Thumb has to adjust the power of the motor instead of use as accelerator?

Joined
Oct 17, 2023
Messages
8
Location
imqy8J2pXiXkZC8qGnh4
I have equipped a recumbent bike with front wheel motor and a KT controller.

Due to mount position it’s hard to reach the display to change the support level. Driving in the city, it would be great to adjust the level fast. Hence, I asked myself wether it’s possible to use the throttle gas as support level selector?
 
Not with the KT controller (or any other I am aware of)..

But you can buy button units with momentary switches on them that you can put on your handlebars where you can reach them more easily, open up the display and wire them in parallel with the existing buttons inside the display.

There are some displays that have those buttons as a separate unit that clamps to the bars closer to the grips while the display is in the center, and some of those are separate parts you can purchase. (the displays themselves wont' be compatible with your controller but if you buy just the button module....)

There are also motorcycle / etc button units that have momentary NO (normally open) switches, anywhere from one to several per bar clamp unit, that could be used.


Another option is to use the Cycle Analyst v3 from ebikes.ca to read your throttle and PAS sensor, and create a throttle signal to feed to your controller, so that you aren't limited to the way the KT and pretty much all other controllers use assist levels (where you always get the maximum assist of whatever that assist level is). Instead, the CA3 will actually let you control the motor current, power, or speed (your choice) directly from the pedals as if they were a throttle.

I use this to control the SB Cruiser heavy cargo trike via pedals alone, without having to have assist levels at all. I just ride normally and get assist proportional to the speed at which I'm pedalling. (there are a number of ways to set it up, I use the simplest).
 
Last edited:
Hi amberwolf,

Thanks for your elaborate answer. I really appreciate it!

I think I will go down the momentary button route of some kind. Although auf the wires to the small display I have (the smallest KT, can’t remember the number now) will be a PITA for sure!

But the cycle analyst also looks really interesting piece to play around! I really like, it also supports torque sensors. Just bad my bike does support normal bottom brakes but has a custom solution. Would need cranks or pedals, which I am not available with wire connections…

Would be great to build something similar with a little microcontroller like the ESP32, as I would like to find a way to remote control my hammerhead Karoo via ANT+. But this alone will be a time sink I am not sure I am willing to dive into! 😆

I guess I have to meditate over this decision a bit.
Thanks again!
Andreas
 
I think I will go down the momentary button route of some kind. Although auf the wires to the small display I have (the smallest KT, can’t remember the number now) will be a PITA for sure!
It might be; the hard part is usually opening the case, and modifying it to pass a new cable thru it with enough wires for the switches (probably six, two for each button***). Once you can see the buttons, you can usually access all the solder pads for them easily enough, so as long as you have the proper soldering station or iron for that type of work, and are practiced enough at using it, you shouldn't have too much trouble.

***if you can't fit six wires in, it's a little more complicated, as you then have to do some measuring to find the common pin for each switch, and use one wire for that, and then a separate wire for each switch's "signal" side. Then you have to wire the switch cluster you use on the bars for a common switch connection on one side, isntead of independent switches.



But the cycle analyst also looks really interesting piece to play around! I really like, it also supports torque sensors. Just bad my bike does support normal bottom brakes but has a custom solution. Would need cranks or pedals, which I am not available with wire connections…

Knowing the details of your build**** we can help suggest solutions.

Without that, we can only guess, such as:

If you're using hand-pumps, or treadles, etc, to run the "pedal chain" drivetrain, there will still be some point at which the linear motion is converted into circular motion. After that point, a "standard" torque sensor can be inserted into the drivetrain as a jackshaft (requires some custom work but isn't impossible in most cases).

If you're using a rotary PAS sensor, you can use that same point to install a torque sensor instead (with the appropriate mounting hardware).

There are also torque sensors that replace the drive-wheel's rightside dropout (or attach to it) to detect the chain pulling on the axle.

There are hubmotors like the TSM-A5 that have a torque sensor built into them.

****(for some reason, there are many people that come here for help but are so secretive about what they have that we can't really help them, since they refuse to tell us enough details about their bike/etc, system, situation, problem, etc. It's pretty frustrating since they often demand help but we can't give it to them without info they won't provide...some of them eventually tell us some of it but we have to literally drag every bit of info out of them one tiny piece at a time)


Would be great to build something similar with a little microcontroller like the ESP32, as I would like to find a way to remote control my hammerhead Karoo via ANT+. But this alone will be a time sink I am not sure I am willing to dive into!
That device comes up in a google search as a cycling computer / GPS unit, so it is not something the Cycle Analyst or any similar device could control or interface with.

What were you wanting to do with the HK via the CA?
 
Last edited:
Hi amberwolf,
thanks for your elaborate answer! I didn't want to be secretive, just had been short in my descriptions. Sorry for that.
It might be; the hard part is usually opening the case, and modifying it to pass a new cable thru it with enough wires for the switches (probably six, two for each button***). Once you can see the buttons, you can usually access all the solder pads for them easily enough, so as long as you have the proper soldering station or iron for that type of work, and are practiced enough at using it, you shouldn't have too much trouble.
It is this display and I doubt there is much space in the case. Maybe I can get away with thin enameled wire, as used in transformers, and just add a shrink tube outside of the case...
***if you can't fit six wires in, it's a little more complicated, as you then have to do some measuring to find the common pin for each switch, and use one wire for that, and then a separate wire for each switch's "signal" side. Then you have to wire the switch cluster you use on the bars for a common switch connection on one side, isntead of independent switches.

