Z-Problems Code

[R.Alward]

I wrote the following problem I am having a few days ago and was not able to get back with the requested information. I have attached the manual and CDD alginments. I am using 4.2 version.


I wrote a program for a flat round disc with a center hole. I used two standoffs with threaded holes for alinement then moved the origin to the center hole of the blank. The first 6 blanks ran find then the cmm had trouble finding the first hole on the alinement. It would stop just short of surface when checking the plane or it would crash into the surface. I hit PF9 and redid the hits on the hole and ran test they worked ok. Program asked if I wanted to change the alinement with the data from the test, I hit ok. The program ran ok for two blanks then it did it again. When it crashed into the surface when doing the plane checks I raised the probe back up then it took off in another directions and hit the table. The next time this happened I grabbed the job box and tried to run the probe up as it ran along the table. If I let it go it would go down and crash. Can anyone tell me what is going on and how to stop it? I shut the system down several times and restarted. It did the same thing after running the first blank ok. The last time I tried this it crashed on the second hole during the alginment.

[JamesMannes]

Are you running the entire program when this happens. What I mean is this..Do you measure the manual alignemt and then have the progarm continue on htrought the dcc alignment?

Also, What type of machine/comtroller?

[Matthew D. Hoedeman]

I wrote the following problem I am having a few days ago and was not able to get back with the requested information. I have attached the manual and CDD alginments. I am using 4.2 version.


I wrote a program for a flat round disc with a center hole. I used two standoffs with threaded holes for alinement then moved the origin to the center hole of the blank. The first 6 blanks ran find then the cmm had trouble finding the first hole on the alinement. It would stop just short of surface when checking the plane or it would crash into the surface. I hit PF9 and redid the hits on the hole and ran test they worked ok. Program asked if I wanted to change the alinement with the data from the test, I hit ok. The program ran ok for two blanks then it did it again. When it crashed into the surface when doing the plane checks I raised the probe back up then it took off in another directions and hit the table. The next time this happened I grabbed the job box and tried to run the probe up as it ran along the table. If I let it go it would go down and crash. Can anyone tell me what is going on and how to stop it? I shut the system down several times and restarted. It did the same thing after running the first blank ok. The last time I tried this it crashed on the second hole during the alginment.

OK, one thing I saw was that your DCC circles do not have a perfect vector (0,0,1)

Another thing, I do NOT use the plane feature, I would measure single vector points and construct a plane

I would also construct a line between the 2 circles and use that for rotate.

I would also say you have a few too many alignments.

AL-1= manual features, plane, cirlce, clicle, constructed line, level to plane, rot to line, origin to cirlce and plane
DCC Mode
AL-2= DCC features, 4 (or more) vector points, construct a plane, cirlce and circle, construct a line, level plane, rotate to line, origin to circle and plane

Also, give these things SQUARE vectors. If they should be square, make them sqaure, get rid of those odd vectors, make them 0,0,1

Do more in your alignments, level, rotate and origin in them, don't level, rotate, origin in one, then just level in the next, then just rotate in the next, etc.

[Underspec]

OK, one thing I saw was that your DCC circles do not have a perfect vector (0,0,1)

Another thing, I do NOT use the plane feature, I would measure single vector points and construct a plane

I would also construct a line between the 2 circles and use that for rotate.

I would also say you have a few too many alignments.

AL-1= manual features, plane, cirlce, clicle, constructed line, level to plane, rot to line, origin to cirlce and plane
DCC Mode
AL-2= DCC features, 4 (or more) vector points, construct a plane, cirlce and circle, construct a line, level plane, rotate to line, origin to circle and plane

Also, give these things SQUARE vectors. If they should be square, make them sqaure, get rid of those odd vectors, make them 0,0,1

Do more in your alignments, level, rotate and origin in them, don't level, rotate, origin in one, then just level in the next, then just rotate in the next, etc.

Having alignments in steps is good practice to allow fine tuning of the next dcc measurement. But you should always follow all that up with a final complete alignment.

[R.Alward]

Thanks

To answers some of your questions.
Machine : B&S Gobla Status. Not sure of controler. Installed in 2004.
I do not run the manual alignment after I write the program.

