentary switch can be used to enable the function instead of auto mode), then I simply pedal. If the grade increases the bike will slow, as I said it's not cruise control, but this is just like what happens for a regular cyclist. As the bike slows up the grade motor power naturally increases. It's not any kind of active control, it's simply the way the motor responds to a greater load. And just like a regular cyclist I will often have to downshift to maintain my cadence. When heading down a grade the bike speed picks up and motor output decreases. The rider upshifts and picks up speed. If the speed passes the original setpoint the motor output go to zero. On the road it feels pretty seamless, I sometimes forget the motor is even running. It works great and the best part is, it's simple. There are no extra wires or anything to install unless you want to mount a manual activation switch on the handlebars.

-R
 
Cruise control conventionally means constant speed, right? Cars would be more efficient with constant fuel flow I believe, I wish they had that. I'd actually probably be happy with constant energy or an efficiency seeking controller. I'm happy to pedal in accordance with the most efficient use of the tech.

I think the "try it and decide from there" perspective is correct. I'll get an awesome motor and batteries (and suitable controller, etc) and iterate from there. I'll start super simple. As a programmer with arduino experience, I might look into arduino for phase 1.5 stuff.
 
I meant to ask for clarification on "cruise control" and ask about efficiency seeking controllers in that post. If anyone knows anything... A quick google did not give a product.
 
elbiot said:
I meant to ask for clarification on "cruise control" and ask about efficiency seeking controllers in that post. If anyone knows anything... A quick google did not give a product.

This subject is a little hard to grasp, I think. Effeciency can mean more than one thing. If you mean nothing more than using the least amount of power concerning your electric bicycle components, then I may be able to help explain.

Higher temperatures are going to reduce efficiency. What is most likely to increase the temperature of things like your motor and controller involve 'high' (relative to their capacity) power output. It is more effecient, and things will stay cooler, if you accelerate more slowly. The way to achieve this is fairly simple, you set a watt/amp/power limit up. The way I choose set a power limit is through a cycle analyst. The power limit I set, that works for me, is in tune with an ideal top speed. Configuring this can be a little complicated, and I doubt I would really understand precisely how you could get there as we don't know what components you will use and I have no clue how this would be accomplished with something other than a throttle. But I am fairly confident that what will help you get to where you are going is focusing on the idea of setting up a power limit. Either a cycle analyst or a programmable controller are the routes I would suggest looking into. The cycle analyst would probably be the best all around here for this.

There is more to this subject, relating to a 3 speed switch (at least that is how I do it). But I am not sure how this would translate into your goals. How a three speed switch works is, you program a controller so that each speed turns the maximum level of the throttle into a different percentage, so if you set (this is a setting you can modify in a controller, which is kind of a complicated process) speed 1 to be 50% maximum throttle, that would mean if you had the throttle half way turned, you would only actually be at 25% throttle and having the throttle fully turned would only turn the throttle to 50% instead of truly 100%. This is useful in at least two situations. I won't go into the first one, but the second one means that if you find what level of throttle is suitable at a voltage to an ideal speed, you can set one of the throttle speeds to basically be sort of like what you are trying to achieve. Lets say we know that at 71% throttle at 52v my bike will be shooting for 20mph and that after that the power output will begin to ramp down and attempt to maintain that speed (RPM), I can then program that throttle percentage to one of the speeds (in the controller settings) and use that sort of like cruise control.

Combining a low maximum power (I use 300w and it serves me perfectly fine, I've ran with less and that was just fine as well for the most part) and setting up the three speed switch appropriately, I can run my bike in a way that is extremely efficient with no room for error. This is sort of like cruise control. I tried cruise control and I feel like it's far less than ideal, cruise control is easy enough to activate for that moment, but it's hard to get like the way I just described (impossible in the idea of genuine power restriction).

Not to bash anyone, but there are plenty of hot rodders on this forum who like to really pump out the power, having little concern for things like range or the ability to meaningfully pedal the bicycle still. It sounds like you really value being able to pedal plenty, so maybe the pedal assist level will work out for you, but what I described with a cycle analyst will allow you to setup a low power limit that is very easy to setup and change to how you like it. If you want 200w instead or whatever, that can easily be done with a cycle analyst. I think you will really value the ability to set a power limit to your liking.

I am sure there are other ways to achieve what I am suggesting, but this is what has worked for me so far.
 
elbiot said:
Cruise control conventionally means constant speed, right? Cars would be more efficient with constant fuel flow I believe, I wish they had that. I'd actually probably be happy with constant energy or an efficiency seeking controller. I'm happy to pedal in accordance with the most efficient use of the tech.

I think the "try it and decide from there" perspective is correct. I'll get an awesome motor and batteries (and suitable controller, etc) and iterate from there. I'll start super simple. As a programmer with arduino experience, I might look into arduino for phase 1.5 stuff.


What I have found is the single most important factor that affects battery power consumption between one E-bike and another for someone who contributes via pedaling isn't the electric bike conversion components, it's the BIKE chosen for conversion. I know this goes against the grain here where many say "oh just buy a cheap bike" to convert. Many of those folks however are basically building "mopeds" which they then ride at average speeds significantly above what they could do by pedaling that bike. However if you are after an E-bike which you will still be an active participant but want some relief from headwinds and hills at a pace that's higher than on your best day with pedaling only, and with less exertion, then the choice of bike IS important. For example, I presently have 3 E-Bikes; a flatbar road bike, a trekking bike and a (moped-like) cruiser. For one particular 31 mile loop I ride for the same average trip speed of 18 mph pedaling with normal effort (a bit less on the cruiser) I will typically use 7.5Wh/mi, 10Wh/mi and 13.5Wh/mi for each of those respective bikes.

If what you are building is essentially a moped that you drive at higher than normal bike speeds then the number one factor is SPEED. Unless you're going to push the speed envelope you can pretty much forget about this motor or that controller, simply buy the system which delivers the speed you want and the battery which can sustain that power. Frankly unless you choose a motor wound way too fast for your needs the power consumption is going to be very similar. In this instance aerodynamics will be the primary route to lowering power consumption.

-R
 
Back
Top