Edge Points in DataView Not Reporting Directions Correctly

[Jeff Hunt]

I have been having issues with edge points in Dataview. It seems that the automotive deviation letter is wrong most of the time. First I would measure an edge with 1 sample hit & then report on the T value, but if the sample surfaces deviation was greater then the edge's it would use the surfaces automotive deviation letter in the dataview report. What sense this makes to do this, I have no idea, but tech support is aware of it & it dosent seem to be an issue with them.

Now they have me measure the edge point the same way, but I force the report direction in the variable set with a .!Z identifier in the variable. Now it sometimes reports a point that is down in car position as being up. I am so sick of this **** ****** up software that is never updated that I might switch to something else soon. Does anyone know of a software package that works like dataview, but actually works?

[Matthew D. Hoedeman]

Have you tried using the GAP ONLY mode in Pcdmis? If you are NOT reporting the surface deviation (S) of the edge point, then by using this option, it makes the edge point like a snap point is for the surface, the actuals are projected onto the nominal vector line perpendicular to the edge check direction. This would make it appear that there is no surface deviation, but if you are only reporting the T and not the S, this would be the best method for you to use.

[Jeff Hunt]

I'll look into that, but it still makes no sense as to why dataview would screw that up.

I'm looking into this CM4D software now...Anyone have any experiance with it? It looks alot more robust & user friendly then Dataview.

[gsepan]

Jeff,

I too have had the same problems with edge points and dataview. I was told by tech support that it has to do with the vectors being close to 45 degrees i.e. I = .7, J = .7, K = 0 or something to that effect. That confuses the software and causes it to report wrong. No word on if it will ever be fixed as they never called me back with a solution. Go figure!!!

[Jeff Hunt]

Ok, I dug a bit deeper...When I report on the deviation of a sample of 5 in the normal mode, everything comes out right. But once I switch to chart mode, it gets screwed up...Who can explain this?

Heres the proof:

[Jeff Hunt]

Bump.