The most correct way to know that would be to look at the datasheet of the motor manufacturer. Or use a simple equation of the movement (rotation) to predict the system acceleration and velocity in time. Another aspect would be the response time between the electrical-mechanical part. It would be proportional to PWM frequency (the system should recognize the duty cycle to start moving).

Dear Olexiy Shynkarenko

How can you do it by the datasheet and the equations of movement? Can you set an example of using the data of the datasheet and the corresponding equations?

The time required to generate rated torque is purely electrical.The time for rated

velocity is dependent on how you model the load torque . Actually it is a coupled

electromechanical equation.For design approx selection purposes ,try :

1. Tmotor = J dwm/dt. + B.wm. ..Inertia , damping coefficient estimated parameters .

2 Knowing rated speed and time you can get an idea of required torque for selecting the motor.

Cheers

I would do it empirically

I know the best way is to do it empirically Hugh Lachlan Kennedy, but I haven't got the actuator(s) yet and I want to use the "time to get to a certain torque" as a constraint in my controller. If one motor isn't good for my robot to follow my required trajectory, I would test another motor to find the best out of them that fits my application.

Dear Narasim Ramesh, I know the robot link inertia (J_link) that the motor should apply torque on. Suppose that this is the datasheet of the motor that I'm working with:

https://www.maxongroup.com/maxon/view/product/motor/ecmotor/ecflat/ecflat90/607930

Is there any equation to find out how much the required time to get to a certain torque is (e.g. from 0 to 30N.m.) by only having the datasheet and load moment of inertia?

Thanks to all of you guys.

Dear Mahdi

Please see if attached pdf is helpful.

Merry Christmas Peace Joy and success.

Cheers

That was great Narasim Ramesh

Thank you!

Dear Mahdi:

you can use this helpfull article at: https://www.industrial-electronics.com/industrial-brushless-servo-mtr_4.html

Thanks for sharing the link, Hassan Nasser

Hi Mahdi,

I am sure you ask sincerely, but it is a trick question: the motor you selected has a peak torque of 14.5 N*m—it will NEVER get to 30 N*m before melting.

The motor torque is purely current dependent, and its rise time is purely inductance- and resistance-dependent. The characteristic current rise time is Tau_elec = L/R, which is also the characteristic torque rise time. For the selected motor, L/R ~ 1.3 ms.

The mechanical time constant (100% speed) is Tau_mech = (R*J)/(Ke*Kt) [R=resistance, J=rotor inertia, Ke=back-emf constant, and Kt=torque constant] and determines the motor speed rise time. For the motor you selected, Tau_mech ~ 7.2 ms.

For excitation at 30 V, you can reach max torque in ~3*Tau_elec, but the motor inertia cannot catch up until Tau_mech. Note, as the motor speed catches up, the back-emf generates a voltage which reduces the applied excitation, slowing down the catching up.

In summary, you can change the torque fast, but the motor cannot follow, even unloaded.

Dear David Cope

Thank you for your comprehensive response.

And you are right about the maximum torque. The torque that I meant was after using a gear (which I have chosen 30:1 transmission ratio).

If dynamics, i.e., time of response, are an issue--and your question suggests they are--be aware that the transformation ratio you mentioned (30:1) means your load inertia get multiplied by ~1000 times (~30^2) when reflected onto the motor. That will likely greatly slow down your dynamics, especially time to achieve a motor speed. Specifically, J -->J' = J + 30^2*Jload in the eqn for Tau_mech. Try to live with a smaller transformation ratio, 2-3:1.

Thanks for your kind advice David Cope .