Re: [CAD_CAM_EDM_DRO] step/dir servo (Gecko drives)

Tim Goldstein wrote:

> Carlos,
>
> I have some of the Geckodrive servo units. I have not yet gotten them
> attached to my machine (hopefully tomorrow. I got my last encoder
> hooked to
> the motor), but I have been playing with then on some bare shaft
> motors. The
> rapids I am getting depends upon the specific motor. I have one nema
> 42 unit
> that is maxing out at 135 ipm (on my machine this works out to 1350
> rpm),
> but have a nema 34 that is fine to at least 180 ipm and a nema 23
> that is
> fine at 240 ipm. The question is how much of this speed will I be able
> to
> retain when mounted on the machine?

Probably most of it. The PWM drive can only produce an output voltage a
little less than the
DC supply, perhaps 90 - 95% of the supply voltage, depending on how
close it can go
to 100% modulation. (Mine can go to 100%, but you still lose a little
voltage across the
transistors. Also, if you drive the servo into saturation, it
introduces a discontinuity
in the transfer function, which upsets the stability of the servo loops,
so you shouldn't
really count on going to saturation for best results.)

Also, the series resistance of the motor armature windings is in series
with the
source. If your motors have a 1 Ohm winding, and the amplifier's FETs
have
an equivalent of, say, 2 Ohms, then you have 3 Ohms in series with the
power
source. So, if your motors need 3 amps to move the machine under some
(presumably light) load, the voltage drop is 3 Ohms x 3 Amps = 9 Volts.
If your supply is 75 V, 90% of that is 67.5, then 67.5 - 9 = 58.5 V.
Given
these rough numbers, you should be able to get a rapid feed rate equal
to
the motor's free-running speed at 58.5 V when loaded by the machine.

This should give you a very rough idea how to guesstimate the required
voltage to get some specific performance, within maybe 15% accuracy.

Jon