[confused] Torque sensor, gear ratio, pedal assist

[Winfried]

Hello,

I'm confused about how mid-drive motors with torque sensors work in conjuction with mechanical gear ratios.

Am I correct in understanding that the stronger you push on the pedals, the stronger the assistance from the motor?

And conversely, the faster you spin after selecting a lower/better gear ratio, the lighter you push on the pedals… and the less assistance you get from the motor?

If that is correct… it's in the user's interest to actually select a wrong (too hard) gear ratio to get strong assistance from the motor.

In other words… to make the best of the motor, the battery, and the transmission, the user should select the right gear ratio to get the right amount of assistance.

Is this correct?

Thank you.

 

[Doggyman1202]

Yes, if I understand your questions correctly. Someone here will probably offer a more detailed and better qualified explanation but if, for example, I want to ride north of 20mph on my Priority Current, I'll likely need to be in the highest gear (5th gear in this case). The amount of pedal assist has less to do with top speed, although it's obviously easier to get there with more assistance.

 

[harryS]

I'm of the opposite opinion. It's really not about cadence. It's the gearing. Select the gear that makes it easier for the motor and that usually means you will have to spin the pedals faster, hence a higher cadence. An extreme case is climbing a hill in too low of a gear, forcing the motor to work harder, If it's a mid drive, it's worse because the multiplication of the gears is lacking. I want to be climbing a hill where my pedals can spin faster, which means I'm helping.

Watch the roadie bikers. Even when they're going slower, they spin pretty fast, Motors like to spin faster too. More efficient. For an ebiker, that means more exercise, better motor life, and lower battery consumption. By the way, I ride a T/S midmotor, but also ride cadence sensor ebikes.

I'm sure Stefan will chime with the technical apsects,

 

[Winfried]

That's also what I thought (spinning at least 60-70 RPM to keep the motor in its sweet spot), but I read somewhere that some mid-drives with torque sensors at least provide less power the faster you spin because it figures you don't need much help if you can spin fast.

Hence the confusion.

Incidently, someone I know got stuck riding uphill with a Dutch cargo trike that had a Bafang mid-motor with a torque sensor and an Enviolo gear hub with an automatic shifter: In the "fast cadence" mode, she spun like crazy with the bike going nowhere (because the Bafang thought she didn't need help?), while with the "low cadence" mode, the motor did provide significant power… but the cranks would get stuck in place — we'll never know if it were because the shifter chose a gear ratio too big for hills and/or the bike came with a chainring+sprocket meant for flat terrain.

 

[Jeremy McCreary]

Welcome aboard!

I have a geared hub motor with well-implemented torque-sensing assist. Guessing that the following is largely true on mid-drives as well...

I have a very strong preference for a cadence of 85±5 RPM. To maintain that cadence on a given grade, I can choose a lower gear with lower assist level or a higher gear with more assist. The difference: In the latter case, I'll be going faster.

So to go substantially faster at constant cadence and grade, I upshift and increase assist at the same time. When the grade increases, I tend to maintain cadence first by downshifting, letting speed drop as it may. If speed gets too low in bottom gear, I start upping assist to maintain cadence.

Nothing magic about this — just an example based on the way I like to ride. I can monitor cadence and ground speed on my displays, but the gear/assist changes are usually just done by feel.

Having a 10-speed cassette and 9 assist levels makes it easy to dial in the combo of cadence, exertion, and speed I'm after at any given moment without thinking about it. For better or worse, I give ZERO thought to motor speed and efficiency, but I have plenty of battery for the riding I do.

Someone please correct me if the game is substantially different on a mid-drive.

 

[fooferdoggie]

That's also what I thought (spinning at least 60-70 RPM to keep the motor in its sweet spot), but I read somewhere that some mid-drives with torque sensors at least provide less power the faster you spin because it figures you don't need much help if you can spin fast.

Hence the confusion.

Incidently, someone I know got stuck riding uphill with a Dutch cargo trike that had a Bafang mid-motor with a torque sensor and an Enviolo gear hub with an automatic shifter: In the "fast cadence" mode, she spun like crazy with the bike going nowhere (because the Bafang thought she didn't need help?), while with the "low cadence" mode, the motor did provide significant power… but the cranks would get stuck in place — we'll never know if it were because the shifter chose a gear ratio too big for hills and/or the bike came with a chainring+sprocket meant for flat terrain.

all the higher end mid drives have their sweet spots on cadence. bosch its 80 rpms below that and the motor wont give its max. I think the max for the most assist is about 110 rpms. the motor still works but if you pushing up a super steep hill you wont get its all. I have a 20% grade hill that was my test bed. both my bosch powered bikes would get up it with me putting out 450 watts about 6mph and all I could do was about 40 to 60 rpms. that was with the older bosch motor and the gen 4. the old had 63nm of torque the new 83nm. but when I went from a 46 to a 42 chainring I got it up at about 8mph 80 rpms and I was able to put out 550 watts. I dont think I even hit the peak of the motor I ran out of steam before it did. I did a bunch of testing with cadence and 80 is the min to get the max out of it.

