alignment setting?

[pcmac]

Is there a setting that affects alignments? I use 3.7 MR1 and wasted a day trying to get a correct alignment; last shot went to V4 and it nailed it first try. I am still pretty green to this so I thought it was me. The alignment wasn't off the board (about .030") but I always quickly check by another method and knew part was good. Very simple part - large o-ring with 14 pins and ids around the circumference so alignment is plane, id using cylinder, and then lock on two pins opposite each other. I figured I must have changed some setting because sizes are fine but locations are off center. I had too many problems with V4 so not switching just yet but I am downloading 3.7mr3 after this post. Any ideas?

[spazus_maximus]

recalibrate?

what do you mean by alignment being off .030" That sounds huge! Is it an iterative threshold?

[than_357]

Ideas? Lots of them...but that can be bad.

Really- not enough information to give an answer. Either you have a 3d alignment and a wrong axis, or something in the layout is wacked to give .03". Heck- that's eyeball territory.

[pcmac]

I was hoping I somehow changed a setting in PCDMIS. Yes .030" is lots but I would have thought that it would either be totally out of sight or right on as it worked in V4 the first shot and repeated again. I can't see how such as simple alignment doesn't work unless I changed something. I do manual, then DCC, hitting a minimum for the manual (3 top plane, 4 on id, 1 hit on the side of a pin on either side to establish a line). Then a second alignment with more points for plane in DCC, construct a cylinder from two circles for id and then indexing and constructing a line from 2 constructed cylinders (using 2 circles for each again) from ID's at opposite sides of OD. I just downloaded MR3 and will load tomorrow - hopefully that will fix whatever I appeared to have screwed up. It is showing off center in both X and Y, Z is correct, sizes are correct. One more scotch, sleep, and try again tomorrow - thanks.

[Jeffrey Lovely]

Try 3 rows of circles for your cylinder- then intersect to a plane
or use auto cylinder 3 rows - no constuction

[\v/inston]

like jeff said you should always use aleast 3 levels when measuring a cylinder. also you can level to a cylinder.

[cmmguy]

Need more info

You need to post the exact steps of the alignment process that you are taking. All features, all alignment steps(what was done) and in the right order. The order is important. It sounds like a step is missing or in the wrong order. Or you are setting an origin on a 3D feature and its "centerpoint" is not in the same place each time. Also what is .03 off to what...

[pcmac]

Problem was something to do with probe build or calibration; I don't know what exactly and how I forgot this but all programs worked when I deleted the probe file, built a new one, and recalibrated. I have had problems from probe build (or calibration?) in the past and I should have tried it earlier. Any ideas on that? - I am positive it is something with this and not the programming.
Question - I usually only pay attention to size (D) and form in results of calibration; do I need to look at X,Y, and Z theo vs. measured. Thanks again - I will try the cylinder build methods also in future programs.

[Kingkrab]

If Your Using A Tp-200 Make Sur The slow Button Is On!