"Current-based" cadence-sensing from Ride1up

[Jeremy McCreary]

Trying to understand the advanced "current-based" (CB) cadence-sensing assist now offered on some Ride1Up models. Follow the link for their description.

Sounds like CB assist doles out power as a function of motor current, up to a fraction of max motor power set by assist level. Ride1Up claims that motor current can be tied somehow to "desired exertion" without measuring rider torque.

Question: Anyone know of an exploitable connection between desired exertion and motor current?

 

[PedalUma]

Trying to understand the advanced "current-based" (CB) cadence-sensing assist now offered on some Ride1Up models. Follow the link for their description.

Sounds like CB assist doles out power as a function of motor current, up to a fraction of max motor power set by assist level. Ride1Up claims that motor current can be tied somehow to "desired exertion" without measuring rider torque.

Question: Anyone know of an exploitable connection between desired exertion and motor current?

I read it and it was just marketing jargon smoke and mirrors. They did not explain what it does and how it does it, just that it is better, trust me.

A cadence sensor that is not just off or on would be a big improvement. It would increase assistance on a cadence curve, so when you pedal faster it gives you more assistance. Downshift, rev up and go.

 

[PCeBiker]

Trying to understand the advanced "current-based" (CB) cadence-sensing assist
Sounds like CB assist doles out power as a function of motor current, up to a fraction of max motor power set by assist level.

That sounds just like a KT controller.
Kunteng calls it "Imitation Torque Control".

The difference with Ride1Up is that you only need 10° of crank rotation to turn the power back on, so they must have a 36 magnet PAS sensor.

Regular PAS sensors have a maximum of 12 magnets.

The KT controller allows for a 5, 6, 8, 10, or 12 magnet cadence sensor input.
12 magnet would be 30° of rotation.

So maybe Ride1Up has developed a cadence sensor that gives 36 pulses per revolution, or they are making stuff up.

 

[PCeBiker]

Sounds like CB assist doles out power as a function of motor current, up to a fraction of max motor power set by assist level.

That's how Kunteng works.

My KT-LCD8H display allowed me to cut my controller power in half, so I turned my 25 amp (48v) controller down to 12.5 amps.
Then I turned the power for first Throttle Gear (or PAS mode if I chose to use my PAS sensor), down to 20% of maximum power, so 2.5 amps or 125 Watts when my battery is at 50 volts.

The KT-LCD8H allows me to adjust first Throttle Gear (or PAS mode setting) power between 20% and 60% of maximum power with maximum power being 5th gear.

Each Throttle Gear (or PAS mode setting) increases the power output progressively up to 12.5 amps (or 25 amps if I turn up my controller).

Each Throttle Gear (or PAS mode setting) also has a speed limit and if you can get to that speed with the alloted power, then the power reduces to maintain the preset speed for that Gear (or PAS setting)

 

[harryS]

KT controllers are current based, and I think they feel better on pedal assist than other brands of controllers I owned.

This is my belief on how it works. Easier to visualize it with a front drive bike, Total speed depends on power to both wheels, right? With a current based system, the motor gets X watts all the time. Speed is not considered. If the rider adds Y watts on the rear wheel, it adds to the bike's speed.

Change to a speed based system. The front motor will get enough power to hold a given speed. Add more power to the rear wheel, and the motor gets to slack off. The rider says, jeez, this assist is crappy. In real life, it's not this drastic. The motor feedback isn't that fast, so the rider does feel his effort being additive, but it's less responsive, You can solve this by setting the speed limit high. But then the assist seems overboosted.

 

[Jeremy McCreary]

That sounds just like a KT controller.
Kunteng calls it "Imitation Torque Control".

The difference with Ride1Up is that you only need 10° of crank rotation to turn the power back on, so they must have a 36 magnet PAS sensor.

Regular PAS sensors have a maximum of 12 magnets.

View attachment 174609


The KT controller allows for a 5, 6, 8, 10, or 12 magnet cadence sensor input.
12 magnet would be 30° of rotation.

View attachment 174610



So maybe Ride1Up has developed a cadence sensor that gives 36 pulses per revolution, or they are making stuff up.

Yes, the higher-resolution rotation detection is another selling point. But I'm wondering how tracking motor current leads to "desired exertion".

Suppose you encounter a bump in total mechanical resistance — hill, headwind, softer surface. At constant rider power, the bike and motor will slow at least transiently, the motor's back EMF will drop, and at constant input voltage, motor current will increase.

You could interpret the bump in motor current as a cue to add assist so as to keep exertion more or less constant. But you'd have to consult other sensors to know when it's safe to do so. Absent a torque sensor, what could those other sensors be? Would measuring true cadence (crank RPM) help?

 

[PCeBiker]

KT controllers are current based, and I think they feel better on pedal assist than other brands of controllers I owned.

I wouldn't know about that.
I didn't hook up my PAS sensor and I don't pedal.

This is my belief on how it works. Easier to visualize it with a front drive bike, Total speed depends on power to both wheels, right? With a current based system, the motor gets X watts all the time. Speed is not considered. If the rider adds Y watts on the rear wheel, it adds to the bike's speed.

That sounds right to me, but if you were coasting down hill, and kept pedaling (just enough to activate the PAS sensor), the power to the motor will shut off at a certain speed depending on your PAS mode setting.

 

[PCeBiker]

Yes, the higher-resolution rotation detection is another selling point. But I'm wondering how tracking motor current leads to "desired exertion".

You just exert yourself as much as you want.