 

[Winfried]

So, unlike what I read, mid-motors with torque sensors do provide more power the faster you spin, so you really should downshift and spin like on a muscle bike so as to keep the motor in the zone it's most efficient?

 

[Doggyman1202]

So, unlike what I read, mid-motors with torque sensors do provide more power the faster you spin, so you really should downshift and spin like on a muscle bike so as to keep the motor in the zone it's most efficient?

The motor might be happier spinning away in low gear but you won't be. There's a reason higher gears are considered the fun gears. I'll always start in a lower gear to put less stress on the motor, but wind up in a high gear to put less stress on me. In other words, higher gears provide greater pedal efficiency.

 

[Jeremy McCreary]

Is it safe to assume that all mid-drives with torque-sensing assist also use true cadence sensors (not just pedal rotation detectors)? That all use the same algorithms to turn that sensor data into motor outputs? And that all their motors have the same cadence sweet spots WRT peak mechanical power, peak torque, and peak efficiency?

Seems extremely unlikely to me on all counts. Worse yet, most manufacturers hold the details of their sensor packages and assist algorithms very close to the chest. So hard to know what's really going on in these blackbox assist systems beyond what an astute rider like @fooferdoggie can ferret out in the saddle on a particular bike.

Maybe @Winfried could tell us the mid-drive makes and models they have in mind.

 

[Doggyman1202]

Is it safe to assume that all mid-drive manufacturers use both torque and true cadence sensors (not just pedal rotation detectors)?

No. My priority Current is torque sensor only. My other bike, an Evelo Omega uses both.

Worse yet, most manufacturers hold the details of their sensor packages and assist algorithms very close to the chest. So hard to know what's really going on in these blackbox assist systems beyond what an astute rider like @fooferdoggie can ferret out in the saddle on a particular bike.

Yes. I wonder what the secret sauce is on the Omega as I can't tell for sure when riding.

 

[Tom@WashDC]

You should probably do a long range test drive on a mid-drive with torque sensing over various terrain. Your confusion will dissipate.

 

[PedalUma]

Different systems feel different. Some have opensource programing. Some put more emphasis on cadence such as the Aventon Level 2 which can almost be ghost ridden and has some lag and overrun similar to a BBS02, although it is a mid-drive torque sensor bike. The ones I prefer reward fast, light, and smooth pedaling and do not have lag or overrun. You still need to provide watts of input to get more watts added. You still need to chew your steak, it is not spoon fed. It is rewarding and feels good.

 

[Tom@WashDC]

...It is rewarding and feels good...

That about sums it up!

 

[Jeremy McCreary]

Different systems feel different. Some have opensource programing. Some put more emphasis on cadence such as the Aventon Level 2 which can almost be ghost ridden and has some lag and overrun similar to a BBS02, although it is a mid-drive torque sensor bike. The ones I prefer reward fast, light, and smooth pedaling and do not have lag or overrun. You still need to provide watts of input to get more watts added. You still need to chew your steak, it is not spoon fed. It is rewarding and feels good.

Good points, but 2 questions:

Q1. I've ridden a friend's 2023 Aventon Level.2 — a torque-sensing hub-drive similar in most respects to my 2023 Surface 604 V Rook. The Aventon may well have a cassette or crank rotation detector for safety reasons, as mine does.

But do you really think its assist system measures and figures in actual crank RPMs?

The Aventon's assist feels much like mine, and mine has no reliable way to measure RPMs at the crank. My TMM4 torque sensor senses only 2 things: (1) Rear dropout strain due to chain tension — a good proxy for crank torque at fixed crank length — and (2) the presence or absence of cassette rotation. (Motor's enabled only when the cassette is turning.)

If my Rook has a sensor capable of returning a reliable crank speed, it's well-hidden. Certainly none in or around the bottom bracket.

Q2. What do you mean by "overrun" in this context? Same as surging?

 

[fooferdoggie]

So, unlike what I read, mid-motors with torque sensors do provide more power the faster you spin, so you really should downshift and spin like on a muscle bike so as to keep the motor in the zone it's most efficient?

yes more like a regular bike. you keeping the motor in the most efficient range. but its really the best way to ride anyway it gives you more power and more efficiency so you can ride longer.