Knowing the details of your build**** we can help suggest solutions.
Just in case you are interested what I am doing, I have written a little bit of project documentation in a german forum... I am pretty sure google translate will give a somewhat good enough translation. :)

Without that, we can only guess, such as:

If you're using hand-pumps, or treadles, etc, to run the "pedal chain" drivetrain, there will still be some point at which the linear motion is converted into circular motion. After that point, a "standard" torque sensor can be inserted into the drivetrain as a jackshaft (requires some custom work but isn't impossible in most cases).

If you're using a rotary PAS sensor, you can use that same point to install a torque sensor instead (with the appropriate mounting hardware).

There are also torque sensors that replace the drive-wheel's rightside dropout (or attach to it) to detect the chain pulling on the axle.

There are hubmotors like the TSM-A5 that have a torque sensor built into them.

****(for some reason, there are many people that come here for help but are so secretive about what they have that we can't really help them, since they refuse to tell us enough details about their bike/etc, system, situation, problem, etc. It's pretty frustrating since they often demand help but we can't give it to them without info they won't provide...some of them eventually tell us some of it but we have to literally drag every bit of info out of them one tiny piece at a time)

That device comes up in a google search as a cycling computer / GPS unit, so it is not something the Cycle Analyst or any similar device could control or interface with.

What were you wanting to do with the HK via the CA?
I don't want to connect the HD with the CA. They are completely unrelated, but I would like to remote control HD and also the display.

It would be nice to remote control the display to change the power of the motor. The display is mounted on the left side of the handle bar just below the grip now. It's hard to see, as it's black...

I currently use a (now nearly ancient) Suunto Ambit 2 to record my rides and to monitor my heart rate. But I always forget to start or stop the watch. Also uploading the activity recordings to my computer is a bit annoying. In opposite the HK is always directly in front of me and I never forgot to start or stop an activity when I used it. Also it connects to my home WiFi and uploads the activity as soon as I stop the recording... Makes it really trouble free.

It's just impossible to reach the HK when it's mounted just behind the front light, right there where there is a little bike computer now.

IMG_5588.jpeg

As you can see, the chain is internally with this bike. Also the motor is a front wheel one, hence there is no way to detect the pedaling torque there...
And for the PAS, I have a custom solution due to the frame.

Hope this clarifies everything a bit. :)
 
If you attach an MTB bar end to one of your handlebar stubs, you could mount your display there (if the cable length allows it).

I once made a handlebar extension for a customer who got a motor kit for her Terra Trike. I made a double ended expanding plug to graft on a short length of cut-off handlebar.

IMG_20191007_194749.jpg
 
Last edited:
If you attach an MTB bar end to one of your handlebar stubs, you could mount your display there (if the cable length allows it).
I did something similar: 3D printed a very short round thing to be mounted with a zip tie, but I need something better, as it is really flimsy. :D

A MTB bar end is a great idea. I think I even have lying around somewhere in the basement... Just use a saw to shorten them a bit, not to get caught at an obstacle.
 
I did something similar: 3D printed a very short round thing to be mounted with a zip tie, but I need something better, as it is really flimsy. :D


LIke this?

1702525636147.png
 
It is this display and I doubt there is much space in the case. Maybe I can get away with thin enameled wire, as used in transformers, and just add a shrink tube outside of the case...
If it's the one pictured below, there's probably quite a bit of space around the buttons, though you have to make sure the wire doesn't interfere with the buttons themselves if you still want the built-in ones to function easily. If you have any unused USB cables (or devices with them built in) laying around, you can probably use the very thin wires these use to fit right in there. There will be at least five wires in any good USB cable--red 5v, black ground, green / white data lines, and a probably uninsulated shield wire. Some cables have more wires in them (though they may be unconnected at one end or the other). Apple Lightning cables are similar.

If you don't ride in the wet (or can waterproof the connection) you could even install a port on the case of the display so you can unplug the button module for whatever reason.


1702525692102.png



Just in case you are interested what I am doing, I have written a little bit of project documentation in a german forum... I am pretty sure google translate will give a somewhat good enough translation. :)
Thanks! I'll go read up on the project, it might help me give you better suggestions. :) (or at least ask better questions).

I don't want to connect the HD with the CA. They are completely unrelated, but I would like to remote control HD and also the display.
Ok, I guess there is some missing info from your original statement I replied to about that, since that statement implied you wanted to build something like the CA that would let you remote control the Karoo (HK)...but the CA doesn't remote control anything. ;)

For the remote contro
It is this display and I doubt there is much space in the case. Maybe I can get away with thin enameled wire, as used in transformers, and just add a shrink tube outside of the case...

