Experience with hub-drives with BOTH torque- and cadence-sensing assist?

[Jeremy McCreary]

Considering the new Velotric Breeze 1 hub-drive for my wife — in part, because it allows you to choose between torque- and cadence-sensing assist while riding.

Anyone have experience with this dual-assist scheme on any ebike?

I'd never choose cadence-sensing assist for myself, but it might make sense for my wife, who likes to pedal but doesn't like the effort needed to ride the torque-sensing hub-drive she has now. Ghost-pedaling might be more to her liking at her current fitness level, but I just can't bring myself to spend money on a cadence-sensing-only ebike.

My evil scheme: Let her enjoy cycling with me with the cadence-sensing option, then switch when she's built up enough strength and stamina to appreciate the advantages of torque-sensing assist.

NB: Mid-drives are out, as she demands a throttle, and I'll only buy a bike we can test from an LBS able and willing to service it. Luckily, we have Aventon and Velotric dealers nearby.

Thanks!

 

[6zfshdb]

My Pedego Platinum Interceptors have both torque and cadence sensors. I almost never use cadence.
Still, it's a backup if I get injured and can't exert much pedal pressure.

 

[Base]

I do have experience with the combined sensor with a trike and a big directdrive hubmotor. The systems operates in way that it first uses the rotational sensor fro the first pickup and than the combi of torque combined with the crank rpm's to calculate the motor torque. With our firmware you cannot choose between the 2, but other trikes use a similar controller but with a geared motor, a rotational sensor and a pure sinewave firmware. Thats a completely different story in drive experience. The rotational units feel really on/off although the controller firmware is rather refined for this type of system.

I think some testrides and proper instructions would result in the best advice and experience riding.

 

[Jeremy McCreary]

I think some testrides and proper instructions would result in the best advice and experience riding.

Absolutely. No point replacing one ebike she doesn't like with another.

 

[6zfshdb]

In addition to torque sensing and a throttle, you might also consider a bike with cruise control. It made a big difference for my SIL, who now rides much more often.

 

[Doom1]

I have a Velotric Nomad 2 which allows switching between cadence and torque by just pushing a button. It's really great having that option because I use them both for different reasons.

 

[Chazmo]

My Juiced RipCurrent S fat bike is supposed to use both torque and cadence sensors. When turned “on” torque sensor is in some kind of combined mode with cadence. When turned “off,” the torque sensor is presumably disabled and only cadence is used. I have not ever turned off the torque sensor (easily done with the display). I’m going to try that.

Having said all that, I don’t even know where the cadence sensor is on my bike. I’ve been trying to learn as much as possible about it since Juiced went in the toilet, but I don’t understand this. Presumably it’s in the bottom bracket at the crank (where else would it be?), but there’s no cable from there that goes to the controller, so I don’t know how this works. The torque sensor on the other hand is called out as some electronics attached to the cassette-side rear dropout, and there definitely is a cable for that. Is it possible that piece of electronics at the rear dropout does both?

 

[Jeremy McCreary]

I have a Velotric Nomad 2 which allows switching between cadence and torque by just pushing a button. It's really great having that option because I use them both for different reasons.

Excellent! Could you elaborate on when you prefer which one?

Bears on how an added cadence-sensing option might address some of my wife's gripes about her torque-sensing hub-drive.

 

[Doom1]

Excellent! Could you elaborate on when you prefer which one?

Bears on how an added cadence-sensing option might address some of my wife's gripes about her torque-sensing hub-drive.

I commute to and from my office and I use the cadence sensor (with the throttle sometimes) for that since I don't want to show up to work all sweaty. On weekends, I'll switch to torque sensor in order to get in some exercise and, you know, just ride a bike like a bike.

 

[Jeremy McCreary]

The torque sensor on the other hand is called out as some electronics attached to the cassette-side rear dropout, and there definitely is a cable for that. Is it possible that piece of electronics at the rear dropout does both?

