Which would you trust more?

[Marshal]

Ok, now that I feel like an idiot...

 

I finally managed to get the spindle to rotate with the probe in the spindle (remember, I'm a newb at this). And the probe is most certainly not centered. Considering I can actually see that it's not spinning about the center of the spindle. Now I suppose I'm going to have to figure out how to fix that issue....

[robk]

Have you checked Renishaw?

[Marshal]

Have you checked Renishaw?

 

many times, and I've read through the manual a half a dozen times. I think I've figured it out now though, we'll see in a few minutes.

[Guest CNC Apps Guy 1]

Ok, now that I feel like an idiot...

 

I finally managed to get the spindle to rotate with the probe in the spindle (remember, I'm a newb at this). And the probe is most certainly not centered. Considering I can actually see that it's not spinning about the center of the spindle. Now I suppose I'm going to have to figure out how to fix that issue....

 

NO need to feel that way. If you've never done it, you've never done it. You're not even sure what to look for. On the control side of your machine I know nothing, but the mechanical side, I can give youa few pointers....

 

There shoudl be 4 screws around the holder (not the probe body), break them all loose. Get a .0001" indicator, get the probe tip centerd on the ball in X/Y and Z with one of the screws in line with the indicator. Zero out the dial on your indicator. Spin the probe 180. tighten the screws a little getting it to 1/2 the amount of runout. Spin and repeat until you get it to better than .0003. Use 2 allen wrenches when tightening because you may have to tighten one and loosen the opposite one.

 

It took me about an hour the first time I ever did it, but nobody ever told me what to do.

 

HTH

[Marshal]

NO need to feel that way. If you've never done it, you've never done it. You're not even sure what to look for. On the control side of your machine I know nothing, but the mechanical side, I can give youa few pointers....

 

There shoudl be 4 screws around the holder (not the probe body), break them all loose. Get a .0001" indicator, get the probe tip centerd on the ball in X/Y and Z with one of the screws in line with the indicator. Zero out the dial on your indicator. Spin the probe 180. tighten the screws a little getting it to 1/2 the amount of runout. Spin and repeat until you get it to better than .0003. Use 2 allen wrenches when tightening because you may have to tighten one and loosen the opposite one.

 

It took me about an hour the first time I ever did it, but nobody ever told me what to do.

 

HTH

 

Thanks for the pointers. I figured that part out from the manual. The calibration cycle completed and the numbers look a LOT better. Tried to check the ring gauge and now the damn battery is low! One thing after another lol

[Joe788]

DON'T calibrate the XY offset on the ring gage, even if you're confident that you've indicated it precisely.

 

Mill a 1 inch square boss on the top of a good chunk of material. Calibrate the XY offset on it immediately after.

 

Once you've got the XY offset calibrated, THEN calibrate the stylus diameter on your ring gage.

[Marshal]

DON'T calibrate the XY offset on the ring gage, even if you're confident that you've indicated it precisely.

 

Mill a 1 inch square boss on the top of a good chunk of material. Calibrate the XY offset on it immediately after.

 

Once you've got the XY offset calibrated, THEN calibrate the stylus diameter on your ring gage.

 

Really? Even though the manually explicitly says to calibrate it off the ring gauge?

[Joe788]

Yep. It's always better to calibrate the XY offset on a machined feature. The reason for the square, is so you aren't trying to measure on axis reversal spots on the four quadrants of a circle. You'll be measuring in a spot where the machine was moving nice and straight during it's cut.

 

I don't even use a ring gage for my diameter calibration either. I use a 1.00000 Deltronic gage pin.

[taperlength]

First, the machine should be tested for backlash and calibrated. Also check for tram, the Z axis must be square to the X-Y plane in both directions. You would be surprised how far off that can be. For real precision, calibrate the probe on a bored hole in the fixture every time it's changed into the spindle.

[Marshal]

First, the machine should be tested for backlash and calibrated. Also check for tram, the Z axis must be square to the X-Y plane in both directions. You would be surprised how far off that can be. For real precision, calibrate the probe on a bored hole in the fixture every time it's changed into the spindle.

