Re: Relation of a DRO to a CNC modification of a machine?
Posted by
caudlet
on 2004-09-28 15:42:26 UTC
--- In CAD_CAM_EDM_DRO@yahoogroups.com, R Rogers <rogersmach@y...>
wrote:
Then report an error if they are out of sync or maybe correct it
self and tell you it did. 2 encoders per servo? Seems like they
should be able to use the existing encoder and make it a true closed
loop system. Mine never loses position but in the event that it did
it would be nice to know.
system. If you make the response too slow, its worthless since you
may well be past the point where the correction needs to be
applied. You would be making up for skipped steps in a part of the
cut that may not benefit from added steps! If you make the
loop "gain" too tight you then have timing problems. Since the look
ahead buffer loads up steps in advance you would have to deal with
injecting the steps into the pulse engine cache. Servo systems are
complex to deal with because you have to balance the feedback to the
load. A little too much feedback and you get the oscillation
(screeching microphone effect) and too little makes the system have
poor accuracy and spongey response. Think of it as pulling a weight
with a spring and trying to accurately position it over a changing
surface. Oh, and you can only pull with jerky moves; not a steady
even pull.
As you stated, yours does not loose position since you have it
designed right and are operating it within the boundaries of the
mechanics you have. I have seen industrial systems with rotary
encoders on steppers and a BUNCH of supporting electronics to modify
the pulse rate based on postional feedback. They were very slow
moving machines which makes the loop design easier.
Servo's are not a substitue for poor design but they will correct
for small variations and easily compensate for wide load changes.
I have machines built with steppers and machines with servos. They
both work fine but getting faster speeds with better torque was
easier with servos.
While Art may well surprise us and pop up someday with a closed loop
(closed at the sofware end) MACH? for step based systems I'm not
sure I would hold my breath.
For the relative small delta between the step to servo cards like
the Gecko, Rutex or for lower current/volts apps the Imserv card and
the price of the motors, you could have a REAL servo system and
experience the smooth moves you can get with those systems.
wrote:
> Andy ,position and secondary rotary encoder or linear scale position?
>
> Wouldnt it be nice if Mach2 would keep an eye on the machine
Then report an error if they are out of sync or maybe correct it
self and tell you it did. 2 encoders per servo? Seems like they
should be able to use the existing encoder and make it a true closed
loop system. Mine never loses position but in the event that it did
it would be nice to know.
>fundimental problems anytime you start closing a loop in any
> Ron
>
>There have been discussions about that topic but there are some
system. If you make the response too slow, its worthless since you
may well be past the point where the correction needs to be
applied. You would be making up for skipped steps in a part of the
cut that may not benefit from added steps! If you make the
loop "gain" too tight you then have timing problems. Since the look
ahead buffer loads up steps in advance you would have to deal with
injecting the steps into the pulse engine cache. Servo systems are
complex to deal with because you have to balance the feedback to the
load. A little too much feedback and you get the oscillation
(screeching microphone effect) and too little makes the system have
poor accuracy and spongey response. Think of it as pulling a weight
with a spring and trying to accurately position it over a changing
surface. Oh, and you can only pull with jerky moves; not a steady
even pull.
As you stated, yours does not loose position since you have it
designed right and are operating it within the boundaries of the
mechanics you have. I have seen industrial systems with rotary
encoders on steppers and a BUNCH of supporting electronics to modify
the pulse rate based on postional feedback. They were very slow
moving machines which makes the loop design easier.
Servo's are not a substitue for poor design but they will correct
for small variations and easily compensate for wide load changes.
I have machines built with steppers and machines with servos. They
both work fine but getting faster speeds with better torque was
easier with servos.
While Art may well surprise us and pop up someday with a closed loop
(closed at the sofware end) MACH? for step based systems I'm not
sure I would hold my breath.
For the relative small delta between the step to servo cards like
the Gecko, Rutex or for lower current/volts apps the Imserv card and
the price of the motors, you could have a REAL servo system and
experience the smooth moves you can get with those systems.
Discussion Thread
gfull
2004-09-27 11:30:14 UTC
Relation of a DRO to a CNC modification of a machine?
R Rogers
2004-09-27 11:42:54 UTC
Re: [CAD_CAM_EDM_DRO] Relation of a DRO to a CNC modification of a machine?
Alan Rothenbush
2004-09-27 11:58:04 UTC
Re: [CAD_CAM_EDM_DRO] Relation of a DRO to a CNC modification of a machine?
Andy Wander
2004-09-27 12:03:25 UTC
RE: [CAD_CAM_EDM_DRO] Relation of a DRO to a CNC modification of a machine?
Andy Wander
2004-09-27 12:07:19 UTC
RE: [CAD_CAM_EDM_DRO] Relation of a DRO to a CNC modification of a machine?
gfull
2004-09-27 14:24:49 UTC
Re: Relation of a DRO to a CNC modification of a machine?
caudlet
2004-09-27 14:59:02 UTC
Re: Relation of a DRO to a CNC modification of a machine?
R Rogers
2004-09-27 16:19:47 UTC
RE: [CAD_CAM_EDM_DRO] Relation of a DRO to a CNC modification of a machine?
Vince Negrete
2004-09-27 16:25:33 UTC
Re: [CAD_CAM_EDM_DRO] Relation of a DRO to a CNC modification of a machine?
Andy Wander
2004-09-27 18:22:14 UTC
RE: [CAD_CAM_EDM_DRO] Relation of a DRO to a CNC modification of a machine?
caudlet
2004-09-28 15:42:26 UTC
Re: Relation of a DRO to a CNC modification of a machine?
R Rogers
2004-09-28 16:19:02 UTC
Re: [CAD_CAM_EDM_DRO] Re: Relation of a DRO to a CNC modification of a machine?
erie
2004-09-28 17:18:47 UTC
Re: [CAD_CAM_EDM_DRO] Re: Relation of a DRO to a CNC modification of a machine?
caudlet
2004-09-29 16:06:28 UTC
Re: Relation of a DRO to a CNC modification of a machine?
R Rogers
2004-09-29 16:35:28 UTC
Re: [CAD_CAM_EDM_DRO] Re: Relation of a DRO to a CNC modification of a machine?
gfull
2004-09-29 19:38:02 UTC
Re: Relation of a DRO to a CNC modification of a machine?
Joe Albeanese
2004-09-29 20:48:22 UTC
Re: [CAD_CAM_EDM_DRO] Re: Relation of a DRO to a CNC modification of a machine?
caudlet
2004-09-30 06:20:53 UTC
Re: Relation of a DRO to a CNC modification of a machine?
caudlet
2004-09-30 06:51:26 UTC
Re: Relation of a DRO to a CNC modification of a machine?
Joe Albeanese
2004-09-30 09:03:05 UTC
Re: [CAD_CAM_EDM_DRO] Re: Relation of a DRO to a CNC modification of a machine?
caudlet
2004-09-30 15:00:37 UTC
Re: Relation of a DRO to a CNC modification of a machine?
Joe Albeanese
2004-09-30 17:34:40 UTC
Re: [CAD_CAM_EDM_DRO] Re: Relation of a DRO to a CNC modification of a machine?
Alan Marconett
2004-09-30 17:39:34 UTC
Re: [CAD_CAM_EDM_DRO] Re: Relation of a DRO to a CNC modification of a machine?
Joe Albeanese
2004-09-30 18:56:40 UTC
Re: [CAD_CAM_EDM_DRO] Re: Relation of a DRO to a CNC modification of a machine?