Another TP problem

[Kevin Watts]

pcdmis 3.5 mr 1
ph10t
tp20
good morning, i have a question i cant seem to find the answer to. please see attached drawing. it is the screen capture from graphics display. My alignment is level y minus to plane 2. rotate cyl 1 to x plus about y minus, origin y to plane 2, origin x and z to cyl 2. My problem is when i measure 2d distance from plane 2 to circle 2 parallel to y axis i get 7.3093. when i measure TP of circle 2 using only y and z as datums i get 7.3188 as y distance which is hrowing TP out to .018. ive noticed this difference many times but only .0003 to .0015 difference. this .009" difference is throwing up a major red flag to me that im doing something wrong. if my y axis is cyl 2, and in tp cyl 2 is primary datum, why would there be any difference in the calculation of the distance. using perp to centerline makes no difference in readings. i have made sure i am in the correct workplane ( x plus)thanks in advance for any help.

[#2#]

Well, I just have a guess.

Perhaps your circle's vector is not perpendicular to the y axis.

When dimensioned one way it slides along the plane until a line perp to the y axis passes through the circle's center. Then it measures that distance.

When dimensioned the other way, perhaps is moves up the y axis to the point where the circle's vector crosses the y axis. Then it measures that distance.

Was the circle measured using the alignment it was measured in?

[Kevin Watts]

yes it was measured using its own alignment.

[#2#]

I'm out of guesses.

[Underspec]

My guess is it's your alignment. I usually measure two circles, construct a 3d line through it. But it should be the same as using a cylinder.

Level Y / Origin Y to plane.
Origin X,Z to Circle1.
Rotate Line to Y about X
Rotate Line to Y about Z
Origin X,Z to Line

Dimension locations for both circles and they both should be 0,0. If not, then something went wrong. Also, make sure when rotating you rotate to the matching vector of the line otherwise it will flip the origin.

[Jan d.]

Sorry to be totally unhelpful, but I have a very similar call in V4.1/2. I pretty much follow your set-up and get the right result.

Be aware that you can NOT just split the dimension in 2 individual calls. They must apply both at the same time. I never used anything before V4.1. It may be that your version is not up to this task? This did NOT work properly in V4.1. They fixed it for V4.2. It seems to be functional now.

Hopefully this is somewhat helpful...


Jan.