Yes, quite possible. My hub-drive has a nicely implemented torque-sensing PAS based on the TMM4 rear dropout torque sensor with built-in cassette rotation detector. The latter provides the same info to the controller as a typical so-called "cadence" sensor at the crank. No external sensor at the crank, so I assume that the TMM4 has taken over that job.

I put "cadence" in quotes here, as these yes-no sensors detect only the presence or absence of rotation. All ebikes need this info to disable the motor assist when the crank/cassette isn't turning.

True cadence in RPM is never measured in most torque-sensing ebikes, but more sophisticated motor power algorithms do take it into account. The best assist I've ever ridden factors in true cadence by measuring rider power and torque directly. True cadence can be calculated, as rider power is proportional to rider torque × cadence.

Measured rider power is the main determinant of motor power in this power-sensing PAS. Cadence enters indirectly through rider power but may also enter directly in special situations like starts.

 

[linklemming]

I think the use case mentioned by Doom1 is the perfect example

I have a PAS option on my DIY Grin GMAC hub motor bike. I originally set it up this way due to the torque sensor not working after crossing water puddles (still happens despite sealing all connections across 3 different torque sensors...go figure). It is a nice backup option to have and I have used it on several other occasions like knee acting up during a ride, overly hot temps causing me to be on the edge of heatstroke and running out of water, or just wanting to cruise on an off day The PAS on the Grin is power based per assist level.

Some of the cheaper PAS are speed based and I wouldnt recommend one of those. It can work if you also muck with the throttle but that seems like a bandaid.

My juice CCX (circa 2019) which had the same TMM4 sensor you mention had the option to use just PAS. Ironically this option seemed to limit the overall power as well. It like it was operating at 50% power on PAS only and the torque sensor just added the rest. Probably an edge case but worth mentioning.

 

[Chazmo]

Yes, quite possible. My hub-drive has a nicely implemented torque-sensing PAS based on the TMM4 rear dropout torque sensor with built-in cassette rotation detector. The latter provides the same info to the controller as a typical so-called "cadence" sensor at the crank. No external sensor at the crank, so I assume that the TMM4 has taken over that job.

I put "cadence" in quotes here, as these yes-no sensors detect only the presence or absence of rotation. All ebikes need this info to disable the motor when the crank/cassette isn't turning.

True cadence in RPM is never measured in most torque-sensing ebikes, but more sophisticated motor power algorithms do take it into account. The best assist I've ever ridden factors in true cadence by measuring rider power and torque directly. True cadence can be calculated, as rider power is proportional to rider torque × cadence.

Measured rider power is the main determinant of motor power in this power-sensing PAS. Cadence enters indirectly through rider power but may also enter directly in special situations like starts.

Outstanding. That explains it, Jeremy! Though I’ve removed the rear wheel to repair stuff, I have not removed the sensor on my bike so I don’t know exactly what it is, but no doubt it’s providing the cadence/rotation info. Perhaps it’s even reporting wheel rotation info as well (for the controller to calculate speed). Thanks so much. I’ll have to test ride the bike with torque sensing off and see what that feels like and what the controller actually does. Also, next time I have the wheel off (probably to clean it soon), I’ll have a close look at the sensor. I’ve been getting an occasional “err 8” from my Juiced controller which has something to do with a Hall effect… I think that might be the torque sensor causing the intermittent fault. Plus I have a spare bike to raid for parts, so maybe if I need another…

Here’s a pic of the torque sensor on my spare parts (very dirty) bike.

 

[Chazmo]

...

My juice CCX (circa 2019) which had the same TMM4 sensor you mention had the option to use just PAS. Ironically this option seemed to limit the overall power as well. It like it was operating at 50% power on PAS only and the torque sensor just added the rest. Probably an edge case but worth mentioning.

Thanks, Link… I will see if I have the same experience on my RCS (2022).

