Translating drone motor specs for use in robots?

I want to use a small out-run “drone motor” in a robot application. Let’s say it is a knee joint in a leg. So much of the time the robot is just standing and the RPM is zero. It is just holding in torque/position mode. I’ll have a 9:1 reduction system made with timing belts.

Here is the question: These drone motors are not spec’d for this. They typically give Kv and max current. Sometimes they give thrust with a certain size prop. What I want to know is the torque vs current graph. I could buy motors and measure them but there must be a way to estimate what I want from what they specify.

As an example, I see a 40x15mm motor rated 40 amps and 350KV. What is the zero-RPM torque v. current function? Can it be worked out?

Torque is approximately current x 8.3 / kv
Things in Motion: How to estimate the torque of a BLDC (PMSM) electric motor using only its Kv and current draw

Generally I assume that the rated current is about twice what can be used continuously. So for your motor, 40x8.3/350=0.95Nm for short bursts (e.g. jumping), and half that continuously.

Current ratings are not well standardized, and even if they were it’s hard to say exactly what a motor can handle. Ultimately you’re trying to avoid burning the windings or overheating the magnets. Windings can burn quickly, but it takes time to heat the whole motor to the point that the magnets are damaged. Airflow makes a huge difference, and if it’s mounted to aluminum structure then that can act as a heatsink.

The true peak torque of a motor comes when the stator reaches magnetic saturation, and I’m not sure exactly how close that is when running at the full current rating. Sometime I want to try building a motor with a more slender stator to leave room for more copper, so for example you could run continuously at 2/3 saturation (less burst torque, but more continuous for the same weight). And make a rotor using mu-metal foil and carbon fiber instead of an iron ring. Usually the iron is about half the rotor weight and 20-25% the total motor weight, and since it’s all located at maximum radius you could potentially reduce the inertia by 1/3 or more. Very good for robot actuators that need to reverse direction quickly. Then you could use higher reduction, reducing the overall motor size too.

EDIT: Dang it, no cheating physics after all :stuck_out_tongue: Upon further research, I’m fairly sure mu-metal won’t be any help in this case due to magnetic saturation. High permeability only means that it will reach full magnetization with a lower applied field, not that it can conduct more field from one rotor magnet to the next. And special high saturation alloys or halbach array would only be marginal improvements, not really worth doing. In a way I guess that’s kind of nice that cheap mass production motors are already about as good as it gets :slight_smile:

3 Likes

If the mass of the magnets are increased, it will have better magnetic properties, compared to cheap mass produced motors, explicit for the hallbach_array. Also worth noting is the various magnet qualities/compositions.