Need Help!- Probe macro to find flat on A axis

[metlshpr]

What im doing is screwing a sqare tube onto a threaded shaft fixture an the A axis and would like to make a macro using the probe that would find the flat surface and set my G54 A cord. to that plane. Any help would be great.

[Geof]

Donkey Hotey where are you?

Or Helicopter John and others.

I have to admit I am less than no help when it comes to probing but you have come up with an application that I could occasionally find useful. Setting the A axis to a particular orientation is sometimes tedious and sometimes a fixture has to be more complex to get the correct part alignment; your approach could simplify the fixture design if the parts could be held at random orientations with the machine sorting it out.

[HuFlungDung]

Ideas: ask the operator for the diagonal diameter of the part.
Then, command the A axis to zero.
Devise a 2 vector probe sequence based on the following:
-calculate the Y offset required for a touch off at 45 degrees from top dead center. Move the machine to that Y.
-have the probe descend in Z, looking for a touch off point. Store the point.
-do a mathematical comparison to determine if the touch off point occurred within a given window of ranges between R (the radius of the part diagonal) and (R-R/5). You would do this in case the probe came in directly on the corner, or came in on top of a rounded over corner. If the point was in that region, you'd want to reject it, and index the A axis 10 degrees and start over.

-retract the probe to safe height, then offset in Y to the other side of centerline. Probe for the point.

Calculate the arctan of the angle of a line drawn between the collected points and the horizontal plane.

If the arctan is > than 45 degrees then add the derived angle to the current A position and store it in G54 A.
If the arctan is < than 45 degrees, then subtract the derived angle from the current A position. (or I might have the directions inverted, but a trial run should quickly debug that).

Exactly how you'd write that up, I don't know

[Donkey Hotey]

I have not done a single custom probe application yet. I'm eagerly watching these posts and getting ideas. My problem is the syntax of the macros. I get that it has functions. I can read somebody else's code and see abs[#501-5] and know what it's doing but I don't know where to get a list of all of the functions available and what form it wants them in (Haas really needs to publish a decent Macro manual--hint...hint).

My guess would be that you would need to manually index the A-axis to be visibly close, probe two points, trig the height difference into an angle, adjust the A-axis, then verify (probe again). That's kind of how all the other work offset routines work except they don't check their work (the second probing).

I've been thinking about this same thing for determining the 5-axis origin of my T5C. I wanted to use a 5-C with a tooling ball but I can't get my brain around how to find the ball in 3D space, without knowing the position.

I have a solid 5C disk (4") that I bought for that purpose. I figure that the A-axis method above would be the first step (to find horizontal), then it would kick the axis 90 degrees, approach from one side, kick the axis -90 and do the other side. Divide that in half and you have the axis offset distance.

[Geof]

Originally Posted by Donkey Hotey ..... (Haas really needs to publish a decent Macro manual--hint...hint).

Have you searched through the Manual pdf file on the Haas website? I have a copy and it has all the Macro stuff between pages 80 and 111 or something like that. It seems pretty complete and I was able to puzzle through okay for the one and only macro I have ever written.

I read the two procedures you and Hu propose and wonder if it could be done somewhat easier in this particular situation.

The part is square which means no matter what orientation it pulls tight at there is always a flat facing up in some manner. My thoughts are that you could have a macro in which you enter a Y value for the width of the square then probe at plus and minus this value divided by 1.5 to make sure you don't run off the edge. Then it would be a matter of trigonometry to figure out how much to rotate the A axis and which way, to bring the flat level. Maybe do a repeat for better precision then writew the A axis position in machine coordinate into the G54 A register.

Maybe I am missing something because I don't do this sort of thing.

[omy005]

I have attached a few pages out of the Renishaw manual for HAAS machines that deal with this procedure.

Andrew.

[Geof]

Originally Posted by omy005 I have attached a few pages out of the Renishaw manual for HAAS machines that deal with this procedure. Andrew.

I feel vindicated.

[HuFlungDung]

The gist of my method was to allow a brain dead operator to probe the part without paying any attention to the start orientation of the part.

[Geof]

Originally Posted by HuFlungDung The gist of my method was to allow a brain dead operator to probe the part without paying any attention to the start orientation of the part.

Yes, I was taking advantage of the symmetry of a square part, my approach and actually the approach in the pdf is not general in application; the brain dead guy would have to be awake enough to rotate the single flat to the top.

But I do have to be honest and confess that this is an application for which I would be inclined to get a probe package for.

[Donkey Hotey]

Originally Posted by omy005 I have attached a few pages out of the Renishaw manual for HAAS machines that deal with this procedure.

Don't anybody go getting too cocky about that. That's from the Renishaw Inspection Plus manual. I can say for sure that my 2004 VF-2, with the factory installed probing and software did not have most of the programs described in those pages (even though I was given those manuals when I bought it).

I'm also not optimistic that the newer IPS-driven probing routines include it either. I think programming it on your own may still be the only option for most users.

[mt92]

Did anyone get this to work yet?