Re: [CAD_CAM_EDM_DRO] Re: Re edm software/ use hardware
Posted by
Marcus and Eva
on 2005-01-14 23:06:34 UTC
Hi William:
The control system you refer to, is for a sinker EDM machine.
This process drives a shaped electrode straight down into the workpiece,
therefore no motion control in the sense of a CNC path is required, and the
needed single axis motion can be controlled directly from the spark gap
conditions.
Wire EDM, on the other hand, requires that the electrode travel in a 2 axis
contour if you hope to get any useful work out of it.
That means that you must create a non-linear motion path for the wire, and
then superimpose on it, a capacity to advance, stop or retract depending on
the spark gap conditions and their persistence over time.
Reversing the G code in order to be able to back up requires you to have
on-the-fly ability to know where you are even before you reach your current
move destination and the means to calculate and write the code to go
backward, also on-the-fly, maybe many times per second.
This is because G code works by stating : "Go from this known point to this
new point, and do it using this motion algorithm" ie:
From a start point of X0 Y0:
G1 X1.0 Y 1.0 F20.0 says: Move from X0 Y0 to X1.0 Y1.0 using a linear move.
To make the move work, both the starting point (X0 Y0) and the ending point
(X1.0 Y1.0) must be known, as well as the motion algorithm (G1) and the
feedrate (F20.0).
Suppose you need to move backward for a little bit once you've gotten to
X0.5 Y0.5 while you're on your way to X1.0 Y1.0.
You have to tell the controller where you are (X0.5 Y0.5)
You have to tell the controller what the move destination is, say X0.498,
Y0.498 (backing up along the path 0.002" to let the dielectric back into the
spark gap)
You have to tell the controller what motion algorithm to use (G1)
You have to tell the controller how fast to do it (Let's use a fast feedrate
F 100.0)
Then you need to write more code to get you back to where you were, so:
G1 X0.5 Y0.5 F100.0
Then you write more code to continue the path:
G1 X1.0 Y1.0 F20.0
But the code ultimately just makes a pulse stream , so if you can
interrogate the pulse stream that got you to where you are, you can back up
simply by reversing its sense, without needing to calculate the code to
create a new pulse stream to back up.
This is the complexity of the problem that makes Robert Langlois' stepper
controller unsuitable for wire EDM applications that do anything more than
straight line cuts using only one axis at a time.
So, can anyone out there describe if this piggyback program can be written
as I envision?
Cheers
Marcus
The control system you refer to, is for a sinker EDM machine.
This process drives a shaped electrode straight down into the workpiece,
therefore no motion control in the sense of a CNC path is required, and the
needed single axis motion can be controlled directly from the spark gap
conditions.
Wire EDM, on the other hand, requires that the electrode travel in a 2 axis
contour if you hope to get any useful work out of it.
That means that you must create a non-linear motion path for the wire, and
then superimpose on it, a capacity to advance, stop or retract depending on
the spark gap conditions and their persistence over time.
Reversing the G code in order to be able to back up requires you to have
on-the-fly ability to know where you are even before you reach your current
move destination and the means to calculate and write the code to go
backward, also on-the-fly, maybe many times per second.
This is because G code works by stating : "Go from this known point to this
new point, and do it using this motion algorithm" ie:
From a start point of X0 Y0:
G1 X1.0 Y 1.0 F20.0 says: Move from X0 Y0 to X1.0 Y1.0 using a linear move.
To make the move work, both the starting point (X0 Y0) and the ending point
(X1.0 Y1.0) must be known, as well as the motion algorithm (G1) and the
feedrate (F20.0).
Suppose you need to move backward for a little bit once you've gotten to
X0.5 Y0.5 while you're on your way to X1.0 Y1.0.
You have to tell the controller where you are (X0.5 Y0.5)
You have to tell the controller what the move destination is, say X0.498,
Y0.498 (backing up along the path 0.002" to let the dielectric back into the
spark gap)
You have to tell the controller what motion algorithm to use (G1)
You have to tell the controller how fast to do it (Let's use a fast feedrate
F 100.0)
Then you need to write more code to get you back to where you were, so:
G1 X0.5 Y0.5 F100.0
Then you write more code to continue the path:
G1 X1.0 Y1.0 F20.0
But the code ultimately just makes a pulse stream , so if you can
interrogate the pulse stream that got you to where you are, you can back up
simply by reversing its sense, without needing to calculate the code to
create a new pulse stream to back up.
This is the complexity of the problem that makes Robert Langlois' stepper
controller unsuitable for wire EDM applications that do anything more than
straight line cuts using only one axis at a time.
So, can anyone out there describe if this piggyback program can be written
as I envision?
Cheers
Marcus
Discussion Thread
Mark
2005-01-10 19:22:54 UTC
Re edm software
TomP
2005-01-10 21:43:27 UTC
Re: [CAD_CAM_EDM_DRO] Re edm software
alflewerken
2005-01-11 12:47:34 UTC
Re: Re edm software
Stephen Wille Padnos
2005-01-11 13:42:42 UTC
Re: [CAD_CAM_EDM_DRO] Re: Re edm software
TomP
2005-01-11 18:53:00 UTC
Re: [CAD_CAM_EDM_DRO] Re: Re edm software
Stephen Wille Padnos
2005-01-11 19:04:42 UTC
Re: [CAD_CAM_EDM_DRO] Re: Re edm software
TomP
2005-01-12 08:20:54 UTC
Re: [CAD_CAM_EDM_DRO] Re: Re edm software
Alan Marconett
2005-01-12 09:40:34 UTC
Re: [CAD_CAM_EDM_DRO] Re: Re edm software
Raymond Heckert
2005-01-12 19:25:46 UTC
Re: [CAD_CAM_EDM_DRO] Re edm software
TomP
2005-01-13 07:46:04 UTC
Re: [CAD_CAM_EDM_DRO] Re edm software
TomP
2005-01-13 07:52:28 UTC
Re: [CAD_CAM_EDM_DRO] Re: Re edm software
Marcus and Eva
2005-01-13 08:25:30 UTC
Re: [CAD_CAM_EDM_DRO] Re: Re edm software
Alan Marconett
2005-01-13 10:21:04 UTC
Re: [CAD_CAM_EDM_DRO] Re: Re edm software
William Alford
2005-01-14 21:20:04 UTC
Re: [CAD_CAM_EDM_DRO] Re: Re edm software/ use hardware
Marcus and Eva
2005-01-14 23:06:34 UTC
Re: [CAD_CAM_EDM_DRO] Re: Re edm software/ use hardware
TomP
2005-01-15 17:22:13 UTC
Re: [CAD_CAM_EDM_DRO] Re: Re edm software/ use hardware