My experience with torque sensors like T9 or T17

[B_c]

I'd like to highlight that my experience suggests that torque sensors work quite well once certain issues are dealt with.

1: The initial issue doesn't stem from the torque sensor; instead, it originates from the motor to the controller, which can cause problems with dirty signals reaching the torque sensor and leading to its reset.

A possible solution is to ground the battery zero to the frame, or if possible, shorten the motor cable so that the controller can be mounted closer to the motor.

For example, today I was riding my bike home and felt as it was the PAS mode with maximum assistance. So, I didn't have any feeling that the torque sensor was in operation. I stopped and noticed that the grounded wire had disconnected from the frame, and when I secured it back to the frame, everything worked perfectly.

In the picture below, you can see one of my bicycles with a detail of the torque sensor and hub motor.



2: The second issue involves a component, as shown in the picture below, that I cannot screw tightly because, if I do so, the torque sensor's shaft does not rotate smoothly and encounters slight resistance.



The solution is to place several flat washers, as shown in the picture, so that I can tighten it as it should be.





3:The third issue is moisture, especially in cases of heavy rain or snow when riding the bicycle.

Moisture somehow enters the torque sensor through the path where the cables go. I have never experienced this issue during sunny weather without humidity and rain-snow.

The solution is to protect it with a bit of silicone from the inside within the torque sensor casing. Additionally, on the outside where the cable enters the torque sensor, I applied a light coating of water-repellent spray.



There is a groove in the housing which I fill with silicone.



On the picture below is a damaged bearing that I will replace with a new one. It is also damaged by moisture.
It was heavily rusted when I removed it.



I hope that what I wrote above can help someone, and if anyone has any tips, feel free to share them.

 

[FuzzyWuzzy]

Thank you. Can you post more images of the frame grounding and the wiring setup

 

[slaphappygamer]

Much appreciated. I'm looking at getting this type of sensor. What controller are you using this with?

 

[stancecoke]

problems with dirty signals reaching the torque sensor

Do you use the torquesensor with a CycleAnalyst?! I read a lot about this grounding issue, but I never had any problem with the sensor directly attached to a Lishui controller, maybe Lishui has a better input filtering on their PCB?!

regards
stancecoke

 

[B_c]

Grounding is quite simple and sufficient with a single thin wire from the battery's negative pole to the seat.

 

[B_c]

I can respond to others that I use my own controller with the VESC driver. My son used to be active here on the forum with the username "buba," but I think he doesn't have time for it anymore. However, he helps me achieve what I want, so for two years now, I've been using my own, much more modern controller. But I can talk about that another time when the time comes.

 

[B_c]

Do you use the torquesensor with a CycleAnalyst?! I read a lot about this grounding issue, but I never had any problem with the sensor directly attached to a Lishui controller, maybe Lishui has a better input filtering on their PCB?!

regards
stancecoke

I have only used this controller, and I was not satisfied because sometimes it works well, and sometimes it doesn't work well or not at all. Does it go by the name Lishui or something else? I don't know for sure.

 

[stancecoke]

I have only used this controller,

Yes this is a 12FET Lishui controller. But I think the original firmware is not satisfying. I use my own firmware

Regards
stancecoke

 

[B_c]

I will show you the old controller that I have been using for two years. I made a picture of it to easily connect the cables and avoid short circuits.

I no longer use it because I have a new controller, and I can access it through the website by entering the IP address. I think it's great that I can make changes using my mobile phone without having to connect to a computer to modify the settings.

I will present the new controller sometime next year when everything is ready the way I want it.

 

[B_c]

Yes this is a 12FET Lishui controller. But I think the original firmware is not satisfying. I use my own firmware

Regards
stancecoke

I have their, as they call it, 500-watt controller, and it didn't work at all; one overheated significantly, while another worked perfectly for several days.

And that is the reason why I wanted something else where I can connect all the lights with different voltages and some other things.

 

[B_c]

Here's a picture of me building a new e-bicycle that will get a T17 torque sensor. And, of course, Rockshox.

