Re: mill threading, Acceleration, Axis

A very good reply Jon, but i'd like to add a couple
of things. Mainly, "moment of inertia" figures are
fine for us nerds but not a lot of good for someone
building a machine.

The main thing that will stop a stepper during an
acceleration curve is a resonance. The motor will have
a few resonant frequencies, (harmonics) each one
getting worse as the speed increases, and then
you get past resonant frequencies.

There is a really good picture of the ramp graph
showing resonances here:
http://www.ezy.net.au/~fastvid/ramp.gif

With a home built system you can forget the formulae
as the entire machine will have it's own resonances,
and these will greatly affect the stepper acceleration.
The first step is to set up the system so it can be
slowly ramped and then held at any speed. Then
manually find all the resonant speeds, probably four
or five speeds. Write down how bad each resonance is
and what speed it starts/ends.

Then if you can design your own ramping rates,
accelerate the motor through the bad resonances
quickly. Normally a 3 stage ramp gets the job
done fairly well:
1. Medium ramp from start speed to before resonance.
2. Fast ramp through main resonance.
3. Fairly slow final ramp to max speeds.

Also note that many motors have full torque even
at speeds of 1 rev/sec or more, you may be able to
have a reliable motor start speed at 1 or even 2 rps,
if so, you can START the motor at a speed above
some of the lower resonant speeds, and accelerate
it from there. Steppers can achieve incredible
acceleration if you accelerate it at a small and
definite amount (2%??) per step. My test mill here
goes from stopped to 10rps in about 1/4 of a motor
revolution.

A final technical note, low inductance steppers
move quicker from step to step. They accelerate
harder within EACH step, quickly moving to the
new position and slamming to a stop at that step.
For this reason, they will have worse resonance
at low speeds. If you have a low inductance motor
and never use it faster than 10rps, it would give
low speed performance gains to change to a higher
inductance motor, or add a simple inductor in
series with each motor coil. So if your motor has
plenty of power at your normal higher speeds,
but bad resonances at lower speeds, this might
fix it. I should mention that going to a
microstepping driver will help, especially one
with special current ramping etc built in.
-Roman

> Alan Marconett KM6VV wrote:
>
> > In a related topic, is there a way to determine the "best"

acceleration

> > curve for a stepper motor?
>
> Yes, but it is not easy. You need to get some very tricky numbers
> from the motor manufacturer, such as rotational moment of inertia,
> 'viscous' damping (which isn't really viscous, but comes from eddy
> currents in the stator) and such. You then combine these figures

with

> rotational moment of intertia from the rest of the machine, and

compute

> where the resonances are, and how strong they are. If they are

strong

> enough, you have to stop right there and either make the machine

more

> rigid or add a real viscous damper (one of the best is a hollow,

fluid-filled

> disc with a free-floating steel disc inside) to damp out those

resonances.

> Then, you can compute the required torque to provide the desired
> acceleration, and work from there to find an acceleration that will

not

> exceed the motor's available torque at any speed along the

acceleration

> ramp. Motor and drive makes have torque vs. speed curves for

various

> combinations of motors and drives. They are extremely informative
> and useful.
>
> Jon