Simple diameter measurements not matching reality

[[Th...]]

Struggling to figure out what my Contura is doing right now. I've just measured a part with a .1875" nominal internal bore that my CMM is reporting as .188" across 3 different stylus systems. I have checked with a class X Deltronic pin that the hole is machined to .1875", and I have checked with two Duramax CMMs in two different buildings, both of which report the correct value and agree with the pin. V-clamped the Deltronic on my Contura and measured, it thinks it's .1872". I've already verified the condition of the pin with a set of high-accuracy micrometers we keep on hand, and the pin size is .1875", as it's supposed to be. Reference sphere is undamaged, values are as recorded in the certificate. Master probe appears undamaged. CMM passed calibration in mid August.

What on Earth have I missed here? It's driving me up the wall.

[[Da...]]

Try using a reference sphere from one the other machines. If your results change, get your sphere re-certified. Don’t just use made-up values (I’ve seen that happening)!

If not, temperature might be an issue. Since I don’t think that you have a 50 °C temperature difference, which would roughly explain the difference of .0002", please consider the following questions:

1. Is your lab temperature-stable?
2. Do you happen to have a heater or radiator next to the machine?
3. Are there any vibrations in the vicinity of the lab, like forklifts going by?
4. Is this a recent problem? Can you find older results that seem to be good?
5. Measure a certified gauge ring to confirm the bad values and confront the technician who checked your machine.

[[Ri...]]

Something I came across once in a similar situation.
The user who setup the Reference Sphere made a typo entering the radius value of the Reference Sphere.
Once that was corrected, everything was fine.

[[Ow...]]

Great suggestions above but, I'll add a few.
Are you scanning or taking points and is there a difference when you check both ways? If there is a difference, especially on the certified pin or ring gage, it's either the strategy, probe or the head.
What is the form/roundness of the pin and hole when checked on your Contura vs the two other CMM's?
Last but not least, how long is the probe you're checking it with?

[[Je...]]

Because your ID feature measures O/S and your OD feature measures U/S, the issue lies in the geometry (radius) of your stylus. Requalify your styli. Go to your stylus system properties and review the quality of the stylus. View in metric. The radius should be within 1-2 micron of the stated size of the stylus. Any more than that and you should discard it. If it is U/S, it is worn and has been for quite some time. If it is O/S it likely has buildup (this is more likely the case and causes similar symptoms to what you are seeing).

If you find that the styli is non-conforming, you should review your qualification schedule. Daily "geometry requalification" runs on common styli should be the standard for a production reliant CMM. Daily cleaning schedule of measurement equipment should be reviewed as well. Less commonly used styli should be done weekly or per-use-basis. A qualification program should be created to report data for each styli (size/sigma - ~1-2 micron limit for each is standard). I can't tell you how often we troubleshoot a CMM measurement to find out the styli haven't been qualified in several months and find that they are cracked or have 5 microns of aluminum build up on a Ruby stylus (use Silicon Nitride for scanning Aluminum and Zirconium for Cast Iron - these are a game changer. Diamond/coated are only necessary for Xenos or HIGH VOLUME Production as in 24/7 automation).

Measure the artifacts with different styli. If the issue persists across several styli, review your Ref Sphere data. If there is a QR code on the stem, scan it with your phone (with QR reader) and it will open the DakkS Ref Sphere cert with all the data. Otherwise refer to the spec sheet or calibration cert that you have with the machine. Last resort, utilize a different Ref Sphere (I advise against this as mixing up spheres across machines can jeopardize you Quality Management System.

Verify the size of the MasterProbe after locating the Ref Sphere. It should be EXACT for XXT at 5.0mm (R2.5mm) or 8.00mm (R4.00mm) for VAST. If these values are not within a micron or .000039" and Sigma less than that, check the MasterProbe for damage/dirt. If the size is out more than this but the Sigma is good, QUADRUPLE check your REF SPHERE data and ensure you have the matching sphere on the machine. This is a hugely common mistake that is very easily overlooked when multiple machines exist in close proximity.


If the issue still persists, I advise calling HW Tech Support to assess your sensor.

[[Jo...]]

Per Jeff's recommendation. You can set a global sigma value, and interval for stylus systems. You can also set individual stylus system limits.

This does several things:
Forces "operators" to clean reference sphere
stylus systems must be clean or they won't qualify
catches stylus damage, and compromised connections, extensions
and with automated report, DOCUMENTS your measurement system condition

[[Ro...]]

If you're comparing hole size to pins then you need to be using Max Inscribed in your evaluation settings.
However, in my experience the small diameter drilled or reamed holes have very little form deviation, so the difference between max inscribed and least squares is almost nothing.

[[Ch...]]

I've tried to explain to my wife how we can talk about one circle all day. And all the intricacies nuances of that one feature; Then come in the next day and still be talking about the same circle.