If it's the one pictured below, there's probably quite a bit of space around the buttons, though you have to make sure the wire doesn't interfere with the buttons themselves if you still want the built-in ones to function easily. If you have any unused USB cables (or devices with them built in) laying around, you can probably use the very thin wires these use to fit right in there. There will be at least five wires in any good USB cable--red 5v, black ground, green / white data lines, and a probably uninsulated shield wire. Some cables have more wires in them (though they may be unconnected at one end or the other). Apple Lightning cables are similar.

If you don't ride in the wet (or can waterproof the connection) you could even install a port on the case of the display so you can unplug the button module for whatever reason.


1702525692102.png





Just in case you are interested what I am doing, I have written a little bit of project documentation in a german forum... I am pretty sure google translate will give a somewhat good enough translation. :)
Thanks! I'll go read up on the project, it might help me give you better suggestions. :) (or at least ask better questions).

I don't want to connect the HD with the CA. They are completely unrelated, but I would like to remote control HD and also the display.
Ok, I guess there is some missing info from your original statement I replied to about that, since that statement implied you wanted to build something like the CA that would let you remote control the Karoo (HK)...but the CA doesn't remote control anything. ;)



As you can see, the chain is internally with this bike. Also the motor is a front wheel one, hence there is no way to detect the pedaling torque there...
And for the PAS, I have a custom solution due to the frame.

Well, the CA does a great job even with just a cadence sensor like you installed inside the housing,
1702526602662.png 1702526630224.png 1702526641535.png
of variable control of the system via cadence. (like I do on my SB Cruiser)

If you have situations that require the motor to run before you can pedal (like getting started from a stop while pointed up a steep hill), a torque sensor would be the "best" solution to that. But the *easiest* solution is a "go button" you can press for a specific amount of throttle (or the actual throttle) just to get started and then use the PAS by pedalling normally.


The only "easy" way I can think of ATM to use a torque sensor on that type of system is one that rides on the chain, like the no-longer-available BEAM TS did, so it detects tension on the top of the chainline and outputs a proportional voltage to that. From what I can see of the system you would have to cut a hole on the top of the chain housing for the guide wheel of the sensor arm to ride on the chain, and presumably print a housing to cover that unit. It can be anywhere on the chainline, wherever is out of the way...but I don't presently see another option for this.


Well, there *are* sensors for bike training that go on (or replace) your pedals / crank arms, and wirelessly transmit their data to their system's main unit...so you could either build a clone of such a sensor or buy one and then either replace the transmitter with one of your own or figure out their protocol and data format to read it with your own system, that then creates a throttle signal (or creates a 0-5v signal to feed to a Cycle Analyst v3 to let *that* create such a signal).




It would be nice to remote control the display to change the power of the motor. The display is mounted on the left side of the handle bar just below the grip now. It's hard to see, as it's black...

I currently use a (now nearly ancient) Suunto Ambit 2 to record my rides and to monitor my heart rate. But I always forget to start or stop the watch. Also uploading the activity recordings to my computer is a bit annoying. In opposite the HK is always directly in front of me and I never forgot to start or stop an activity when I used it. Also it connects to my home WiFi and uploads the activity as soon as I stop the recording... Makes it really trouble free.

It's just impossible to reach the HK when it's mounted just behind the front light, right there where there is a little bike computer now.

FWIW, a simple mechanical solution to pressing buttons (that works on touch screens or on mechanical switches) is to use an air button. It's a button (that looks and feels like the buttons on an arcade game machine console) that has seals in it to push some air thru a tube to push on some remote thing; they're commonly used in aquatics systems since wires corrode but these don't. ;) Garbage disposal switches mounted on sinks also use a variant of them.

Often they're used by routing the air tube out to a "normal" electrical switch that's operated by this tiny amount of air pressure, but they can also be used to push a "reversed" mechanism that pushes a rod out a tiny bit with the air's force, to press another button on something else.

The nifty thing is you could design and print both of those parts instead of trying to find ones that fit your specific application, and just use regular rubber O-ring seals on the moving parts to keep the air between them / in the tubing, and use stiff-walled tubing (1/4" diameter is easy to find and is usually used) between them.

Then you mount the button on your bars and the remote pusher on the face of the device, with a touch-screen stylus's tip on the end of it to touch the button on the screen of the device. The primary disadvantage to this is it covers that portion of hte screen, but a properly designed one would only cover the amount of screen necessary to hold the stylus tip over the button.



Hope this clarifies everything a bit. :)
It does, and your thread over there will probably clarify more. :)
 
LIke this?

View attachment 344407
Yes, very close to this one. I put it online on printables, if somebody needs something similar.

And thanks for the other info! I don't have time to go over it now, but the torque sensor looks interesting.
Something one could build himself maybe, with a hall sensor and an arm to be bent...

Have to take a closer look. :)
Thanks
 
Yes, very close to this one. I put it online on printables, if somebody needs something similar.
Thanks.

And thanks for the other info! I don't have time to go over it now, but the torque sensor looks interesting.
Something one could build himself maybe, with a hall sensor and an arm to be bent...
Or a load cell from a scale (which you can salvage from one, or buy the cell itself new separately, along with the instrumentation amplifier/etc to detect and scale it).
 
Back
Top