Probe Cal Correlation Problem

**John Kingston** · 11-03-2006, 07:09 AM

Hopefully Matt will read this and step in. I think you may need to check to see that your error map from your last calibration is being used. Matt has pointed out numerous times that the "upgrade" to new versions doesn't always bring this over. This is just what my feeble mind can remember. Matt H. would be better at explaining this.

**Wes Cisco** · 11-03-2006, 07:12 AM

You are using a star probe? Is tip #4 shanking out when you calibrate it? Or is tip#1 making contact when tip #4 should be? Are you using a dual sphere artifact? These are my best guesses based on the info you provided. HTH

**Matthew D. Hoedeman** · 11-03-2006, 07:31 AM

Originally posted by John Kingston View Post

Hopefully Matt will read this and step in. I think you may need to check to see that your error map from your last calibration is being used. Matt has pointed out numerous times that the "upgrade" to new versions doesn't always bring this over. This is just what my feeble mind can remember. Matt H. would be better at explaining this.

Well, this upgrade/comp map thing showed up in V3.5, the FIRST version to go to the registry settings. If he was running V3.5 before the upgrade, then installing V3.7 should not be an issue, when I went from V3.5 to V3.7, it didn't turn it off, but installing V3.5 didn't turn it on.

Well, the one question I have is, Did you tell it the ball was in a new place for the FIRST tip and that it was in the same place for all the OTHER tips?

**#2#** · 11-03-2006, 08:39 AM

Originally posted by Matthew D. Hoedeman View Post

Well, the one question I have is, Did you tell it the ball was in a new place for the FIRST tip and that it was in the same place for all the OTHER tips?

This is my first question also. Additionally, are you calibrating A0B0 on all tips. I believe PC-DMIS uses the A0B0 cal of a tip to relate it to other tips.

Try this... Load tip #1 calibrate A0B0 telling pc-dmis that the shpere has moved. Then load tip #4 calibrate A0B0 telling pc-dmis that the sphere has NOT moved. Next, write a quick program to measure the distance between two simple features (two holes in a part) using both tips. Compare the results.

HTH

**cmmuser** · 11-03-2006, 08:52 AM

Answer to some Questions

If I am going to look up to see if the map settings are on, where would I look and what would I be looking for the setting to be set at?
I upgraded from V3.5 to 3.7.
I am not using a star probe. When I refer to Tip# 1, I mean the tip located in slot #1 in my tool changer. Tip# 2 is the tip in slot #2 in the changer, etc.
It does not appear to be shanking on the calibration sphere on any of the tips. Tip#5 is a 1/2mm ball and is one of my “Good” tips (using on a 75° arc on the cal sphere).
I an using only the top ball of a double ball set. It sits I0J45K45 on the table. I don’t believe this type of problem has ever surfaced here before, so it’s either recent, we don’t encounter it often, or “normal” tolerances cover up for the problem most of the time.
When I ran the full calibration (my tips did not come over to V3.7 from V3.5), I went into Probe Utilities, selected Tip# 1 in the changer, selected A0B0, hit measure, “Yes – sphere has moved” and let it go. Then, I selected no tips, hit measure, “Yes – measure all tips” and “No – sphere has not moved” and then ran cal on all 5 tips using “No – has not moved”.

**Wes Cisco** · 11-03-2006, 09:03 AM

What are your probe builds like? Tip size, length, extension material. It could be that numbers 1 & 5 are too heavy/long. Do you get the same results with both probes? You said tp200 and sp600 right?

**cmmguy** · 11-03-2006, 09:11 AM

What brand of CMM also, cause it is handled differently on different cmm's

**cmmuser** · 11-03-2006, 09:36 AM

Answer to some Questions (Rev 2)

Tip# 1 is a 1mm x 20mm carbide shank ruby with a 20mm SS extension. Tip# 4 is a 5mm x 50mm ceramic shank ruby with a 10mm SS extension. These tips have all been “In-Service” for many years in relatively unchanged states, so it is either a recent problem, or always been wrong.
CMM is a Brown & Sharp Xcel 9-15-9 with a TP-200 and a PH10MQ head and an SCR-200 changer.

**JamesMannes** · 11-03-2006, 09:49 AM

Are you using the Sharpe 32z controller? If so, load the "problem" tip and try contacting a surface not normal to the works planes. Something like I=.707, J=0, and K=-.707. When the probe contacts the part watch when the probe is triggering. I too have a BNS 9-15-9 with a Sharpe 32z controller and we do not use the TP200 probes on it anymore. At first we thought the operators were being too rough with the equipment so we switched over to the TP20 probes. One day I had to use the TP200 on the same machine and happened to notice that there was a large delay between the time the probe contacted the part surface and when it actually registered the hit(s). We have not remedied this situation.

