Wrecked! Need Help!

[DOUG MORTON]

Long story short. When I came-in this morning, I noticed a nick (shiny-spot) on my PH9A probe head. Obviously one of my CMM operators on the off-shifts had wrecked my CMM without telling me. I calibrated the tip at various angles and noticed a distinct problem. I performed the "ink-pen test" ... I've seen the calibration tech's do this before. I placed an ink-pen in one of the mounting holes in the granite plate, rotated head to A90B0 and aligned on the ink pen ball. Then locked X-axis rotated head to A90B180 and moved back to the ink pin ball. I was approx. 0.060 off. I re-aligned the probe head to eye-ball spec. My question is how square do I need to be and how do I get there?

[Matthew D. Hoedeman]

Long story short. When I came-in this morning, I noticed a nick (shiny-spot) on my PH9A probe head. Obviously one of my CMM operators on the off-shifts had wrecked my CMM without telling me. I calibrated the tip at various angles and noticed a distinct problem. I performed the "ink-pen test" ... I've seen the calibration tech's do this before. I placed an ink-pen in one of the mounting holes in the granite plate, rotated head to A90B0 and aligned on the ink pen ball. Then locked X-axis rotated head to A90B180 and moved back to the ink pin ball. I was approx. 0.060 off. I re-aligned the probe head to eye-ball spec. My question is how square do I need to be and how do I get there?

How square? Within a row of apples. That is what the calibration does, it locates the tip in 3-D space around the 'ball' of the head (the pivot point). If you are square by eye (say, within 0.020"(, you are good to go, just calibrate and measure.

HOWEVER, you may want to remove the head from the quill and MAKE SURE that the mounting bolt that holds the head to the taper is still tight, then put it back in and square it up.

[Wes Cisco]

Eyeball is enough if you eyeball is calibrated traceable to National Standards. Yours is, isn't it?

Seriously, eyeball should be ok. Run calibration again and see how it comes out. Hopefully you have on hand a part you ran yesterday, (or before the crash) that you can run again to compare results. HTH

[#2#]

Eyeball should be OK.

UNLESS you do a lot of measurements at angles such as A90Bxxx. In that case, you may find that the off-square causes shanking problems if you set your part up square to the machine axes.

Probe tip calibration will comp. the off-square situation but it won't do anything for shanking out in holes. You will either need to square the part with the probe or square the probe with the machine axes. I would do the latter as it is easier to square the part with the machine axes than with the probe.

[Hilton Roberts]

Long story short. When I came-in this morning, I noticed a nick (shiny-spot) on my PH9A probe head. Obviously one of my CMM operators on the off-shifts had wrecked my CMM without telling me. I calibrated the tip at various angles and noticed a distinct problem. I performed the "ink-pen test" ... I've seen the calibration tech's do this before. I placed an ink-pen in one of the mounting holes in the granite plate, rotated head to A90B0 and aligned on the ink pen ball. Then locked X-axis rotated head to A90B180 and moved back to the ink pin ball. I was approx. 0.060 off. I re-aligned the probe head to eye-ball spec. My question is how square do I need to be and how do I get there?

I prefer to make sure my probe is mechanically square to the machine. Many times I use a 0/0 probe to "eyeball" a part parallel to the machine and then use dcc alignment in the part program.

If you have parts with holes in the sides that require a 90/90 probe angle for example and your probe is reaching fairly deep into a bore or small counterbore, you can run into probe shanking. If the probe is out of square and you shank, the error will show up in the Y axis or the X axis if using for example a 90/180 angle.

If you have had a serious crash in Z, you would probably see the change just by eyeballing the probe. Your error would show up in the Z component of the hole location.

You can get very good results squaring up your probe in the X/Y plane by setting a right angle iron on the plate and "tramming" which ever face you choose to use parallel to either X or Y axis.

Once it is parallel ( assume we are parallel to X axis here ) change the probe angle to 90/90 and set a zero on the Y face. Get the probe clear and set the angle to 90/-90 and measure the Y location on the opposite side of the angle plate.

Rotate the probe to take out one half of the error and repeat as often as required to get the readings from each side of the angle plate to within .001 or so. It IS easier than you might think.

You should be good to go.

Try to use an extension as well as a probe cartridge and stylus. The longer length you have will make it easier to see the difference in the probe pointing in the two directions although it will make it a little harder to get the squareness right on. You can take that stuff out for your regular tool configuration.

Making everything as square and parallel as you can get it will help control some of the variables you will see in everyday part measurement.

Where you do not want to be is in the position of explaining why a part brought in for reinspection of some disputed feature does not repeat.

Fortunately for me, I am the only one who uses my machine. If I were you, I would have a test part to measure periodically ( you KNEW I was gonna get that in here....right Matt??? )

That way if I saw something on my machine that indicated a crash, I would throw that part up on the machine and measure it to be certain nothing had changed.

HTH

Hilton

[Matthew D. Hoedeman]

Fortunately for me, I am the only one who uses my machine. If I were you, I would have a test part to measure periodically ( you KNEW I was gonna get that in here....right Matt??? )

Yeah, I was pretty sure, but YOU DIDN'T BACK ME UP on making sure that the screw was still tight. Does not do any good to get it sqaured back up if the screw is a little loose, the head can than move and you wouldn't know it. Oops, there goes the credibility.

[Hilton Roberts]

HOWEVER, you may want to remove the head from the quill and MAKE SURE that the mounting bolt that holds the head to the taper is still tight, then put it back in and square it up.[/QUOTE]

WMS

HOWIZZATMATT

H

[Matthew D. Hoedeman]

HOWEVER, you may want to remove the head from the quill and MAKE SURE that the mounting bolt that holds the head to the taper is still tight, then put it back in and square it up.

WMS

HOWIZZATMATT

H[/QUOTE]

Much better, now go check your 'master' part!

[Cumofo]

No credibility here?

WMS

HOWIZZATMATT

H

Much better, now go check your 'master' part! [/quote]

What is an acceptable part variance?
You mean a GR&R?
Do I need three appraisers?
What would be the acceptable equipment variation?
Do I use ANOVA?