Auto Circle Inverting Bug?

[derekvegeta]

This has happened on a couple of other programs and I have never been able to pin down why it's happening: When measuring a radius (>/= 90 Deg) My direction and start/end angle changes without warning and even the angle changes. Its very frustrating as some of these radii have very limited access or are themselves only 90 deg (so I need to be probing the correct 90 deg. span of the theoretical circle). When it happens, it seems to effect at least all of the partial circles (I didn't notice anything on my full circle measurements though since the direction and start/end angle aren't critical so long as I get 360 deg.) My vectors do not appear to change. Visually, it often looks mirrored about my angle vector BUT that could just be my imagination.

I had not made any alignment changes (to the programs) and the program has executed as initially programmed at least once in all cases.

It appears to be happening from resetting measured values to nominal: However when I tested it, it didn't consistently do it (sometimes it would show mercy to some of my circles)

This time around it occurred 2019R2, Build 362 (SP3), based on my fuzzy memory the previous instances were 2019 R1 or 2018 R2.

[JEFMAN]

I also saw it on auto ellipse...
Don Ruggieri , any idea ?

[Don Ruggieri]

T.......When measuring a radius (>/= 90 Deg) My direction and start/end angle changes without warning and even the angle changes. .......

Just to be clear, it happens on circles (and ellipses?) greater than 90 degrees and less than 360 degrees?

[derekvegeta]

Greater than or equal to 90 Deg. but Auto Circles only for me; I have not used ellipses. In some cases there were negative Degrees being used (but the difference between start and end was still 90 or greater, less than 360)

[Mike Ruff]

I have had this happen to me on older versions (I don't remember what version though- I want to say 2010). My workaround was to use vector points and a constructed circle to force the hits to be in the same spot. All 6 DOF were locked down, and the circle in question was in the YZ plane measured with T3 of a star tip (it was an undercut from a turned part), and had about 100° of arc.

[derekvegeta]

Did/do you use Reset Measured Values to Nominal at all?

[Tater]

Make sure the void detection is not on.

[vinaykumar.sandhu]

Hi derekvegeta ,

I tried to reproduce the issue in 2019R2 build # 362 and in prior versions as well, but was not able to duplicate the change in direction and start/end angles issue. I have tried reproducing the issue by creating various circles/ellipses with different combinations(inner/outer, CCW/CW, with sample hits, w/o sample hits etc) but still had no luck. I guess having some supporting files like registry settings, .PRG and .CAD would really help in reproducing the issue.

[derekvegeta]

The CAD is ITAR Restricted so I cannot provide that. However, I can share my code here or send the program somehow (without CAD). I would need help with getting you registry settings however. Alternatively, a team viewer session could work where I show it happening.