I’m designing a task where I would like to apply periods of random noisy (smoothed white noise) mechanical perturbations to subjects, and I’m wondering what range of perturbation frequencies can be safely used without damaging the KINARM, i.e., how smooth do the torque profiles have to be?
To my knowledge, this is not a question we have a specific answer to. In general, the motors should not be capable of producing enough force to break anything as long as your hand is controlling the end point (Kinarm EP) or your arm is in the troughs (Kinarm Exoskeleton).
Having said that, the robot type, force magnitude and duration will determine if any issues may arise. These factors and the time between the forces will determine if you will run into issues with torque limiting. The motors can produce ~4.7 Nm forever (the number varies depending on the model), but at higher forces you will reach limits where the torques will be clipped to avoid motor overheating.
Which robot type are you using? What is the peak load you’d like to apply and for how long?
Thanks for your answer. This is for the human exo. The loads I would use would be quite small (<1Nm per joint), but change at high frequency (0-100Hz), producing noisy vibrations.
Hi Jonathan, our engineer has passed on these comments:
The motors are definitely able to produce torques in this frequency range. For example, we apply sine wave signals from 1-250 Hz for the open loop testing. The magnitude of the sine waves is in the 1-2 Nm range, so I think that should be no problem. I don’t know what the ramifications of doing this for extended periods of time might be. We apply them for 1-3 seconds at each frequency.
Practically speaking, once the frequency gets up to above ~10 Hz, it really is just a vibration. As it gets higher, it rattles the other components in the robot a fair bit and can be very loud (audible noise). In the region of ~60-90Hz, there are some resonances where it is particularly loud.
Thanks Koloman! This all makes sense and is working as intended.