Re: [CAD_CAM_EDM_DRO] Question Re: Work Offsets

inetd_conf wrote:

>I had some odd behavior on an EMC system running BDI 2.16:
>
>I spent the last two days PID tuning to get the axes running
>cleanly. They're running cleanly now, and in the process I had the
>opportunity to set up and run a number of toolpaths, mostly
>prototyping in pine so a tool crash wouldn't be disastrous.
>
>Today I tried to run an actual part. I used an edge finder to
>indicate in the workpiece, and set my zeros by right-clicking on
>each axis in turn and setting it to zero.
>
>

Unless your edge finder has a zero radius tip, O don't know how
this can work. I usually use an edge finder with a .200" end, and
set the axis to either -.100 or +.100" depending on which side of
the part I am on. Just to protect against dumb mistakes, after setting
both axes, I then do X0 Y0 in MDI mode, and manually set the edge
finder to be centered. If I have made the wrong sign entry above,
the error will be obvious, as it will be .200" off center from the
part corner.

>When I ran my program, it proceeded to crash the tool into the
>workpiece (even though the first move is to go to (0,0,0) and raise
>the quill to 2.0" off the top of the workpiece.)
>
>I carefully checked my zeros and loaded a new part. Same behavior,
>only this time it snapped the end mill off.
>
>Before changing a tool and crashing it again, I played around with
>it.
>
>In MDI mode, if I typed "G0X0Y0Z0", it would go to (0,0,0) in the
>relative coordinate frame. If I ran the program, it would attempt
>to go to (0,0,0) in the absolute machine coordinate frame (the Z
>value of which was well inside the workpiece).
>
>

I know you have set X and Y with the edgefinder, but did you set the Z
axis to the right coordinate offset?

>I stripped the entire header off the program file so it starts with
>that "G0 X0.000 Y0.000 Z0.000" statement. Same behavior.
>
>What am I missing here? I thought MDI and running a program was
>essentially the same thing. I didn't know one would run in one
>coordinate frame and one would run in the other. I'm REALLY puzzled
>why this behavior didn't show up during the two days of PID tuning I
>did.
>
>

It may require you to have the SCREEN DISPLAY set to "Relative"
mode when you set the axes from the screen. If the screen is displaying
absolute coordinates, then the screen buttons to set the axes will set
the ABSOLUTE coords to zero, not the RELATIVE coords, which is
what you want.

Jon