**Wes Cisco** · 11-03-2006, 10:17 AM

I use a tp20 and a tp200, both with changers. I run tips similar to what you have describedI . While you might have a tp200 that is on the verge of going bad, 40mm of metal is a lot of weight. I would try making most if not all of that ceramic or carbon fiber. A 5mm ruby on that long of an extension could also give problems. I tried a 4mm x 40mm ceramic tip on the tp200 unsucessfully because of the weight of the ruby generated too much centrifical force at the critical moments of inertia. (I love geekspeak). I have found that while the tp200 is supposed to be able to handle longer tips than the tp20, the opposite is actually true. I use shorter builds on the tp200 and really only use it when I need roundness. It is far to prone to false triggers, even with short tips. I too have a 9-15-9 with sharpe 32. I was told tp200 is really better suited to newer machines with smoother drive systems. Unless the lobing error is a problem for you, I would look into swapping for a tp20 system. HTH

**cmmuser** · 11-03-2006, 10:34 AM

Answer to some Questions (Rev 3)

Controller Model Number is Sharpe 32 180-202-1. At this point in time, we have been using this set-up for approximately 3 years (at least). We, as a company, are unwilling to accept that it’s always been wrong (I hope it’s not the case).

I have entered Probe utilities, selected Probe# 1, A0B0, “Yes – sphere has moved” and run the cal. Then, I have selected probe# 4, A0B0, and “No – sphere has not moved” and run cal. I then wrote the small program to use probe# 1 to check 2 sized holes and found a simple distance between to be .5861”. I then repeated the process with Probe# 4 and found the distance to be .5861”. That check looks pretty good to me, but I’m not taking anything for granted anymore.

I have still not had any luck in finding any information on this comp map issue. Where do I find it? How do I see if it’s on.

**Matthew D. Hoedeman** · 11-03-2006, 10:40 AM

OK, once you have 0,0 calibrated on all the tips, go back and calibrate all the angles in each file, telling it NO the tool has not moved, and they should all be golden.

As for the comp file, it won't work for one and not the other, so that isn't even an issue here, but if you want to see if it is on, you ened to go into the SETTINGS EDITOR and check both of the following:

COMP METHOD
and
COMP FILE

The method should be 14 (I think) and the file would be COMP.DAT for the Sharpe32 controller. I can't look right now, my machine is running and you have to close Pcdmis to look at the settings.

**cmmuser** · 11-03-2006, 11:11 AM

Answer to some Questions (Rev 4)

I have looked into the Settings Editor. Under the Heading “Sharpe32”, I have found a CompensFileName with a value of Compens.Dat and a VolCompMethod with a value of 14. I also ran a search on the hard-drive for Comp.Dat and found a file in the PCDMISW 3.7 folder, but I did not find anything called Compens.Dat anywhere on the hard drive (I even went back to check the spelling).
Are these the things you are talking about, or should I be looking under another heading (there are a lot of heading in the Setting Editor)? Should the Compens.Dat file really list Comp.Dat?

**Matthew D. Hoedeman** · 11-03-2006, 11:27 AM

Originally posted by cmmuser View Post

Answer to some Questions (Rev 4)

I have looked into the Settings Editor. Under the Heading “Sharpe32”, I have found a CompensFileName with a value of Compens.Dat and a VolCompMethod with a value of 14. I also ran a search on the hard-drive for Comp.Dat and found a file in the PCDMISW 3.7 folder, but I did not find anything called Compens.Dat anywhere on the hard drive (I even went back to check the spelling).
Are these the things you are talking about, or should I be looking under another heading (there are a lot of heading in the Setting Editor)? Should the Compens.Dat file really list Comp.Dat?

That would be them, I am not sure about the name. I have heard that the name does not matter, BUT, I would think that the comp name should match the comp name on the computer. The 14 is right.

If your SMA is up-to-date, get on the phone and start screeming at B&S about that file name. I have heard that the name does not matter, but if so, whyinthe****inghell is there a place for the name? What you can do is place 5 tooling balls on your table, one at each corner of your reach and then one in the middle as high as you can. Measure all of them, program them in DCC, run it 10 times or so, saving the stats, then change the name to COMP.DAT and run it all again (10 times). If it all matches, great, if not, well, then you are on your own and it is doubtful if your machine can be called 'calibrated'.

Probe Cal Correlation Problem

Probe Cal Correlation Problem

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Related Topics