Possible probe issues / what is the XYZ in the probe geometry?

[[Be...]]

We are measuring a part that has several holes going around the perimeter and for two holes on the bottom have to switch to an L-probe.  Really its a star probe that we only put one leg on.  And it works ... until it doesn't.  So it works for a week or a month then it stops working.  We don't have a reason why.  So I get introduced to the problem and my first thought is that the parts are actually bad.  We verified with two separate independent machines that the part is good.  The part is measuring oot in the +Y direction on just the one machine.

So I inspect the program and it seems fine.  No changes.  It is still running fine on other machines and comparing the programs they match.

So I am thinking maybe it is the probe or probe calibration.  They build a new probe from a working set.  Same problem.  We delete the probe and recreate it.  Same problem.

So I start looking closer at the calibration and I guess I don't know what I am looking at.  Forgive me that the probe numbering does not follow convention.  For whatever reason, they made probes 2 and 3 and those are the only orientations that are actually used.  I made the rest of them just to have more data.  Probes 1,2, and 4 are done on one reference sphere location.  Probes 3, 5, and 6 are made on the other.  

So, what does the X, Y, and Z mean? The main stem is maybe 3 inches long and the probe sticks out to the side maybe 1 inch.  So when A = B = 0 the main stem sticks straight down and the probe sticks out to the left in the -X direction.  When A= -90, B = 0, the main stem sticks straight down and the probe sticks out in the -Y direction.  I haven't been able to think of anything that makes sense of the XYZ for any two probe orientations.  

 

 

 

[[Ma...]]

Hello Benjamin,

the X/Y/Z values are the distances to your Masterprobe:

 

Best Regards

Marcel

[[Be...]]

I appreciate the answer, but this does not match the numbers shown in the graphic.  In the first position, it looks about right.  Assuming the numbers are in inches.  But in position 4 the XY values should be remarkably similar but they are not.  Same with position 5.  Rotating the probe 90 or 180 degrees about the Z axis should give predictable results following the idea that you proposed.  Based on the graphic I am guessing that this comes to me from Zeiss through you.

Right now, I have two guesses I am mulling over.

1) the brand new probe that is giving great results in terms of the sigma is actually a very bad probe.  But this appears to be the case with more than one new probe then. So I am disinclined to believe this is the answer.

2) The software has a built in filter that assumes very specific angles for specific probe numbers and it will just give you garbage if you program them at some other angle.

 

[[Ma...]]

How do you do calibration? When i was doing manual calibration, then i had to touch ref. sphere in direction of probe tip - i had difficulities with A45 B60 ( i had to scan right bottom corner ).

Isn't your probe shanking?
Also i would considered lowering dynamics and probing force.

Could you post photo of your probe setup?

Edited 7 hours ago

[[Ma...]]

Make sure the probe is using the correct reference sphere for its associated position or you will get some crazy numbers. This is important as you are using two different spheres.

[[Be...]]

Please sign in to view this quote.

The calibration procedure we are using is pretty generic.  It is basically what they would teach in a beginner calypso class.  I don't know if that counts as manual as I haven't learned any other way to calibrate these probes.

We did switch to a 3mm probe that has more shank clearance just in case, but I don't think it is shanking.  The sphericity, probe diameter, and sigma are all really good.  Maybe I should reduce the angular range of the scan from 180° to 160°? 

I have not tried lowering the probing force and dynamics.  It seems like there is the one pop-up when you create the new probe.  Maybe that is the way to go.  

It just seems so odd that the two features are measuring off about 0.03" in the +Y direction even though the A is 90 for one and -90 for the other while they both have B=-65.  The two orientations have to be calibrated with different reference sphere orientations. We have gotten the same results with a 2mm ball and a 3 mm ball over multiple calibrations.

When the calibration starts we get a popup window asking about the angle.  As far as I can tell we are supposed to enter the angle from the center of the reference sphere to the center of the probe if everything is right where we want it.  It seeds the answer and is usually between 5°and 10° off in the X/-Z and Y/-Z directions.

We have a duplex set up with two reference spheres set up roughly 15" apart on the same plate that get loaded into the machine at the same time.

I would have to ask if I can post a picture as they are pretty strict here.  I would have to get approval to take the picture and then have the picture reviewed before posting it.  It is a big hassle.  

[[Ma...]]

I didn't have any trainings - just what was told and shown to me in roughly 5 weeks.

If you are scanning ( tensor ) then you can not change angular range - it's still force 180°. But passive should work.

All i can think of is either shanking ( which would result at least in wrong probe tip diameter ), loose probe tip or it's not rigid enough for probing force.

Also i am not sure if those numbers are in inch - i think they are in mm's.