 

Oh I have no doubt the machine itself is off a bit, but for the vast majority of work we do that's ok. I just wanted to experiment with the probe and learn some new things with it to see what it's capable of, and figure out what more we can do with it in our limited field.

[Müřlıń®]

First, the machine should be tested for backlash and calibrated. Also check for tram, the Z axis must be square to the X-Y plane in both directions. You would be surprised how far off that can be. For real precision, calibrate the probe on a bored hole in the fixture every time it's changed into the spindle.

 

 

+1

 

Be sure to set tram on an angle plate or a cylinder square and not just sweep the table....that also checks the quill or ram at the same time.

[Müřlıń®]

Grrr

[Guest CNC Apps Guy 1]

+1

 

Be sure to set tram on an angle plate or a cylinder square and not just sweep the table....that also checks the quill or ram at the same time.

Actually, they should be using a granite square and test bar for dealing with geometric issues in machine tools but MOST shops are not so equipped.

[Joe788]

Actually, they should be using a granite square and test bar for dealing with geometric issues in machine tools but MOST shops are not so equipped.

 

LOL! I know a guy who bought a VF9 and the HFO tech was trying to square the machine by sweeping a 12 inch circle like a Bridgeport!

[Chris Rizzo]

LOL! I know a guy who bought a VF9 and the HFO tech was trying to square the machine by sweeping a 12 inch circle like a Bridgeport!

 

They stock the HFO vans full of spare spindles and mocon boards, but no tools.

[Guest CNC Apps Guy 1]

LOL! I know a guy who bought a VF9 and the HFO tech was trying to square the machine by sweeping a 12 inch circle like a Bridgeport!

PLEASE tell me that was not one of our guys. We have squares and test bars.

[Müřlıń®]

Actually, they should be using a granite square and test bar for dealing with geometric issues in machine tools but MOST shops are not so equipped.

 

 

Pfft. I ground my angle plate to within .00005 of square mehself.

 

Well..I did use a granite angle plate to do it rofl.....

 

Funny stuff joe788....some guys got all da dollars but no sense....

[crazy^millman]

That is where an outside device like the Master3DGage is ideal for checking parts. You take the variables of the machine out of the equation and can check parts very quick. For setting up parts you are okay, but only as good as the quality of the machine and last time it was laser calibrated and volumetrically checked. No offense James, but squares only do so much. By using the correct device to Volumetric check the machine you know that is working correctly everywhere. Squares and ball bar test only tell you so much, by using something that looks at all axes and travels through out the whole machine you get a real idea what is going on everywhere and not just in limited areas. The arms are calibrated volumetrically so anywhere you put it and use ti you are confident you are getting good results. Most CMM machines are using the old 10/1 rule and the biggest reason is none are check volumetrically. They go by the little area they check is good so the rest of it is good since they are engineered and designed to be accurate then because that little area they check is the rest must be. When is the last time you seen someone bring in equipment to check a CMM at the top of it's Z travel? They always check it on the table and assume since it is good there is has to be good 3' off the table if a big machine. Do you see the calibration techs bring their equipment in 24 hours before the test and record the temperature to know what the barometric pressure and humidity along with temperature is before calibrating a CMM? I have seen only one or two even come close to the right way and they still checked the machines on the table and not everywhere. Think about your machine the same way. Were is it checked how is it checked and have you run a circle, square diamond test in 40 different places on the machine and just the little area you always machine it. The best advice I could give a production shop is move that vice move that vice and move that vice. Do not put a vice in one place and run it there to 10 years. Move it around the machine. Run the machine though weekly extents of the travel to even out wear and make sure everything is getting lubricated.

 

Just my 2 cents on this topic Sorry to go off topic a little.

[Müřlıń®]

+1 on moving vice.

 

Yea Ron, that's why knowing manuals applies to CNC's.

 

Once you actually succeed in holding 50 millionths on a manual grinder, all those procedures have to be adhered to and your thought process changes on how to approach a task.

 

A surface gage and an interrapid speak volumes to someone that knows.

 

That arm will let ya get rid of a huge potential for error vs checking

things the old school way.

