Getting a Star probe square to your machine

I stumbled upon this thread sometime back when looking for information about this subject. I don't have a Rayco plate (The boss vetoed getting one when she saw the price), and I have no idea what a 123 block is. Anyway, I got to thinking about how I could do something similar with the setup I have.

I set up my star, but only qualify the "straight down" tip to start with.

What I did was to clamp a parallel in a vise. I have some pins screwed into the table and I put the vise up against the pins. However, this only gets things in the ballpark as the inserts in my table are pretty banged up.



I have a short program I wrote that helps the operator square the parallel with the table. You park the probe 1/4 inch out from the left front corner of the parallel and it takes a read point. It then takes a hit on the top and left end of the parallel and uses those in a new alignment to take a hit in the center of the parallel with the ball just barely below the top face. This point becomes the new center of alignment.

Two hits are taken, one on each end of the parallel. These are dimensioned from the machine's alignment in the X axis. I pull the MEAS value from the first dimension into a variable, then use that variable as the nominal for the dimension for the second hit. Then I pull the DEV value from that dimension into another variable and do some IF/THEN/ELSE stuff to see how big the difference between the two hits is. I figure that if the parallel is square to the table within .02, that is good enough.



Once I have it square, I use the end of the parallel as a guide to eyeball squaring up the star tip. Then I qualify everything.

Attached is the routine to square the parallel, in case any body is interested.

Anyway, I am just curious to know if there are any flaws in this methodology...

I stumbled upon this thread sometime back when looking for information about this subject. I don't have a Rayco plate (The boss vetoed getting one when she saw the price), and I have no idea what a 123 block is. Anyway, I got to thinking about how I could do something similar with the setup I have.

I set up my star, but only qualify the "straight down" tip to start with.

What I did was to clamp a parallel in a vise. I have some pins screwed into the table and I put the vise up against the pins. However, this only gets things in the ballpark as the inserts in my table are pretty banged up.



I have a short program I wrote that helps the operator square the parallel with the table. You park the probe 1/4 inch out from the left front corner of the parallel and it takes a read point. It then takes a hit on the top and left end of the parallel and uses those in a new alignment to take a hit in the center of the parallel with the ball just barely below the top face. This point becomes the new center of alignment.

Two hits are taken, one on each end of the parallel. These are dimensioned from the machine's alignment in the X axis. I pull the MEAS value from the first dimension into a variable, then use that variable as the nominal for the dimension for the second hit. Then I pull the DEV value from that dimension into another variable and do some IF/THEN/ELSE stuff to see how big the difference between the two hits is. I figure that if the parallel is square to the table within .02, that is good enough.



Once I have it square, I use the end of the parallel as a guide to eyeball squaring up the star tip. Then I qualify everything.

Attached is the routine to square the parallel, in case any body is interested.

Anyway, I am just curious to know if there are any flaws in this methodology...