Postion Move's

[B. Jacobs]

:confused: I have a question I hope someone can help me with. I have a round part. I am standing it on it's end. I am taking the top as a plane and then measureing the circles. I am aligning the plane as zplus the circle as x,y center.
=ALIGNMENT/START,RECALL:A2, LIST= YES
ALIGNMENT/LEVEL,ZPLUS,PLN3
ALIGNMENT/ROTATE,YPLUS,TO,CIR3,ABOUT,ZPLUS
ALIGNMENT/ROTATE,XPLUS,TO,PLN4,ABOUT,ZPLUS
ALIGNMENT/TRANS,ZAXIS,PLN3
ALIGNMENT/TRANS,XAXIS,CIR3
ALIGNMENT/TRANS,YAXIS,CIR3
ALIGNMENT/END
I am then taking all the hits I need. Planes and circles.
When i run the part as soon as I am done programming it it works fine.
When I put a new part up there to run, it looks as though the x and y move 180 deg. I start to crash and the probe goes all over. Any thoughts what I might be doing wrong.
Thanks ahead of time.

[J. Pick]

Look at PLN4 to make sure your X-rotation isn't screwing it all up.
Stop standing on the end of the part - that can't be good for repeatability (j/k).

[Randy L.]

Is this part cylindircal? Could you give us more information as to what the part looks like?

[\v/inston]

if it is a cylinder, measure it as a cylinder with atleast three levels.

when you go to do the alignment, level to the cylinder.

[Wes Cisco]

I do this sort of thing often. Why do you rotate Y axis to cir 3(the center of the part?) :confused: Once you level z+ to the plane you only need one rotation for either x or y axis, whichever you rotate the other will be 90 degrees to it by definition. You need a feature that is either on the od or on the top face not at the center of the part. (the closer to the od the better). Measure that rotational feature. (Hopefully your print specifies which feature you should use by making it a "clocking" datum). Construct a perpendicular line from the center of cir 3 to the feature. Rotate either x or y axis to that line. Translate origins as you have (z to plane, x&y to cir 3). HTH ;)

[BSchwark]

More info on the part is needed. If this a ring gage or someting similar you can use a machine axis for a rotation. If there are more feature to check you need to define your rotation better. Perferably with a line type feature instead of a plane.

[B. Jacobs]

What I have is a cylinder that is four different diam. If you look at it like a soda can,
The top is .625 diam round, the body is .530 diam round then down in z about 3/4 inch there is a big circle 1.125 diam with a thickness of .06 in, then right below that .625 diam round.
I have made some lines out of the top circle. This seems to help.

[Wes Cisco]

If all of your diameters are concentric, just use machine axis for rotation. Your alignment would be 1)level z+ to plane, 2) x & y origin at cir 3, 3) z origin at pln. If one of your diameters is eccentric, measure it and consturct a line perpendicular from cir 3 to the eccentric dia. Then rotate one axis, (either x or y) to the constructed line as step two. HTH

[CORKSCREW]

I think your rotations are the problem. Why would you rotate something twice about the same axis.

ALIGNMENT/ROTATE,YPLUS,TO,CIR3,ABOUT,ZPLUS
ALIGNMENT/ROTATE,XPLUS,TO,PLN4,ABOUT,ZPLUS

What element do you want aligned, I suspect the plane is what you should align. If pln3 is an x-y plane, which I am assuming it is because you are leveling z to it, Pln4 should be a yz or an xz plane to rotate to. If pln4 is at a compound angle, it could also cause some deviations in you program.

[B. Jacobs]

Got it figured out. Thanks for all the help. I am still learning these alignments. Thats where my problem was.