Hello Guest it is March 28, 2024, 05:29:50 PM

Author Topic: Stupid Probing Question....  (Read 5793 times)

0 Members and 1 Guest are viewing this topic.

Stupid Probing Question....
« on: January 01, 2010, 11:46:11 PM »
I now have one of those spiffy probes with the little, tiny ball end, and the switch consisting of the three little arms at 120 degrees apart, resting on ball bearings.  In terms of precision and repeatability, it's excellent - I get very consistently +/-0.0001".  But, I have two problems:

1) It's near impossible to get it calibrated to where the ball end is *precisely* aligned to the spindle center.  Getting much under +/-0.0005" is a real chore, and I'm not at all convinced it'll stay there for any length of time.

2) Even worse, the distance the probe tip moves between the time it first makes contact with the workpiece and the time the switch opens is completely unknown, and, I suspect, due to the gemotry of the switch, probably varies based on the angular position of the contact point.

So, how in the heck are these two problems resolved?  Certainly, it would work just fine as-is for probing an object to get a point cloud, but I want to use it for machine setup, so I have to know *exactly* where the spindle centerline is relative to the contact point, as I did when I used to use a rigid probe (which Mach3 liked to break off from time-to-time....).

On a related note, I'm working on a "semi-rigid" probe design that will have a 0.200" diameter tip just like the rigid probes I've used, but will ber able to "give" on over-run by having the probe mounted in a spherical ball bearing, with a spring-loaded centering mechanism.  I think this will be far easier to adjust perfectly concentric to the spindle, and should return precisely to that position after being bumped off-center.

Regards,
Ray L.
Regards,
Ray L.
Re: Stupid Probing Question....
« Reply #1 on: January 02, 2010, 12:39:50 AM »
We had the same problem on an old Mazak.

We drill an hole at the table's corner near the Zero Machine. Hole must be precise to the 0.0001.

If we had to machine very precise part, we made a little program, called Calibrate Probe.

It was written so it could modify the data register in the machine.

I think it could be possible to write some VB functions to update your probing macro.

Hope this help,   Jeff
Re: Stupid Probing Question....
« Reply #2 on: January 02, 2010, 07:40:37 AM »
I also have one of those probes

Believe me i am not knocking it because it is great VFM but it is easy to confuse its accuracy or lack of, unless effort is put into the probe routine.

The test i did after getting weird results was to get a piece of stock with a nice hole in it use a say G55 centre the usual way with a DTI

Zero the offset Then probe the hole using the probe using G54

You can then compare the error between the two methods

If all is well both will read zero, but you may be surprised as to the results

So although i am questioning the accuracy "Out of the box" the repeatability is good

If anyone wants to know how i did it just ask
The Good Thing About Mach3, Is It's very Configurable

The Bad Thing About Mach3, Is It's Too Configurable

Offline Graham Waterworth

*
  • *
  •  2,668 2,668
  • Yorkshire Dales, England
    • View Profile
Re: Stupid Probing Question....
« Reply #3 on: January 02, 2010, 09:31:49 AM »
Most probing routines take 2 or 3 hits per reading, first is a fast traverse to find the object, it then backs off say 1mm (.040") and then slow feeds to find position. For very fine detail it backs off again .25mm (.010") and then creeps to find the position. At this point the final reading is taken. The accuracy is then down to the trigger action of the probe.

Graham
 
Without engineers the world stops
Re: Stupid Probing Question....
« Reply #4 on: January 02, 2010, 11:45:02 PM »
I also have one of those probes

Believe me i am not knocking it because it is great VFM but it is easy to confuse its accuracy or lack of, unless effort is put into the probe routine.

The test i did after getting weird results was to get a piece of stock with a nice hole in it use a say G55 centre the usual way with a DTI

Zero the offset Then probe the hole using the probe using G54

You can then compare the error between the two methods

If all is well both will read zero, but you may be surprised as to the results

So although i am questioning the accuracy "Out of the box" the repeatability is good

If anyone wants to know how i did it just ask

Phil,

I'm trying a similar approach - I took three 1-2-3 blocks, and clamped them in a vise, with the outer two standing on-edge, and the third on it's side, so i have exactly 2" between the two outer ones.  I will then probe that to come up with proper calibration, and see how much variation I get based on the rotation of the probe.  This should tell me what the over-travel is between first contact and switch opening.

Regards,
Ray L.
Regards,
Ray L.
Re: Stupid Probing Question....
« Reply #5 on: January 02, 2010, 11:47:50 PM »
Most probing routines take 2 or 3 hits per reading, first is a fast traverse to find the object, it then backs off say 1mm (.040") and then slow feeds to find position. For very fine detail it backs off again .25mm (.010") and then creeps to find the position. At this point the final reading is taken. The accuracy is then down to the trigger action of the probe.

Graham

Graham,

That is basically what I'm doing - I do a first fast probe (50 IPM), then back off 0.020" and repeat at 1 IPM.  This gives me excellent repeatability to 0.0001".  But I still expect to find considerable variation in the over-travel at different rotations of the probe.  I plan to test this tomorrow.

I *really* want to get a good contact probe going, that does not *have* over-travel....

Regards,
Ray L.
Regards,
Ray L.
Re: Stupid Probing Question....
« Reply #6 on: January 03, 2010, 05:51:53 AM »
Hi Ray

When are you going to release the code. ;D

Phil
The Good Thing About Mach3, Is It's very Configurable

The Bad Thing About Mach3, Is It's Too Configurable
Re: Stupid Probing Question....
« Reply #7 on: January 03, 2010, 01:11:11 PM »
Hi Ray

When are you going to release the code. ;D

Phil

Phil,

Shortly, once I've tested them thoroughly.  I've also added a number of new functions for doing measurement, without clearing the DROs.

Regards,
Ray L.
Regards,
Ray L.
Re: Stupid Probing Question....
« Reply #8 on: January 03, 2010, 01:16:15 PM »
So, the more I learn about how the type of probe I have really works, the less I like it.  It's great for doing surface and outline probing, and measurement - any operation where you don't really care if the tip is perfectly concentric to the spindle centerline - but just seems like a huge hassle for doing machine setup, which is what I'm most interested in.  I'm definitely going to make myself a contact probe, with a Mach3-proof spring-loaded tip.  It will be FAR easier to calibrate, and should maintain calibration much better.  And, it will trigger on first contact, not at some later switch opening point.  I've got a nice, simple design in mind.  Hopefully I can get it built up in the next week or so.  I plan to use a ball-bearing as the spherical probe tip, and a spring pre-loaded spherical ball bearing as the primary support.

Regards,
Ray L.
Regards,
Ray L.