Re: Ball Bar Tests & accuracy

Hi Greg,
What you say is true, however you are only looking at the revolutions
of the motor, not the revolutions of the lead screw or the travel of
an axis.

As a general rule, a servo motor is designed to operate at a higher
rpm than a stepper motor. This is why many stepper systems are direct
drive and many servo systems are geared down with pulleys.

If you have system running with a direct drive 200 step stepper
motor, the lead screw will lose a step at 0.005 revolutions (1/200)
of the lead screw.

If you replace the stepper with a servo and gear it down 4:1 then
with a typical 2000 line encoder providing 8000 counts per revolution
of the motor, it is providing 32000 counts per revolution of the lead
screw. With the 128 count fault level of a Gecko G320 this is 0.004
revolutions (128/32000) of the lead screw.

So in this case the dynamic accuracy of the servo system is higher
than it is for the stepper system.


You can also lower the fault count on a Servo system even more by
using a Gecko G340 in step multiplier mode.

With the Gecko G340 the error count is related to the input steps
based on the pulse multipliers setting:
1X mode => 128 input counts (just like a G320)
2X mode => 64 input counts
5X mode => 26 input counts
10X mode => 13 input counts

Using these counts, the revolutions of the lead screw for a fault
become:
2X mode => (64/32000) = 0.0020 revolutions
5X mode => (26/32000) = 0.0008 revolutions
10X mode => (13/32000) = 0.0004 revolutions

The bottom line is that a Gecko servo system can be designed to have
a higher accuracy than a stepper system. It all depends on what
gearing and encoder you are using.

Wayne Whippo
Boulder, CO

>Greg Jackson wrote:
>For the most part, stepper versus servo will not be much of a

comparison in

>accuracy. One of the big advantages of a servo is that, because of

the

>feedback, you can sustain large dynamic errors and still not fault

the

>machine.
>
>With a stepper system it will become apparent very quickly if you

lose

>steps. For the most part, if you are not losing steps, you are

guaranteed

>to have a dynamic accuracy of much better than one full step. If

you get

>anywhere near a step in lag or lead, you will probably lose that

step.

>While running a complicated program, if your position reference is

the same

>after as before, you can know that the dynamic accuracy during the

cutting

>was also good and within better than one step. That would be one

real step,

>not a microstep, as the system cannot "slip" a microstep, only a

real step.

>
>With a servo system, your only absolute knowledge of dynamic

accuracy will

>come from the position fault error. If it does not fault, at least

you know

>that it stayed within that error level. If the servo drive does not

have

>integral position gain, you can get a ballpark idea of dynamic error

by

>plotting dynamic drive current. With a scope or other

instrumentation on

>drive current, physically displace the motor shaft and document the
>relationship between motor shaft error and drive current. This is
>approximately the same relationship that exists while it is

running. You

>can then run a part program and continue to watch drive current. As

the

>part is running, the dynamic position error of the axis is roughly

shown by

>drive current. With an intelligent drive you can monitor dynamic

error, but

>this technique works pretty well for simple dumb drives as well.
>
>The bottom line is that, while servos have a number of advantages,

accuracy

>is not one of them. Steppers can provide a great deal of accuracy

at low

>cost. With a typical 2000 line encoder providing 8000 counts per
>revolution, the 128 count fault level of a Gecko is 0.016

revolutions. A

>200 step motor will lose a step at 0.005 revolutions (1/200). In

operation,

>a stepper will probably be well within 0.002 revolutions. In order

to meet

>that sort of level, the Gecko servo would have to maintain an error

of less

>than 16 counts. This can happen, but you cannot know that it has

happened,

>all you know is that it never got greater than 128.
>G. Jackson