If you need more help, you just turn up your PAS mode setting.

Suppose you encounter a bump in total mechanical resistance — hill, headwind, softer surface. At constant rider power, the bike and motor will slow at least transiently, the motor's back EMF will drop, and at constant input voltage, motor current will increase.

You could interpret the bump in motor current as a cue to add assist so as to keep exertion more or less constant. But you'd have to consult other sensors to know when it's safe to do so. Absent a torque sensor, what could those other sensors be? Would measuring true cadence (crank RPM) help?

You're over thinking it.
The Kunteng version just puts out power at a certain value, but you choose the value. Not just all or nothing.
If you're slowing down you can just pedal harder, change gears or turn up your PAS mode setting to add some more power, regardless of how hard or fast you're pedaling.

It's all kinda manually operated without all the sensors, data, and AI.

You just ride your ebike like a regular bicycle, and add as much power as you want.

 

[PCeBiker]

A cadence sensor that is not just off or on would be a big improvement.

It's not the cadence sensor turning on and off at between 5 and 36 times per revolution, it's what the controller does with that on/off signal.
Is it turning full power on and off, or is it turning a preselected amount of power on and off?

I have my first throttle gear turning 125 Watts on and off and I can have my fifth throttle gear turning 1365 Watts on and off (with a fully charged battery) if I want.
And I can adjust how the power curve ramps up depending on my current speed (if I hook up my PAS sensor).

I'm sure that Ride1Up is doing something similar.

It really does work nicely.
My first throttle gear tops out at 6 kph, but if I'm in gravel or soft grass I might only go 3 kph, or not at all if there's not enough power at 125 Watts to rotate the wheel.
I can pull over to the side of the road for a really tight radius 180° turn, and just hammer the throttle.
My e-bike just rolls ahead slowly at a walk speed until I'm turned around, then I hit the PAS button to speed up again.


Most cheap ebikes come with a bunch of PAS mode settings, but first PAS mode speed is something like 18 kph, and it gets full power assist to get you to that speed.
There is no slower PAS mode speeds and it's not adjustable.

 

[Gionnirocket]

Trying to understand the advanced "current-based" (CB) cadence-sensing assist now offered on some Ride1Up models. Follow the link for their description.

Sounds like CB assist doles out power as a function of motor current, up to a fraction of max motor power set by assist level. Ride1Up claims that motor current can be tied somehow to "desired exertion" without measuring rider torque.

Question: Anyone know of an exploitable connection between desired exertion and motor current?

Sounds like the cadence sensor on the Bafang BBS* mid drives I mentioned in the other post. Not sure why hub controllers seem to be struggling with this.

On the BBS you set each PAS to a certain cadence and desired current. The motor will help you maintain that cadence using up to the set current and taper the power down as you approach the cadence set point. Speed is not in the equation. Since you can somewhat set the acceleration curve... You gradually get assistance up to the cadence level and then it gradually decreases assistance. Minimizing the On/Off feel and eliminating ghost pedaling.

 

[PCeBiker]

Sounds like the cadence sensor on the Bafang BBS* mid drives I mentioned in the other post. Not sure why hub controllers seem to be struggling with this.

I've heard about that.
Apparently you can choose between speed based PAS and power based PAS.
Almost EVERYBODY chooses power based.

 

[PCeBiker]

,.. Not sure why hub controllers seem to be struggling with this.

It makes no sense to me.
Controllers are used to control power.
It's stupid simple to program a controller to tell it how much power to use.

 

[Tom@WashDC]

Sounds like marketing mumbo jumbo. Just go to torque sensing already. A torque sensing bottom bracket can be had for $60-$100 retail!. There is no cost difference for a torque sensing or cadence sensing controller or display. They are just delaying the inevitable.

 

[PCeBiker]

,.. Just go to torque sensing already.

Just go to throttle, and forget all the damn sensors.

 

[Jeremy McCreary]

OK guys, I'm taking Ride1Up's current-based cadence-sensing claim at face value (for now) and wondering how such an assist scheme might be implemented.

NOT asking about KT's system, which might or might not be based on the same concept.

 

[Tom@WashDC]

Just go to throttle, and forget all the damn sensors.

Actaully, torque sensing maps to the throttle circuitry in the controller. It basically adds throttle power in a consistent smooth fashion based on pedal pressure.

 

[PCeBiker]

I map directly to my throttle circuitry using my throttle.
I add power any way I want based on my throttle position.

Question: Anyone know of an exploitable connection between desired exertion and motor current?

I am the connection and choose to not be exploited.
I throttle and/or pedal as I desire.

Just go to torque sensing already. A torque sensing bottom bracket can be had for $60-$100 retail!.

I can't use a torque or cadence sensor because I don't want to pedal.

I am considering mounting a torque sensor to my handlebars.
I could squeeze it to apply power.
That'd be cool .

 

[Tom@WashDC]

I am the connection and choose to not be exploited.

PCeBiker, what do you mean "choose not to be exploited"?

 

[PCeBiker]

PCeBiker, what do you mean "choose not to be exploited"?

If I've got sensors monitoring me and my e-bike, and data being sent to from the cloud and mission control dictating my ride, I would feel exploited.

I like to feel like I'm the one in control with as little AI, software and computer chips as possible.

 

[Tom@WashDC]

If I've got sensors monitoring me and my e-bike, and data being sent to from the cloud and mission control dictating my ride, I would feel exploited.

I like to feel like I'm the one in control with as little AI, software and computer chips as possible.

Gotcha