Calibrating Disk

[ray7523]

I am new to the board so please excuse my stupidity. We have three disks and they all read undersize. Two read .0004" undersize and one .0005" under. I have no history working with the disk probe so I am at a loss. The probe goes through calibration fine.

The guy I work with says that the probes used to take hits on three layers during calibration and now we can only do two layers. Is that the issue????

We are using cad v4.1 release canidate. Any help would be greatly appreciated.

We found the undersize readings by checking to ring gage

[#2#]

What do you mean "read undersize"?

Are you getting the .0004 and .0005 from the calibration results? Are you comparing this number to the diameter measured with a micrometer?

Don't do that. The value reported in the calibration results is the diameter that the cmm "sees" It is a combination of the actual (micrometer) diameter and the motion of the machine.

Think of it this way. The machine is moving it to take a hit. When the probe touches the sphere (or part), it causes a trigger signal to be sent to the controller. This signal can be generated by an electrical switch or a strain gage or possibly other means. Anyway, it will take a little (very very little) time for the trigger signal to run through the wire to the controller. When it gets to the controller, the controller says to itself "oh, there's a hit" and stops the motor and reads the scales. Now, while the trigger signal is traveling to the controller, the motor is still moving the probe. So, when the scales are read, the probe is actually past the point where it contacted the sphere. The end result is that the probe tip appears smaller to the machine than it actually is. Don't worry, the machine comp. uses this value to correctly compensate.

This is why it is important to calibrate with the same movespeed, touchspeed and prehit/retract distances as will be used in the program for maximum accuracy.

The number you do need to worry about is the standard deviation reported in the calibration results.

[Tested]

Make sure that you are not going to far up or down from center of the sphear. The disk probe is nothing more than a ball probe sliced both sides of center. Going to far up or down will cause you much trouble. I am running 4.1 and am not having a problem doing 3 layers.

[ray7523]

I was comparing the measurements to what I got with a 2x20mm ruby.
The touch speed, prehit, and retract are set to the same as calibration. The move speed may be set a little faster.

What am I doing wrong that I can not get the disk to calibrate in three layers??

thanks again for all input

[#2#]

I was comparing the measurements to what I got with a 2x20mm ruby.
The touch speed, prehit, and retract are set to the same as calibration. The move speed may be set a little faster.

What am I doing wrong that I can not get the disk to calibrate in three layers??

thanks again for all input

I always, always calibrate my disk probes with 10 hits spaced at 2 levels. Never have a problem.

If I understand correctly, you measured a diameter with the disk and for arguments sake got 1.9995 and then you measured it with a 2X20mm probe and got 2.0000.

[Matthew D. Hoedeman]

I was comparing the measurements to what I got with a 2x20mm ruby.
The touch speed, prehit, and retract are set to the same as calibration. The move speed may be set a little faster.

What am I doing wrong that I can not get the disk to calibrate in three layers??

thanks again for all input

It sounds like you need to increase the number of hits it is trying to use. If you have 8 or 9 hits, I think it will only use 2 levels, try 15 hits and see what it does.

[ray7523]

Yes I did get 1.0000 & 1.4000" with the ruby and .9995 &1.3995 with the disk.

Also I have the all three disks set at 15 hits per layer

[#2#]

Yes I did get 1.0000 & 1.4000" with the ruby and .9995 &1.3995 with the disk.

Also I have the all three disks set at 15 hits per layer

Are you measuring with the same number of hits for each tip? Are you measuring manually or DCC? What are the diameters of your disk probes? Typically, the closer the size of the probe is to the size of the hole, the less accurate the results.

What about the rumor that you shouldn't calibrate a disk with more than 10 hits? I heard this somewhere on this board but I don't remember from whom and the reason. Maybe you could try it by using 10 hits on 2 levels and see what you get.

[ray7523]

Yes I am using the same number of hit with all tips, and all measurement data is from running in DCC mode.

I was checking a 12, and two 18mm disks.

I will try changing to 10 hits per layer and see if that helps

Thanks

[cmmguy]

What about the rumor that you shouldn't calibrate a disk with more than 10 hits? I heard this somewhere on this board but I don't remember from whom and the reason.

That has to be urban legend... How would it make a difference?

[#2#]

That has to be urban legend... How would it make a difference?

I don't know. I just remember reading it somewhere. I'll go searching and see if I can dig anything up.

[#2#]

OK, a quick search yeilded two posts related to my comment.

POST 1
POST 2

Looks like John Kugler is my source for this. Perhaps John could weigh in and give us some more info.

[John Kugler]

Gentlemen, (and or ladies),

My suggestion is to change it to 5 hits and make sure you are in default mode and not in user defined. PC-DMIS treats disk probes very differently than regular ball brobes. As such, the calibation is treated differently. Try that first and see if that will work, if it does, change your number of hits to 12. Do NOT do any more than that. For some reason, more than 12 will effect the entire calibration sequence. It may appear to work just fine, but in reality, it is not.

Please note, I do not know if this issue is resolved on 4.0 and above.
Possibly Mr. Mannes may have information on that.

Hope this helps...

[cmmguy]

OK, so it an issue with the calibration routine not working properly and not necessarily with the number of point affecting the calculation. A bug?
thank you

[ray7523]

Gentlemen, (and or ladies),

My suggestion is to change it to 5 hits and make sure you are in default mode and not in user defined. PC-DMIS treats disk probes very differently than regular ball brobes. As such, the calibation is treated differently. Try that first and see if that will work, if it does, change your number of hits to 12. Do NOT do any more than that. For some reason, more than 12 will effect the entire calibration sequence. It may appear to work just fine, but in reality, it is not.

Please note, I do not know if this issue is resolved on 4.0 and above.
Possibly Mr. Mannes may have information on that.

Hope this helps...

I tried with 5 hits in default mode (I was in user defined), and got the same results. I changed to 12 hits in default mode, and still no change to diameter