my probe is confused, or is it me

[Kevin Watts]

Good morning,
pcdmis 3.5 mr 1, ph10t with tp20 and star probe.

im trying to do an alignment where i measure a cylinder. i must use tip 5 for 1 circle and tip 3 for the other circle. im using 4 hits per circle and pcdmis is calling it a sphere. i can override the guess and call it a cylinder but why would it guess a sphere and will this cause problems in the alignment when i change it to cylinder? all tips calibrate good. tip 5 is 3 X 40 and tip 3 is 4 x 20.
thanks in advance for all replies

[craiger_ny]

You can't make a cylinder from 4 hits. Or are you measuring a circle, calling a tip, and then measuring another circle thereby actually measuring a cylinder? If that is the case I'd construct from 2 circles. Are you in the appropriate workplane?

[Kevin Watts]

yes i am in the correct workplane, i am taking 4 hits on each circle

[#2#]

You can't make a cylinder from 4 hits. Are you in the appropriate workplane?

He's got a total of 8 hits and I don't think workplane should matter for a cylinder.

I'd try taking more than 4 hits in each circle. Bump it up to 6 or 8 and maybe it won't guess it is a sphere. Or, do the override.

[John Kingston]

Is this a cylinder constructed from 2 circles or a measured cylinder?
PC-DMIS compensates for ONLY the last tip in a feature. That can create a source of error.

[craiger_ny]

I asked about workplane because I do not know if he is measuring circles or cylinders. I am thinking he is doing hits, calling a tip, then doing more hits the final result of which is a cylinder. I do not know how it comps though when you do that. It may be a comp issue between the 3mm tip and the 4mm tip being used in the same feature.

[#2#]

Think of it this way. You've got a basketball. Now take a piece of cardboard and cut a hole in it that is about 1/2 the diameter of the basketball. Put 4 marks on the cardboard approx. 90º apart. Now make another piece of cardboard that is exactly the same. Sandwich the basketball between the two pieces of cardboard with it partially sticking through the holes. Hold the cardboard pieces parallel and look at the marks. Those marks are your hits. They all fall on the basketball right?

[#2#]

Is this a cylinder constructed from 2 circles or a measured cylinder?
PC-DMIS compensates for ONLY the last tip in a feature. That can create a source of error.

Good point!

I asked about workplane because I do not know if he is measuring circles or cylinders. I am thinking he is doing hits, calling a tip, then doing more hits the final result of which is a cylinder. I do not know how it comps though when you do that. It may be a comp issue between the 3mm tip and the 4mm tip being used in the same feature.

I see now. So he should do as John says. Measure a circle with one tip, change to the other tip and measure another circle. Then construct the cylinder from the two circles.

[Kevin Watts]

I am trying to measure a cylinder, where i take hits on tip 5 and then hits on tip 3. i changed to 8 hits per circle and still get sphere. i can measure 2 circles and then create a cylinder. but curious as to why it would guess a sphere. the part is a cyinder and i must measure at both ends and this is the only way i can measure for alignment

[craiger_ny]

Measure circles and construct a cylinder. I do not know what happens when you use 2 different size tips inside of one feature but my guess is not good things.

[John Kingston]

Measure circles and construct a cylinder. I do not know what happens when you use 2 different size tips inside of one feature but my guess is not good things.

Craig is right, you don't want to use more than one tip in a feature.
PC_DMIS compensates for the last tip ONLY! Also, on a cylinder (using one tip) you need to stagger hits. I often add a middle row with only one hit. This eliminates the problem Goodluck described. BUT, in your situation, I would use circles and construct the cylinder. Vector points work too. I use them a lot on some features I just can't get at without the star probe. Cut & Paste with Pattern is an awesome tool for this. I can go into more detail if you want, but right now I need to get this part programmed!

[Matthew D. Hoedeman]

MORE levels. 3 levels of hits will help. 2 perfect circles with equal-spaced hits will also make a perfect sphere.

[Kevin Watts]

thanks to all. Using 2 seperate circles seems to do what i need. i am not sure that the 2 different size tips should confuse things since they both calibrate accordingly. but im still a newbie.

[dwade]

I agree with Matt with using 4 hits and 3 levels. I have run into this just last week on progam that had been written a while back using 5 hits with 2 leveles and all of sudden the cylinders were skewed where they shouldn't have been.

If you need a work around you can construct a 3D line from the 2 circles and use variables from the 1st circle and the 3D line to create a generic cylinder.

[dwade]

Actually my mind went to the gutter when I read the post name. I went straight to thinking "this is right up Craiger's line"