Re: Constant Velocity unwinder
Posted by
Graham Stabler
on 2006-07-18 03:05:20 UTC
--- In CAD_CAM_EDM_DRO@yahoogroups.com, "Fernando" <fer_mayrl@...> wrote:
"Form a summing node of the tach feedback (- voltage)
with a velocity command (+ voltage)"
just means stick the voltage from the tach and the voltage from a pot
you will use to set the speed into a op-amp summing amplifier.
Because the voltages are different signs the output of the op-amp will
only be zero when they are both exactly the same otherwise it will be
outputing a voltage to tell the VFD to do something. The P part is
just the gain of that amplifier and the I and D are obtained by adding
some capacitors etc, I gets rid of the steady state error and D
responds to the dynamics (I doubt you need D).
I have done something quite similar to this recently except it was PI
not PID and I am trying to maintain tension in EDM wire rather than
speed but the principle is the same. My circuit is breadboarded up
and working, I can try and find the schematics if I produced any. It
was driving a DC motor directly (well via a transister) but you could
drive the 0-10v control on the VFD with no problems. You could also
find a frequency to voltage converter circuit on the web to convert
and encoder to a voltage if you fancied a non contact feedback method.
Unlike a typical servo system you need no reversal of direction and as
you seem to be countering the effects of spool size I suspect that the
system does not need to have particularly good dynamic response etc so
it should be fairly non critical.
Graham
>Mariss used the word node which makes it sound worse than it is :)
> Thanks Mariss,
> Not that proficient in electronics, but I will look into it.
> I will let you know.
> Regards and thanks a lot
> Fernando.
"Form a summing node of the tach feedback (- voltage)
with a velocity command (+ voltage)"
just means stick the voltage from the tach and the voltage from a pot
you will use to set the speed into a op-amp summing amplifier.
Because the voltages are different signs the output of the op-amp will
only be zero when they are both exactly the same otherwise it will be
outputing a voltage to tell the VFD to do something. The P part is
just the gain of that amplifier and the I and D are obtained by adding
some capacitors etc, I gets rid of the steady state error and D
responds to the dynamics (I doubt you need D).
I have done something quite similar to this recently except it was PI
not PID and I am trying to maintain tension in EDM wire rather than
speed but the principle is the same. My circuit is breadboarded up
and working, I can try and find the schematics if I produced any. It
was driving a DC motor directly (well via a transister) but you could
drive the 0-10v control on the VFD with no problems. You could also
find a frequency to voltage converter circuit on the web to convert
and encoder to a voltage if you fancied a non contact feedback method.
Unlike a typical servo system you need no reversal of direction and as
you seem to be countering the effects of spool size I suspect that the
system does not need to have particularly good dynamic response etc so
it should be fairly non critical.
Graham
Discussion Thread
Fernando
2006-07-17 15:16:45 UTC
Constant Velocity unwinder
Mariss Freimanis
2006-07-17 16:48:26 UTC
Re: Constant Velocity unwinder
Fernando
2006-07-17 17:30:57 UTC
Re: Constant Velocity unwinder
Mark Whitis
2006-07-18 02:10:05 UTC
Re: [CAD_CAM_EDM_DRO] Constant Velocity unwinder
Graham Stabler
2006-07-18 03:05:20 UTC
Re: Constant Velocity unwinder
Fernando
2006-07-18 08:30:01 UTC
Re: Constant Velocity unwinder
Fernando
2006-07-18 08:45:38 UTC
Re: Constant Velocity unwinder