 

[PedalUma]

Good points, but 2 questions:

Q1. I've ridden a friend's 2023 Aventon Level.2 — a torque-sensing hub-drive similar in most respects to my 2023 Surface 604 V Rook. The Aventon may well have a cassette or crank rotation detector for safety reasons, as mine does.

But do you really think its assist system measures and figures in actual crank RPMs?

The Aventon's assist feels much like mine, and mine has no reliable way to measure RPMs at the crank. My TMM4 torque sensor senses only 2 things: (1) Rear dropout strain due to chain tension — a good proxy for crank torque at fixed crank length — and (2) the presence or absence of cassette rotation. (Motor's enabled only when the cassette is turning.)

If my Rook has a sensor capable of returning a reliable crank speed, it's well-hidden. Certainly none in or around the bottom bracket.

Q2. What do you mean by "overrun" in this context? Same as surging?

Q1. Cadence sensing is supposed to measure 'cadence,' meaning pedaling speed. Many of the older style ones do feel disconnected from cadence and seem to be either Off or On. The one's that I consider nice ramp up assist on a curve that increases with pedaling speed, leveling assistance at around 80 rpm. With the ones that I like best pedaling speed is always the same as chainring speed on the mid-drive. On the Bafang BBS02 and HD for example, chainring speed is disconnected from pedaling speed on the mid-drive. On those types you can pedal at a cadence of 15 rpm with no pedal pressure and be going 45kph, 28mph.

Q2. Overrun. On many systems and typical older systems. A) Assistance does not kick in until after thee magnets on the crank have passed the cadence pickup. Then you get a surge of power. That is what I call lag. B) They are also setup so once you stop pedaling it still gives assistance for about three seconds. That part B is what I call overrun. I find it disconcerting. It also makes technical ridding almost imposable if you are trying to get around roots and rocks on a climb.

 

[mschwett]

Good points, but 2 questions:

Q1. I've ridden a friend's 2023 Aventon Level.2 — a torque-sensing hub-drive similar in most respects to my 2023 Surface 604 V Rook. The Aventon may well have a cassette or crank rotation detector for safety reasons, as mine does.

But do you really think its assist system measures and figures in actual crank RPMs?

pretty sure they do, otherwise they’d have no idea of the power, since power requires knowing speed! since many mid drive motor designs work as a ratio of rider power they’re going to need to know rider rpm, rider torque, motor rpm, and motor torque, although whether the last one is measured or simply known from the power input side, not sure. wouldn’t be surprised if some of the intricacies here are why mid drive makers can be particular about cranks and chainrings.

we’ve seen charts of the mahle motor in the specialized mid drives which show how quickly efficiency drops off below a certain motor speed, trending towards zero. i wouldn’t be surprised if this contributes to their desire for relatively smooth assist curves which ramp up somewhat slowly as your pedal RPM increases.

 

[PedalUma]

There are a series of stepdown gears inside the motor housing to take it down to human scale. Given that, the motors I like best run best at 4,000 rpm before gearing steps that down to a cadence of 80 on the cranks/chainring. There is torque sensor lag but it is running at 1,000 Htz so it is not perceptible.

 

[Jeremy McCreary]

pretty sure they do, otherwise they’d have no idea of the power, since power requires knowing speed! since many mid drive motor designs work as a ratio of rider power they’re going to need to know rider rpm, rider torque, motor rpm, and motor torque, although whether the last one is measured or simply known from the power input side, not sure. wouldn’t be surprised if some of the intricacies here are why mid drive makers can be particular about cranks and chainrings.

we’ve seen charts of the mahle motor in the specialized mid drives which show how quickly efficiency drops off below a certain motor speed, trending towards zero. i wouldn’t be surprised if this contributes to their desire for relatively smooth assist curves which ramp up somewhat slowly as your pedal RPM increases.

Sorry, I'm confused. My neighbor's Aventon Level.2 is a torque-sensing hub-drive, and so is my ebike.

My ebike can only know crank torque, as it has no reliable way to measure crank speed — in which case, it can't know my input power, as you point out. But it still manages to deliver very natural-feeling assist.

Many reasons to believe that the Aventon doles out assist in much the same way as my bike, but your hub-drive could well be more sophisticated.

Assist terminology drives me nuts! The "cadence sensor" on most "cadence-sensing" ebikes doesn't measure cadence (crank RPM) at all. It just detects the presence or absence of crank rotation. And a "torque-sensing" assist system that also measures and factors in true cadence would be better described as a "power-sensing" system.

My ebike is torque-sensing but not power-sensing.

 

[Winfried]

I have a geared hub motor with well-implemented torque-sensing assist. Guessing that the following is largely true on mid-drives as well...

Out of curiosity, what's the brand? How does that torque sensor look like?