[Guest CNC Apps Guy 1]

No offense James, but squares only do so much.

None taken. We're talking about making sure the machine is set up properly... i.e. squareness, etc... That was off topic more than a little Ron..

[[email protected]]

IMNSHO.....probes are quick & dirty. I use them for roughing ops but when finishing a mold insert I have a .00005 B&S indicator and NO, thats not a typo its 4 zero's and a 5. It's only got .004 total travel

And we calibrate the probes daily FWIW

[Marshal]

IMNSHO.....probes are quick & dirty. I use them for roughing ops but when finishing a mold insert I have a .00005 B&S indicator and NO, thats not a typo its 4 zero's and a 5. It's only got .004 total travel

And we calibrate the probes daily FWIW

 

Well, for 99.999% of the stuff we do, we've only got to be within a couple thousandths, nothing particularly critical (yet). I've got the probe calibrated to within X0.0004, Y.0001, and Z0.0004, based off an indicator and a few trials, so I think that's more than close enough for what we do. Heck, when it did the ring gauge diameter I ended up with 2.00009 on a 2" ring gauge. Close enough for the celebrities we make crap for

 

I appreciate all the help and suggestions.

[rsc]

Hi Marshal.

 

Check runout of the stylus. It should not be more than .0002 TIR. There should be four screws (spaced at 90 degrees) at the top of the probe body that will allow you to push the probe body around on the probe shank (similar to a Buck chuck on a lathe). Remember to back off the opposite screw before trying to push the body in that direction!

 

Where I worked previously, my department got the first machine (HAAS horizontal) with a probing system. No one in the company had any experience with probing, so I was on my own. I found out quickly that you need to check it regularly, especially when the night shift fellow breaks a stylus and replaces it without notifying you!

 

When calibrated properly, it worked great! I used it regularly for setup and in-process fixture offset setting. Out of calibration and you are asking for trouble.

 

On the HAAS system you need to calibrate several aspects of the probing system. First, check runout and adjust as described above. Next, calibrate diameter using a gage ring. Next, calibrate the length. That should do it.

 

I hope this helps.

 

 

Bob

[Marshal]

Hi Marshal.

 

Check runout of the stylus. It should not be more than .0002 TIR. There should be four screws (spaced at 90 degrees) at the top of the probe body that will allow you to push the probe body around on the probe shank (similar to a Buck chuck on a lathe). Remember to back off the opposite screw before trying to push the body in that direction!

 

Where I worked previously, my department got the first machine (HAAS horizontal) with a probing system. No one in the company had any experience with probing, so I was on my own. I found out quickly that you need to check it regularly, especially when the night shift fellow breaks a stylus and replaces it without notifying you!

 

When calibrated properly, it worked great! I used it regularly for setup and in-process fixture offset setting. Out of calibration and you are asking for trouble.

 

On the HAAS system you need to calibrate several aspects of the probing system. First, check runout and adjust as described above. Next, calibrate diameter using a gage ring. Next, calibrate the length. That should do it.

 

I hope this helps.

 

 

Bob

 

Yup, I checked the runout and saw how bad it was, fixed that. Now the probe is calibrated as close enough for the work we do. We don't have a night shift, or even a second shift, so if it breaks, it's either me or our normal machinist that breaks it, and I know when that happens. At least now I know how I need to calibrate it though, and it should make him happy that I checked it and got it set up properly for him to use however the heck he wants.

 

Thanks for the tips.

[friggenbozo!]

I'm playing with the probing system on our mill, trying to learn the ins and outs of it, and I noticed a discrepancy today. Basically, when we put something in the mill we usually indicate about a hole in the center to give us our part center. Pretty basic, and we haven't had any problem with it. But I would assume the probe can do the same thing and be just as accurate, if not more accurate. So I tried it, and noticed that the center of the hole using the probed center is 0.0147" different in the X, and 0.0022" different in the Y. Which would you trust to be more accurate, the indicator or the probe?

 

 

Make an x-y calibration on your probe before every pickup. it only takes a few seconds. Automate it. Otherwise it could be out from

just handling it or going thru a toolchange.