I'm looking to use 3-phase motor without gearbox for robotics application. It will always work at near stall situation never rotating more than 360deg. I'm looking for arrangement that will produce max stall torque while drawing min current.

I tried both gimbal (high resistance) and BLDC (low resistance) motors and yes I could achieve high torque, but only at cost of too high current and overheating.

My question is both practical and theoretical: what motor or mode of operation should I choose to maximize stall torque? and what in principle can be done? Imagine I had my motor factory - how would I approach building stall torque motor?


One relevant equation is the lorentz force equation. which suggests that if you want to reduce current then you need to increase the magnetic field.

The magnetic circuit is complex, with different materials moving relative to each other. As the rotor of the motor rotates, the magnetic field within the stator changes direction and there are losses. Choosing optimum materials and geometry is the subject of many books and scientific papers. In most motor design, rotation is expected and materials and geometry that is optimum for a rotating magnetic field are used. If you have a mostly non-rotating application, you may be able to select different materials and geometry that is better.

If you don't want to design your own motors, the type of industrial motor that is most likely to be a bit optimized towards your requirements is called a direct drive motor.

The gimbals motor you linked claims to have 6Kg/cm of torque. Assuming they really mean 6 Kg*cm*9.81*m/s^2, that's about 0.6 Newton meters (Nm) of torque. A good motor that would fit the same volume could be the Kollmorgen TBM(S)-7615-X, which has about 3 Nm of stall torque, but also approximately 3 times the mass. That's a rough idea of what good design and optimization can do. About double the torque per kilogram of motor.

Considering your ~$100 per motor budget, it depends on how motivated you are. If you are really motivated and like to build things, you could learn a bit of how motors are made and make your own motors (which are just some iron sheet cut out (by waterjet for instance), glued in a stack, wound with copper wire, and magnets glued to a rotor) for roughly your budget, assuming your robot has 3 or 6 motors.

However you may want to keep in mind that there are very few direct drive serial chain (arm or leg like) robots. The ones that exist are usually oriented horizontally so that the motors don't have to resist gravity.

Or you could do a bit of searching for motor (+ gears) that fits what you want to do best at your price range. You will probably have to make some compromises, but with a bit of calculation, you can be reasonably sure of the performance of the robot before you buy the motors.

  • $\begingroup$ Thanks @haumptmech! Can you please explain the following: "If you are moving very slowly (almost not at all) you could also optimize the structure and type of metal carrying the magnetic field, as they are typically optimized for a moving magnetic field where the flux changed directions inside the metal rapidly." I was thinking the same: improve magnetic features of the core and/or magnet. Is it possible to get magnets twice as strong as those in gimbal motors? Is there a material twice as magnetic as the standard core? $\endgroup$
    – m_kramar
    Feb 19 '18 at 22:50
  • 2
    $\begingroup$ It sounds like you are trying low cost/low quality motors? Search for better motors (frameless torque motor for example). 1k-5k euro per motor for one optimized for this type of application is what you should expect to be paying. (maybe 70% of that if you are buying in quantity.) If you want to experiment with designing your own, invest in a couple books on permanent magnet motor design. $\endgroup$
    – hauptmech
    Feb 20 '18 at 1:33
  • $\begingroup$ I'm looking at $100 motors, something like this: iflight-rc.com/… $1K is definitely out of my range. $\endgroup$
    – m_kramar
    Feb 20 '18 at 5:02
  • $\begingroup$ I tried to improve my answer based on your budget. $\endgroup$
    – hauptmech
    Feb 20 '18 at 7:11

So, you want to use a servo? Why not just buy a servo?

Three things are required for "motor action" -

  1. A current carrying conductor
  2. A magnetic field
  3. An applied voltage

The result is relative motion between the conductor and the magnetic field.

Three things are also required for "generator action" - 1. A current carrying conductor 2. A magnetic field 3. Relative motion

The result is a voltage developed across the conductor.

This is important because once a motor starts it also becomes a generator of sorts, because it meets the criteria required for generator action.

The generator action in a motor will develop what is referred to as "back EMF". It is a voltage that acts against the applied voltage. The back EMF "fights" the applied voltage, which reduces the net voltage applied to the motor, which reduces the current draw.

The torque a motor can produce will always be directly related to the strength of the magnetic field inside the motor. The strength of the magnetic field is directly related to the current in the motor windings.

You make more torque in a motor by passing more current to the motor.

As you have noticed, more current leads to more heat in the motor, because of "I squared R losses." This term comes from the formula for electric power, $P = IV$, combined with Ohm's Law, $V = IR$, to give a formula for electric power in terms of current and resistance:

$$ P = I^2 R \\ $$

The $I^2R$ losses are the reason why high voltage electric lines are used for power transmission - higher voltages reduce current. Power losses increase with current squared, so it's important to get current as low as possible to reduce the power losses.

So, in order to get a high stall torque, you need a high applied current. As you have seen, you cannot take a "standard" or general purpose motor and use it at stall conditions because most motors are designed with back EMF in mind. The back EMF protects the motor by reducing current draw, which reduces the heat generated ($I^2R$ losses) inside the motor.

To operate predominately in the stall region of a motor requires more cooling to dissipate the heat generated, and/or different wire insulation that is capable of withstanding higher winding temperatures.

A motor spec sheet should list the "nominal" or operating current for the motor. The alternative to buying a specialized motor might be to operate the motor in conjunction with a current-regulated power supply instead of a voltage-regulated power supply. Throttle the motor to no more than the specified operating current and the heat generated in the motor should be manageable.

I say this might be acceptable because most higher-power motors feature an integrated fan. If the motor has a fan, then the rated current for the motor likely depends on the fan rotating at the nominal motor speed.

You can meet the requirement by creating a blower for the motor - this is a common practice where slow speed, high torque motors are used, as in cranes. Essentially, you disconnect the fan from the motor you want to operate at stall condition and attach it to the output shaft of a second motor. Any time you apply power to the first motor, you spin the second motor at whatever speed the fan is supposed to turn.

Now you get all the design cooling for the first motor and, assuming you're operating at or below the rated current for the first motor, you should be able to operate at or near stall torque for as long as you want.

  • $\begingroup$ Hi @Chuck, than you very much for detailed answer, can you please expand on a couple of points. It looks to me that the only way is to improve magnetic properties of magnets and core. Is it possible to get magnets twice as strong as those in gimbal motors? Is there a material twice as magnetic as the standard core? Another confusing point is number of windings: what will give me max torque/current ratio given battery voltage? $\endgroup$
    – m_kramar
    Feb 20 '18 at 0:53
  • $\begingroup$ @m_kramar - Maximize torque by maximizing the magnetic field. Maximize the magnetic field by increasing the number of windings and/or the current through the windings. Increase current through the windings by lowering resistance across the windings, which you could do with lots of windings in parallel. If you're only using permanent magnets, then you are at the limit of whatever the strongest magnet you can find is. If you use coils (windings - electromagnets) then you can vary the field strength by varying the applied current. $\endgroup$
    – Chuck
    Feb 20 '18 at 13:28

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.