One other thing I should mention. I have use this type of algnment for the past two years with no problems. I will clean up the vectors, in this case I got in a hurry to get the program written. But I still do not understand why the computer can not go back to the same location on on the standoffs when checking the holes. It stops just above or tries to go down too far. Is the fact that I did not clean up the vectors casue this?

[Roberto]

To answers some of your questions.
Machine : B&S Gobla Status. Not sure of controler. Installed in 2004.
I do not run the manual alignment after I write the program.

One other thing I should mention. I have use this type of algnment for the past two years with no problems. I will clean up the vectors, in this case I got in a hurry to get the program written. But I still do not understand why the computer can not go back to the same location on on the standoffs when checking the holes. It stops just above or tries to go down too far. Is the fact that I did not clean up the vectors casue this?

Your manual alignment is not right. There is no origin for X&Y.
Double check it.

[JamesMannes]

The only thing I see right now is that in your first alignment you do not align to a specific hole, you have it still in machine cooridnates. I don't know that that will screw it up but I always doa complete manual alignment. I would define a complete manual alignment as...

1. leveled
2. rotated
3.translated in all 3 axis'
4. any rotation translation to be completed at this point

[#2#]

Your manual alignment is not right. There is no origin for X&Y.
Double check it.

The only thing I see right now is that in your first alignment you do not align to a specific hole, you have it still in machine cooridnates. I don't know that that will screw it up but I always doa complete manual alignment. I would define a complete manual alignment as...

1. leveled
2. rotated
3.translated in all 3 axis'
4. any rotation translation to be completed at this point

I too noticed this. If nothing else when doing a manual alignment, you need to set the orgin in all 3 axes. If you don't, your parts will ALWAYS have to be located in the same spot on the machine or you will run into problems.

I often times just do a point and a circle for a manual alignment (set all 3 axes) and then go on to a more detailed dcc alignment.

[Roberto]

OK, one thing I saw was that your DCC circles do not have a perfect vector (0,0,1)

Another thing, I do NOT use the plane feature, I would measure single vector points and construct a plane
I would also construct a line between the 2 circles and use that for rotate.

I would also say you have a few too many alignments.

AL-1= manual features, plane, cirlce, clicle, constructed line, level to plane, rot to line, origin to cirlce and plane
DCC Mode
AL-2= DCC features, 4 (or more) vector points, construct a plane, cirlce and circle, construct a line, level plane, rotate to line, origin to circle and plane

Also, give these things SQUARE vectors. If they should be square, make them sqaure, get rid of those odd vectors, make them 0,0,1

Do more in your alignments, level, rotate and origin in them, don't level, rotate, origin in one, then just level in the next, then just rotate in the next, etc.

Hey Matt,
Is there a specific reason why you use vector points (constructed plane) instead of plane feature? If correct vectors used both should work fine, I'm I right?

Thanks,

[djayne]

Hey Matt,
Is there a specific reason why you use vector points (constructed plane) instead of plane feature? If correct vectors used both should work fine, I'm I right?

Thanks,

I wrote 6 manifold programs a while back in off-line mode and used four seperate auto planes on the four ports of the datum surface and created one plane out of the 4 autoplanes. When I did the alignment while inspecting the part, it skewed everything about 0.200mm for some reason. I had a compound rotation to measure the outlet flange and this really threw it out of whack. I changed the alignment to use vector points and constructed a plane out of those points and everything lined up correctly. Go figure. I had to go back and change all six programs for this - which was on six different types of manifolds but I had used the same method on all for the datum surface origin.

[Roberto]

I wrote 6 manifold programs a while back in off-line mode and used four seperate auto planes on the four ports of the datum surface and created one plane out of the 4 autoplanes. When I did the alignment while inspecting the part, it skewed everything about 0.200mm for some reason. I had a compound rotation to measure the outlet flange and this really threw it out of whack. I changed the alignment to use vector points and constructed a plane out of those points and everything lined up correctly. Go figure. I had to go back and change all six programs for this - which was on six different types of manifolds but I had used the same method on all for the datum surface origin.

Thanks for your comment djayne,
I personally won't construct a plane out of 4 separate autoplanes for align.

I was referring to single planes (whether plane feature or constructed with vector points).
I use both depending on the setup (i.e. when print calls for datum target points I use vector points).
However, if correct vectors used both planes should work fine.
My opinion.

Thanks again!