Calibration, all-tip-test good, but golden unit is out of print

[VONDY]

Last week our CMM 'Z' axis bumped the side of a part as it was running. Didn't seem to hit it very hard though. The PH10MQ is about 6 years old and has been becoming noisier and nosier as it adjust to a new wrist angle. Well after the incident we ran our quick "all-tip-test" which checks 6 different wrist angles with all 9 of our tips. It showed bad. So we than ran the calibration of tips through, our first tip #1 had about 25 out of 145 wrist angles showing "standard deviations exceed calibration tolerance" It is set at .00098 and they where saying the deviation was up to .00150. We ordered a new PH10MQ. We were able to calibrate all the tips and when we ran our "all_tip_test" it all showed OK. Yet when we ran a golden unit part through the program, it showed a deviation of .0210" location, we ran the same part on the other CMM and had .0009". We created three new tips and calibrated them, inserted them into the program and ran the golden unit part through again, same results.

Any thoughts on what I should try next?

[craiger_ny]

Going on the assumption that it is the same part program can you use the probe and module (if it applies) on the other machine too? Perhaps you have a problem with the probe body itself. Are you using the exact same stylus from one machine to the other. Make sure you are only introducing one variable at a time.

[VONDY]

Going on the assumption that it is the same part program can you use the probe and module (if it applies) on the other machine too? Perhaps you have a problem with the probe body itself. Are you using the exact same stylus from one machine to the other. Make sure you are only introducing one variable at a time.

Both machines share the tip #5, but have seperate tips #2 and #3 that are used in program.

[Wes Cisco]

I doubt you want to hear this, but it sounds like you may have knocked your z axis out of alignment. I think you need a calibration. HTH

[craiger_ny]

You do take the probe body off the "bad" machine and install it on the "good" machine as well when you run through this test of sorts right? I'd start with the whole 9 and take the PH10 with everything on it as built from one machine to the other and see what resuls you get. What I am getting at is you want to do your best to make sure that only one thing at a time is different from one machine to the other (as best as you can, I do understand that you can't go all the way with this). You will go through a long trouble shooting process if you have too many variables introduced.

Craig

[VONDY]

You do take the probe body off the "bad" machine and install it on the "good" machine as well when you run through this test of sorts right? I'd start with the whole 9 and take the PH10 with everything on it as built from one machine to the other and see what resuls you get. What I am getting at is you want to do your best to make sure that only one thing at a time is different from one machine to the other (as best as you can, I do understand that you can't go all the way with this). You will go through a long trouble shooting process if you have too many variables introduced.

Craig

One machine has a PH10M and the other is a PH10MQ. We do not interchange these ever. We have tips #4-7 that share between the two machines and the tips #1-3, 8-9 both machine have separate tips. The tips are built from the PAA1 all the way down to the stylus.

I guess what I am having a hard time understanding is how you can calibrate all your tips, run a tip test through all nine tips ( this process has worked for 7 yrs ) and all show good, yet when we run a part, any part that has any type of wrist change, it shows out by about .020". This should easily show up on our tip test which is set a .001".

[craiger_ny]

All I am saying is use as much common items as you can during this test. If you can't switch PH10 then switch as much as you can. Run the part on one machine. Unscrew the probe body from it and screw it into the other PH10 and run it there, then when you get bad result on one machine and good on the other with all of the same hardware you will find out that Wes is right and it is probably your Z rail. I am not just talking about tips. If the probe body is bad on one machine and good on the other you will get what you are seeing. Just as if the Z rail is out on one and in on the other you will also get what you are seeing.

All I am saying is trouble shooting is very difficult with multiple variables. It can be done as there is a mathematical way of doing it that I could explain but I get the feeling it would take more posts and time to do that than it would to unscrew the probe body from one machine and screw it into the other thus insuring that you are using the same physical pieces of hardware during your test. Wes is probably right but the diference between a bad Z rail and a bad probe body is that you don't need a tech to come in for a bad probe body.

If you are using the same hardware from machine to machine then I'd put money on Wes. If not it is a crap shot.

[JamesMannes]

Just a thought...
You don't happen to have the probe name shared by both machines(e.g. on a network) do you? If you do that is a big no-no.

[Jeffrey Lovely]

Also did you square up the head

[VONDY]

Also did you square up the head

Squared to within .0008" - And both machines use there own named probe tip folders\files completely separate of each other.

We are going to get a Tech in to look at it and probably laser\calibrate it. But I'll let that up tp them in what they do.

Hopefully its something I'm missing, but so far all the suggestions don't apply as of yet.

[tking]

I think perhaps Wes is right. You can calibrate probes just fine on an out of square machine. Your program will run fine but the whacked out axis alignment will skew your measurements. I can give you some info on how to check this if you like. E-mail me if you want more info.

TK

[Joe Zastrow]

[quote=VONDY]Squared to within .0008"



How do you check this

[VONDY]

[QUOTE=Joe Zastrow]

Squared to within .0008"



How do you check this

I use a 200mm extension and place the wrist angle at A0B-90. I then use an indicator and release the "Y" & "Z" and run the head back and forth along the "X" axis and as I lightly loosen, I'll adjust the head slightly and re-tighten the head until I have it under a .001".

I'm starting to think that the air bushing washers may have been damaged as the "Z" rail has some slight movement to it that it didn't use to have.

We have a Tech coming in on Monday to take a look and I'll let him decide, i already spent $$$$.$$ on a RBE head, no use in blowing any more.

Thanks for all the advice and help,

Vondy