RE: EMC minor quirk

[Ray Henry <rehenry-at-up.net>]

Dan B. 

Welcome.  You have a good point with tangent lines to an arc.  I seem to
remember reading that EMC's cutter comp attempts to maintain the same
feedrate at the point where the tool touches the part regardless of the
shape of the cut being made.  If you move into the inside of an arc the
machine speed should slow, if outside it should speed up. 

When EMC comes to an outside corner it describes an arc around the corner
using the hard corner as the center, and the radius of the cutter as the
path radius.  Then it uses the approach to the corner as tangent offset by
radius and the line after the corner as tangent offset by radius also.   

I don't know if this is what Jon was hearing but a hard 90 sounds pretty
strange with my steppers.  It's like going around a quarter circle where
one axis slows and the other accelerates.  

Ray


At 09:44 AM 2/13/2000 -0500, you wrote:
>Hi,
>
>I have been following the EMC mail list for a couple of weeks now, and I am
>admittedly a newbie at this, so if I am asking a stupid question I hope you
>will excuse me.
>
>I am involved in programming CNC machines in the woodworking industry, and I
>have noticed that straight line moves that lead into an arc always run
>better if the line is tangent to the arc.
>
>Is this the case of the situation you describe below.
>
>If the line is not tangent to the arc,  I wonder if you notice the same
>acceleration when it is.
>
>I have chalked up this behavior on our machines to the guess that it is
>easier for the controller to calculate the tool path if lines and arcs are
>tangent since no matter how much the tool dia. compensation offset is the
>end points of the line and arc will always meet if they are tangent.
>
>Any comment on my guess would be welcome.
>
>Thanks for letting me comment,
>
>Dan Blouse
>
>-----Original Message-----
>From: emc-at-nist.gov [emc-at-nist.gov]On Behalf Of Jon Elson
>Sent: February 12, 2000 8:41 PM
>To: Multiple recipients of list
>Subject: EMC minor quirk
>
>
>This is one for Fred Proctor or Will Shackleford;
>
>When EMC is moving in a rounded-off rectangular
>path of alternating linear and arc moves, with cutter
>diameter compensation turned on, I noticed that at the
>end of each straight line, the machine would accelerate
>monentarily, before starting into the arc move.
>It is not serious, but seems like there is a quirk in the
>calculation of velocity on the offset path to get the
>desired velocity on the actual cutting line at the perimeter
>of the tool.  Since the velocities on the linear parts
>are the same (offset and actual) it must have something
>to do with the calculation for the offset velocity for the
>arc.  But, the arc seems to be done right, it may be the
>software is wrongly calculating arc velocity when it
>attempts to match velocity at the end of the line with
>the beginning of the arc.  This is the best that I can
>interpret the behavior.  If a stepper system will exhibit
>this same behavior, the audible steps might be very
>obvious.  It is a little less obvious with servos, but
>I can clearly detect the accelerations.  The RS-274
>program has all of this movement programmed at a
>constant speed of 10 IPM.  The tool is set to a
>diameter of 0.5".
>
>My setup is using the 20-DEC-1999 EMC, with
>servos driven by the Servo-to-Go card.
>
>This is not a serious problem, but if I have guessed
>the origin of the problem correctly (a long shot) then
>it might get more serious the larger the tool diameter
>is set to.  There is no evidence of this speed variation
>when cutter diameter compensation is not active.
>
>Thanks,
>
>Jon
>
>
>
>
>
>