New Torque Tester

I’ve built a much more robust setup for testing cube torque, but now I’m running into some new problems.

The IQ control motor I’ve been using the turn a cube layer seems to be a bit under powered for some of the cubes I’m testing, and especially when I try to turn cubes slowly it stalls. I’ve also noticed that even when it is completely free running at slow speeds it isn’t totally smooth:

I’m concerned that this will produce artifacts in the data because the uneven turning by the motor will be transmitted to the force sensor.

So I’m going to modify the setup for a much more powerful (and more expensive!) ClearPath servo motor. (https://www.teknic.com/products/clearpath-brushless-dc-servo-motors/) Luckily I have one of these lying around from another project. This motor should be powerful enough to tear apart any cube I put in the machine.

New Torque Tester

I’ve redesigned the cube torque tester with components that should be much more accurate and should be able to provide better repeatability between measurements.

I’ve planning on the following improvements:

• The cube will be held with 3D printed parts, which I’ve ordered from Shapeways so the tolerances should be pretty good. (±0.15 mm + 0.15% of the longest axis.)

• The rotation axle that supports the cube will use ball bearings to reduce the torque imparted on the cube from the holder as much as possible. Any friction from the bearing effectively reduces the torque that is transmitted to the load cell so it’s important to make this bearing as friction-free as possible.

• Everything will be mounted to a ¼” (6mm) thick aluminum plate which should keep things more stable and repeatable.

• I’ll be able to precisely control the stop to prevent the arm from counter-rotating, and I should be able to use this stop to slightly pre-load the load cell to make sure there is no slop in the system at all.

• An extruded aluminum post will allow me to raise and lower the motor mount. This should be much more rigid than my cardboard setup and hopefully also allow me to position the motor in the same place every time. It will be interesting to see how close of a match I can get between two measurements of the same cube after removing and reinstalling the motor mount.

• I’ve elongated the beam that pushes on the load cell. I suspect that I was overloading the load cell previously and using the load cell above it’s rated range is probably bad for linearity which will distort my measurements.

Motorized Torque Tester

I mocked-up a rig that uses a small computer controlled motor to turn the cube for the purposes of getting more consistent results:

My cardboard setup has some problems with the alignment of axes of rotation, which is why, when the motor turns the cube, you can see everything flex.

I’m also having issues with getting consistent contact between the cardboard holder and the load cell since the load cell only prevents the cube from rotating in one direction, but the motor turns the cube both ways. That’s why you can see a screw preventing the cube holder from turning the other way in the video.

It’s time to get more serious about the construction of this whole setup. I’m going to start designing a more precise and stable apparatus.

A quick note about the motor I’m using: I bought this IQ control motor on crowd supply last year and it just arrived recently. I had left it in its box until I started building this cube testing setup, but I got it working very quickly and it seems to be doing the job very nicely. I suspect that this will be a better motor to use than a stepper motor that might introduce vibrations or other weirdness.

Motor: http://www.iq-control.com/iq-position-module

Motor controller: https://www.sparkfun.com/products/11061