Re: [CAD_CAM_EDM_DRO] Re: Brushless servo drive - Cheap Solution?????

mariss92705 wrote:

> Jon,
>
> The G420 is step and direction input. Same specs as the G320
> otherwise.
>
> Next in the hopper for new drives is is an evisceration of the
> step/direction interface which will be replaced with a +/-10V, 0-5V
> interface. It will be jumperable for current mode, voltage mode or IR
> compensated mode operation. In otherwords, a "classic" servoamplifier
> rather than a complete drive. The model numbers have not yet been
> picked.

I suppose we should go off list on this pretty soon, but...

I wonder if it would be possible to add a small front end (might need a
couple pots, instrumentation amp, etc.) to make this a velocity servo.
Since so many people get the shivers when I mention DC tachs, maybe
we should leave it as you plan it, and put the velocity loop in the
CNC computer, too, as it has pretty much all the info needed
(just not continuous-time). But, it would be nice if we could have
a little plug-in option for those who like analog velocity loops. I
suppose I could always make up the little option board, but then
I'd be adding stuff INSIDE the Gecko drive.

Another possibility is to make a PWM-input servo amp, maybe with
the pulse width and direction signal being opto-coupled inputs.
Oh, the problem there is that for current-mode, you'd have to go
back to analog for the current loop. Hmmm, I wonder if there's
a way to do that, easily. Obviously, this would be simple as a
voltage-mode amp. I have some experience with those, whenever
you get the gain more then 10% of what you want for positioning
accuracy, you get instability.

It would not be much of a hack to add the velocity loop into the servo
calculations that are done now. EMC already computes V and dv/dt
(The "V"s being velocity, not voltage). It would mean adding some
more initialization parameters to give the loop parameters to the
calculation, but that sounds like it might make a lot of people happy.

Jon