 

[Jeremy McCreary]

Outstanding. That explains it, Jeremy! Though I’ve removed the rear wheel to repair stuff, I have not removed the sensor on my bike so I don’t know exactly what it is, but no doubt it’s providing the cadence/rotation info. Perhaps it’s even reporting wheel rotation info as well (for the controller to calculate speed). Thanks so much. I’ll have to test ride the bike with torque sensing off and see what that feels like and what the controller actually does. Also, next time I have the wheel off (probably to clean it soon), I’ll have a close look at the sensor. I’ve been getting an occasional “err 8” from my Juiced controller which has something to do with a Hall effect… I think that might be the torque sensor causing the intermittent fault. Plus I have a spare bike to raid for parts, so maybe if I need another…

At least some hub-drives have wheel speed sensors built into their motors, as in my Bafang G020.

The TMM4 has no reliable way to measure wheel speed. Doesn't even measure cassette speed, just the presence or absence of cassette roation. Since the chain doesn't slip, the same yes-no answer applies to a hub-drive's pedal rotation. Mid-drives are more complicated.

 

[Chazmo]

At least some hub-drives have wheel speed sensors built into their motors, as in my Bafang G020.

The TMM4 has no reliable way to measure wheel speed. Doesn't even measure cassette speed, just the presence or absence of cassette roation. Since the chain doesn't slip, the same yes-no answer applies to a hub-drive's pedal rotation. Mid-drives are more complicated.

Hmm… Well I guess it’s probably the G062 Bafang hub motor that’s measuring wheel rotation then. There’s no other sensor for that either. But, the cable that runs to the controller has only three wires, Jeremy. Is this cable carrying rotation data to the controller in addition to providing battery power from the controller to the motor? Any idea how that’s done? I presume it’s not being done over the power wires, but I may be out of my depth here...

 

[Base]

TMM4 sensors are Torque sensors, they were created in Eindhoven Noord-Brabant. Used on the ION bikes from Sparta and later Batavus and Koga. Later on these were used on the BionX IGH rear wheel motors and licensed to other vendors. Also the Gazelle system worked with that sensor.

Most rear wheel motors use a form of Hall sensor to measure rotation and calculate the Km/h. I am not near the office at this moment so I cannot "show" you the contents of a motor with hall senor. Usually just a few magnets on the hub shell (6 or 8) ands a pcikup on the pcb. cheap bikes also use a hall sensor on the bb axle to measure the crancks movement.

found a picture online of the hubshell with magnets

 

[Jeremy McCreary]

Hmm… Well I guess it’s probably the G062 Bafang hub motor that’s measuring wheel rotation then. There’s no other sensor for that either. But, the cable that runs to the controller has only three wires, Jeremy. Is this cable carrying rotation data to the controller in addition to providing battery power from the controller to the motor? Any idea how that’s done? I presume it’s not being done over the power wires, but I may be out of my depth here...

Sorry, out of my depth here, too. I've always assumed that my Bafang's single motor cable includes a separate wire for the wheel speed signal, but I don't know much about ebike wiring and controllers.

 

[Chazmo]

Sorry, out of my depth here, too. I've always assumed that my Bafang's single motor cable includes a separate wire for the wheel speed signal, but I don't know much about ebike wiring and controllers.

Thanks so much, Jeremy. If I ever get the nitty gritty details I’ll let you know! Much appreciated!

 

[Chazmo]

@Jeremy McCreary any progress/updates on this? Did you make a choice for your wife, or are you considering some choices? Just curious.

 

[Jeremy McCreary]

@Jeremy McCreary any progress/updates on this? Did you make a choice for your wife, or are you considering some choices? Just curious.

Still waiting to test the Velotric Breeze 1, due to arrive at a nearby LBS in early April. We'll try some Aventons and some cruisers there as well.

She really enjoyed the American Flyer E-wave cadence-sensing hub-drive cruiser we rented in Huntington Beach last year — way more than her current torque-sensing hub-drive commuter of similar power and weight. Still unclear to me why, as she's unable to articulate the difference. But the cadence-sensing assist and somewhat pedals-forward geometry may have been part of it.

The setting was probably also a factor — totally flat bikeways largely protected from traffic there vs. unavoidable hills and good but less protected bikeways here at home.

I see a lot comparison testing in our future.