Looking for Torque and Angle Sensor

[horitbros]

Hey guys,

Our company builds a special kind of physical rehab equipment. It is all based on pendulums and fulcrums and has no electronics whatsoever in it. We now want to add some smarts to this line of equipment and i think that by installing a torque and angle sensor into the axis of the pendulum we will be able to take some useful measurements. the problem is im struggling to find a cost effective (under $500 USD) solution which combines a torque and angle sensor. The maximum torque is say 1000Nm and the maximum angle is under 160deg. Ideally this combined sensor would be smaller then a standard brick. Also we would love this sensor to be able to send wireless data out into the processing unit for calculations. Is this something that exists and if so could you please point me to the right direction.

Thanks a lot for the time and help!

A.

 

[3DDave]

Depending on the geometry you may only need angle measurement; unless you are trying to capture very fast, inertial measurements it's possible you can have a microcontroller perform the trigonometry to convert the angular motion of the pendulum to a torque and then compare that to the angular movement at the input to decide how to multiply the pendulum supplied torque.

The latter is only required if there is some reconfiguration that changes the leverage between the pendulum and the input.

 

[IRstuff]

Since it's a known pendulum, the torque is proportional to the angular acceleration, isn't it? If so, then you just need high rate angle measurements, which are much easier to do; a high speed encoder could do that, or even a gyro.

TTFN
I can do absolutely anything. I'm an expert!
homework forum: //

http://www.engineering.com/AskForum/aff/32.aspx

faq731-376 forum1529

 

[horitbros]

Thanks guys,
Maybe i didn't explain it clear enough. The machine has many settings changing the distance between the axis and the point where the force is applied downwards. Here is a picture. At the point labeled axis, I want to measure torque and angle of rotation of the axis.
Also the force changes constantly.

imgur.com/22MFvv7

Thanks

 

[3DDave]

The axis you are pointing at sees no torque, so there is no torque to measure.

Do you have a complete free-body diagram of the entire mechanism?

 

[horitbros]

Hmmm.. maybe the diagram is not clear...

Here's a better one

http://imgur.com/95n1HPN

thanks

 

[IRstuff]

So, even still, even a simple 3-axis accelerometer coupled with some means of determine the change in pendulum length would allow something like an Arduino to measure the angle change and calculate the angle rate and effective torque

TTFN
I can do absolutely anything. I'm an expert!
homework forum: //

http://www.engineering.com/AskForum/aff/32.aspx

faq731-376 forum1529

 

[horitbros]

Thanks IRStuff This is an interesting idea.

Do you know of any cost-effective 3-axis accelerometers that could enough precision and have a wireless connectivity?

Many thanks

 

[3DDave]

Look at a Chronos Watch from TI.

http://www.ti.com/tool/ez430-chronos

 

[horitbros]

Thanks a lot

 

[IRstuff]

Depends on your budget. A 3-axis accel/gyro for $12.95:

https://www.sparkfun.com/products/10937

You can use the accelerometers for an absolute angular orientation and the gyros to determine the angular rates/accelerations. Arduino ATMega328 board:

https://www.sparkfun.com/products/12757

Someone will need to do the programming. You probably also want some sort of keypad input plus a display of some sort

TTFN
I can do absolutely anything. I'm an expert!
homework forum: //

http://www.engineering.com/AskForum/aff/32.aspx

faq731-376 forum1529

 

[3DDave]

The watch has built-in 3 axis accelerometer capability and radio communications, as well as temp and barometric sensors and a receiver and programmer. It comes with demo software that links the watch to a computer and to heart rate monitors. Which is at least all the software and hardware parts one needs for this job. It's as off-the-self as one is going to get and is packaged by TI for getting embedded developers up and running as quickly as possible. It also has it's own built-in display and even comes with an spplication to control PowerPoint.

The same sort of thing can be made with Arduino, but the CC430 series is very low power where one might expect months of use on a C2032 coin cell. The Sparkfun item above is the IMU chip alone; to use it requires a circuit board and some decoupling caps, so the $40

https://www.sparkfun.com/products/11028

item will likely be the required one.

The original idea you had - mostly that is test equipment territory, so figure a few grand. Here's a nice torque cell

https://www.transducertechniques.com/rst-torque-sensor.aspx

or the cheaper

https://www.transducertechniques.com/sws-torque-sensor.aspx

Check the related equipment tab, which isn't wireless. Still no angle sensor or display or recording/transmitting ability.

I think the lack of a free body diagram, any sense of the requisite accuracy or suggestion of budget is the largest problem.

 

[IRstuff]

There are other options like:

https://www.sparkfun.com/products/13339

which can be plugged directly into the Arduino board and is only $19.95.

TTFN
I can do absolutely anything. I'm an expert!
homework forum: //

http://www.engineering.com/AskForum/aff/32.aspx

faq731-376 forum1529

 

[horitbros]

Yep, looks great, thanks. Arduino or even raspberry pi or similar since want to collect the data from the sensors, send to our server in the cloud, correlate with some settings and adjustments to the program and push back to the machine which will have a touch screen display