 

[B_c]

And let me add this, on the new controller, I can connect all KT displays, all Focan displays, and probably all Bafang displays without opening them up and programming them. It will be good news next year.

Alternatively, if someone doesn't want a display at all, they will be able to see all information.

 

[stancecoke]

while another worked perfectly for several days.

I have just one of this new generation 12FET, it works perfectly for me. I use it for several developments, e.g. the Kclamber cassette integrated torque sensor. With the EBiCS firmware there is no additional PCB needed and Kunteng, Kingmeter and Bafang display protocols are implemented
For most of my projects, a small Lishui 6FET is sufficient.

regards
stancecoke

 

[B_c]

I have just one of this new generation 12FET, it works perfectly for me. I use it for several developments, e.g. the Kclamber cassette integrated torque sensor. With the EBiCS firmware there is no additional PCB needed and Kunteng, Kingmeter and Bafang display protocols are implemented
For most of my projects, a small Lishui 6FET is sufficient.

regards
stancecoke

Just to ask you, when you say "it works perfectly for me" do you mean with the original firmware or with your firmware?

 

[stancecoke]

With the custom EBiCS firmware, of course.

regards
stancecoke

 

[B_c]

With the custom EBiCS firmware, of course.

regards
stancecoke

Thank you for the answers.

And, let me ask you this: how far is your cable from the controller? I noticed that when I have a cable that's 120 cm long, I need to ground the controller, but on bicycles, I had a cable of 60 cm, and they could work without grounding.

 

[B_c]

I must add that during grounding, it's always advisable to include a fuse in case of a short circuit. If one chooses not to do that, a smaller wire should be selected so that in the event of a short circuit, the wire would likely burn out and prevent further mishaps.

 

[stancecoke]

how far is your cable from the controller?

I do not cut the original cable of the torquesensor normally, so I think it will be about 1,2m.
I think Lishui uses very elaborate interference suppression for its inputs, with ferrite balls, RC elements, etc.

regards
stancecoke

 

[B_c]

I do not cut the original cable of the torquesensor normally, so I think it will be about 1,2m.
I think Lishui uses very elaborate interference suppression for its inputs, with ferrite balls, RC elements, etc.

regards
stancecoke

I wasn't clear. I mean the motor cable in your case, from the motor to the controller-driver, how many centimeters is it approximately?

 

[stancecoke]

from the motor to the controller

Depends on the frame geometry and the position of the controller, of course. But less than 100 cm in most cases. I tried shielded motor wires when I used my self built strain gauge amplifier, but I saw no effect.

regards
stancecoke

 

[B_c]

I will try your firmware when I have more time for it. Otherwise, I have three of their controllers that I cannot use without errors.
Thanks stancecoke!

 

[B_c]

I have to tell you that riding on snow and ice with both PAS and torque sensors can be very dangerous. The torque sensor performs much better in such weather conditions, although I had a bad fall even with a bike equipped with a torque sensor.

I told my son to program something simple for me to prevent this from happening. He programmed something simple with acceleration, and he calls it "traction control" similar to what cars have. I can tell you that it's a good feature.

And yes, it's much more dangerous with the motor on the front wheel.

 

[B_c]

I would recommend choosing the T17 if one has to decide between these two sensors, even though it might be slightly more expensive.

The T17 comes with one more signal, allowing the controller to detect even when pedaling backward.

Imagine a function that can be programmed into the controller, for instance, determining the speed at which the pedals are turned backward, and correlating that with the amount of electricity regenerated back into the battery.

Another example could be setting a condition that, when you pedal backward a certain number of times, the bicycle becomes legal at 25 km/h, as in my case and so on.

 

[amberwolf]

I use it for several developments, e.g. the Kclamber cassette integrated torque sensor.

Could I trouble you to post your thoughts on those sensors over here in the following thread?

https://endless-sphere.com/sphere/threads/rear-all-axle-v3-with-torque-sensor.122166/