I’m new here. I have an EE degree from over a decade ago but have forgotten most things about motors and hoping to refresh myself via a little project!
I’m looking for a motor which I can control to act as a torsional spring. That is, if external forces move the shaft by, say, 5deg from 0 position, the system will respond with an opposing torque of, say .1Nm, if the shaft is moved an additional 5deg the opposing torque increases to .2Nm, etc. The system should have this symettrical behavior both clockwise and counter clockwise from the 0 angle.
I don’t think I need a motor with super precise positioning control but I need to be able to vary the torque precisely based on a precise reading of the present angle of the shaft (from an encoder I assume). Maybe that’s the same thing as precise positioning control lol, but my control system will not be seeking to control position if that makes sense.
The motor will need to be ok with having its shaft move in the opposite direction to the direction of torque it is applying, with about a 50% duty cycle, as shaft speed&direction will largely be determined by external forces. As I recall that’s exactly how you burn out a motor
Is this feasible? What types of challenges can I anticipate?
If it is feasible, what type of motor is best suited to this application in both a small scale proof of concept setting as well as at larger scales?
If I’m looking for something with, say, a max torque of about 2.5N*m to start playing with as a proof of concept, what would be a good product to purchase?
jremington:
Motors burn out when used in stalled conditions.
A galvanometer is what you need, but one with 2.5Nm torque will probably be quite expensive.
To be clear it's not going to be stalled 100% of the time in practice. If you imagine a mass oscillating back and forth on this torsion spring the stall points are just the moments at the highest shaft angle before the shaft starts turning back towards the 0 angle.
Are there motors out there designed with ventilation in order to allow some stalled operation?
It's built into each force feedback joystick, and it stalls 100%. Just use a Motor that's big enough to handle your desired force, as force will be transformed to heat at a rate of 100%, use a PWM for controlling the force and you are good.
davegravy:
To be clear it's not going to be stalled 100% of the time in practice. If you imagine a mass oscillating back and forth on this torsion spring the stall points are just the moments at the highest shaft angle before the shaft starts turning back towards the 0 angle.
Are there motors out there designed with ventilation in order to allow some stalled operation?
The heat problem is in the copper wire insulation inside the windings.
Torque is proportional to current, so what you need is position control with current limiting. Set target position to zero, and in your update loop set current limit in proportion to the position. This will also prevent the motor from overheating, as long as you make it so the current limit maxes out at some reasonable value if the motor is pushed really far from its zero position.
Something like an 8318 outrunner controlled by ODrive, and an Arduino to update the ODrive current limit should do the trick. You will indeed need an encoder for the motor position.
Torque is approximately current * 8.3 / kv, so if you have a 100kv motor and you want to produce 0.5Nm of restoring force, set current limit to 0.5 * 100 / 8.3 = 6 Amps.
