I find that two different motors both become unreliable at starting at about 5000 full steps per second.

I am using a Motorola MC3479 chip to drive small motors wired as bipolar. I can run the motors as fast as about 750 steps per second, but then they stall with no load. They will not restart until I reduce step frequency to about 500 steps per second.

Is this slower than normal? Do commercial drives die at such a low step rate? I know steppers lose torque at higher step rates due to inductance affects. Am I there yet, or is something else likely wrong?

Eric Ratliff
http://www.icpdas-usa.com/

 

Inertia, resonance and other mechanical factors
are even more important than the electronics in
getting steppers to do what you expect. I would
say your experience is typical. Put them in the
application and deal with the issues there. Even
if you can get them to work great with no load,
things will be different when they are actually
applied. High step rates require damping or a
favorable load to control the motion. Ask me
how I know this. :^)

Regards

cww

 

What is "normal" depends on many things: drive voltage, load intertial, and friction. You should ramp up the step rate to model acceleration. If you are starting at 500 sps and then running at 750 sps full speed, that is not much of a ramp up. Start at 50 sps and then ramp up quickly, but smoothly, to your running speed and you will achieve optimal performance.

Robert Scott
Real-Time Specialties
Embedded Systems Consulting

 

This is fairly typical for this type of drive. The MC3479 is a pretty old chip. There are much newer chips available using MOSFETs. If you are using a small motor try the easystepper from spark fun. To get faster speeds you need to overcome the inductance in the coils. To do this they have developed the chopper drive which switches on and off current rapidly to the coils. By using higher voltage power supplies with the chopper drive you are able to overcome the inductance.

 

Before changing to a chopper drive, first do some simple calculations to see if inductance really is the limiting factor in this case. I assume the drive in use right now is current-limited using a series resistor, R. Find the inductance of a motor coil and calculate the L/R time constant. If that time constant is substantial compared to the time for one step (1.4 msec.) then inductance is indeed a factor. If, however, the L/R time constant is less than, say, 0.14 msec., then going to a chopper drive will not help. It is an easy thing to check, as compared to building an entirely new system with a chopper drive.

Robert Scott
Real-Time Specialties
Embedded Systems Consulting

 

Robert,

Thanks for you great comments. I was starting from zero to 500Hz (for example) to determine what seemed to be a maximum 'reliable' step rate. I had seen ramping up would allow higher rates. I have been inputting steps with an old HP Oscillator with a huge dial.

I did not find the EasyDriver V3 from Sparkfun before in my searches. It is at http://www.sparkfun.com/commerce/product_info.php?products_id=8368 Wow, this could have saved me some time.

I think I can measure the inductance with the oscillator and a resistor in series with the motor.

I do now have the motors connected to their load. Now the maximum step rate that can start from 0 is 500 HALF steps per second.

P.S. Yes, I am using 4 series resistors to limit current to 2 motors. I had hoped that this would make more voltage available to overcome inductance as speed increased. I am using a 12V supply. I have 17 Ohms for the 5 Ohm motor. I have 12 Ohms for the 10 Ohm motor.

Eric Ratliff
http://www.icpdas-usa.com/

 

Curt,

Thanks for your comments. The step speed dropped by 50% when I connected to the machine. I'm not sure if it' inerta, friction, or both that are mattering.

Eric

 

Dear friend please do a simple measurement and let us know the result. If you have an oscilloscope please connect the probes to the inputs of the motor while running and see the shape of the pulses inputted to the motor, while you increase the frequency of the pulses. This test will tell you lots of things.

If the shape of the pulses are changed dramatically or if the ampiltude or duration of the pulses are reduced or the input to the motor no longer resembles a pulse as you go above 500 sps you should conclude that you have an electrical problem. In this case it is most probable that you have a weak power supply for the motor and the motor overloads the power supply and you should use a better drive for your motor.

If there is no change in the shape of the input pulses then we conclude that it is a mechanical problem(maybe the inertia problem or...).
Please do the test and let us know the result.

[email protected]

Good luck