Torque Sensor

Hello guys. I have a rotating shaft, there are two strain gauges connected to that shaft, my goal is to build a torque sensor to get the instantaneous torque values. My idea was to build a wheatstone bridge with the two strain gauges, then connecting the bridge to nano-shield load cell, then connecting that to arduino ATMEGA 1284P, then by coding I can get the values through LabVIEW.
What do you think of that?

It’s feasable. What you describe is how many industrial instrumented torque shafts are constructed.

Have you solved how you will run any signals between measurement points on the rotating member and the roboRIO (which is presumably not rotating)?

This, along with the specifics of the strain gauges, instrumented material, and front-end electronics will determine the accuracy of the device, and therefore its usefulness.

Be forewarned, there are few high-accuracy, robust, inexpensive torque sensors available. If you have success, you very well may have a new market of people who want to purchase your setup. Then again, considering one has not yet been created, you will likely have some challenges ahead of you.

It should also be noted - for most FRC sources of actuation power (motors, solenoids, pneumatic pistons), the force and torque output is usually quite easy to calculate from values that are easily measurable (voltage, current, pressure).


Can anyone help me with building a wheatstone bridge with two strain gauges! , 120 ohms each, on this solderless breadboard? like where can i put the strain gauges wires and the resistors and the output wires.

probably do you have a schematic of what the circuit is

1 Like

Good luck with this! Industrial pressure / torque / force sensors are notoriously hard to make and to map a linear response. Really good ones are pretty expensive and you do not see them in FRC often. I think Rev makes a pressure sensor, not sure how good it is.

1 Like

The starting question is: Do you know and understands what a wheatstone bridge is? and why does it work?

Are you looking to get the values directly from LabView from the Arduino? or you looking to go from Arduino to RoboRIO?

so what I am understanding is that you will need two different wheatstone bridges and the two strain gauges that are each 120 ohms correct? Once I know this I can figure out a circuit that would work. Also for building it what resistors do you have access to or do you not care about what you might have to buy. also as far as I can see you will need two different wheatstone bridges. Other than that you could probably take out the arduino all together and plug it into the analog input of the roborio

If you use an electric motor torque is pretty linear to current. So you can use a talon SRX as a torque sensor - we have done it in the past if we wanted to grip something but not too hard for short periods of time don’t remember the exact code but we polled the Current from the talon SRX and if it was below the limit you slightly increase and if above you decrease. We also used it on occasions to “bounce of walls” in autonomous mode drive until the current rises above then stop and do something like shoot or drop or go left etc

Just as a question how would gearboxes factor into the linearness of the torque to current

I recommend this method. If you build a spinning or stationary torque transducer you are still going to have to calibrate it, and without lab grade equipment you aren’t going to get much better accuracy than measuring voltage and current, calibrated of course.

To account for gearboxes and everything I recommend you look into a Prony Brake and use that to calibrate your torque measurement by way of voltage and current.

I do support you building your own load cell and circuitry to evaluate them, that is a great learning experience, but I don’t think it will be as accurate as you are hoping for if you try to measure torque.

Even if you assume that the torque to current relationship is linear, there are issues measuring how much torque is actually applied to a load, based on the forces external to the motor. This can be demonstrated mathematically as follows:

\tau_{motor} - \tau_{app} - \tau_{fric} = I_{motor} \alpha
\tau_{app} = \tau_{motor} - \tau_{fric} - I_{motor} \alpha

Gearboxes usually have a %efficiency. And otherwise torque is inverse to speed in gearing. It depends all as to how accurate do you have to be with your torque measurement. No matter what you do in engineering there is always be a range of accuracy. So it depends on the circumstances. In the cases we used it the Amps as an indicator were good enough as it was mostly employed to hold something or push against something and the tolerances in those cases were big enough. In years past I used torque motors (outside of FRC) to build wire winding machines that wound gold wires that needed constant torque and adjust the speed accordingly drawing down the wire from one diameter to another and that was quite accurate enough - back then (early 80s) using op amps to do the “thinking” and measuring

1 Like

How can I build a wheatstone quarter bridge with one strain gauge. What are the connections needed and how it can be calibrated ?
Thank you.

This is your 4th thread about the same topic. Please keep everything in the same thread to avoid clutter and so everyone gets the whole story.

Is it possible to use HX711 module with strain gague to measure the torque applied to a rotating shaft?

Sure, no reason you can’t do that.

1 Like

So there is no need for a wheatstone bridge circuit?

If you look on the HX711 datasheet, it looks like it does require a wheatstone bridge input, it just handles the amplification and the digification.

This appears to be a university project. Multiple threads about this topic combined.