Us metrology people are weird people.

[[Jo...]]

Please sign in to view this quote.

I've resorted to telling engineers that a triangle can be expressed as a circle with bad form. Unless you qualify the form, we're not communicating.

[[Th...]]

I'll try to cover as may suggestions as I can.

Temperature stable, no nearby heat or vibration sources. This is a recent problem. Ordering a certified ring gauge is my next planned step.
Ref sphere values are verified as recorded in the book. There is no mixing of spheres on machines.
Oversized condition did persist across a 1, 1.5, and unused 2mm stylus. The undersized condition on the pin was measured in Y/Z with a horizontal 2mm, while the bore was measured in X/Y.
Replaced 30mm ref sphere with a 25mm from one of our newer Duramax machines. Size of Masterprobe and the 1.5mm I'm checking with shrank to just over 1 micron from nominal, but sigma values are under 1 micron still.

On to the hole measurement, requalifying with the alt ref sphere did not change the results, nor did changing from scan to single points at the same step width. I even created a dummy Masterprobe using a very similar silicon nitride stylus we ordered years ago for some reason that I don't recall (P/N 626115-0800-064) which once again had no effect on the measurement.

Looking at the calibration cert from August, the only change between then and now is that I have since updated Calypso from 7.0.20 to 7.6.04. Installing the patch file and reverting to 2020 are the last things I can think of that I can do myself, although I don't know if I should expect a change.

[[Da...]]

Please sign in to view this quote.

In addition I can only think of homing CMM and restarting Calypso afterwards and similar as you did check the gage in every axis combination X/Y, Z/Y, Z/X.

[[Th...]]

Unfortunately, full restarts of both the CMM and workstation was one of the first things I did.

[[Da...]]

Another thing that I personally am struggling with: After some time of not cleaning my reference sphere, I get a lot of trouble of keeping my calibration values good. Since my company is directly situated next to a coal power plant, there seems to be some kind of invisible gunk build-up on the ref spheres. After a thorough cleaning, everything is all right again, but it used to drive me crazy, because there is no visible indication that there is any dirt on the spheres.

[[DW...]]

Please sign in to view this quote.

Germany still burns coal? I thought they were all renewables? Sorry this just stuck out to me 🤣

[[Da...]]

Please sign in to view this quote.

Yes, coal share in Germany is still around 27 % versus 20 % in the USA.

P.S.: Sorry, another website states, it's still 30 % in the USA.

[[Je...]]

Qualify the Ref Sphere with the MasterProbe.
Measure a circle on the the Ref Sphere at the equator with the MasterProbe.
Report the size versus certificate size.

If the values do not match with 1-2microns (filter/evaluation effect), you need to call HW Tech support.
If they do match, repeat the measurement with a different stylus at different X/Y, Y/Z, X/Z orientations.

The last time I saw this, the dovetail screw on the sensor mounting to the Z axis was loose. Luckily that is an EASY fix.

[[Th...]]

Masterprobe verified in X/Y at about a half-micrometer off nominal. Book value for mean sphere diameter is 30.0002mm, measured at 30.0007mm. 2mm -X stylus used to measure in Y/Z, result is 29.9912mm. Checked X/Y with 1.5mm and 3mm, both measured at 29.997mm. No shared components between any systems. Circle path set to Navigator automatic strategy and cookbook standard outlier/filter settings, diameters all reported as LSQ.

The set screw holding the XT in place appears to be secure.

Just checked X/Z with a 3mm stylus in +Y, measurement is 29.9939mm.

[[Je...]]

Please post the reported radius size of your MasterProbe as well as the other styli you are referring to with incorrect reporting. Please post your version of the following screenshots so that we can eliminate 10-12 other questions that will inevitably come up.

My screenshots are from a Simulation seat so they will not look correct but this is just for a reference of what windows to take a screenshot of.

MP (adding one from one of the other styli is helpful as well)
The REF Sphere listed in the previous shot should be posted:
The whole Stylus data set for the MasterProbe

[[Th...]]

Masterprobe stylus data:
Geometry for same:
Quality for same:
Settings for same:
Ref sphere:
1mm stylus qualification window:

[[Je...]]

Thank you for all the info. No huge glaring issues seen here. There are a few values displayed here that lead me to 2 different conclusions however. Please correct me if I am mistaken:

1 - It appears that you are using non-ZEISS styli and/or Ref Sphere. Not saying it is outside the realm of possibility but a 100nm roundness and .00066 sigma is quite a lot for a ZEISS supplied Ref Sphere. Also I can't say I can recall the last time I saw a ZEISS Ref Sphere over 30.0000mm. They are, from memory only, all 29.99950-29.99999. Also a ZEISS supplied MasterProbe should be between 3.9996-4.0000. None of these indicate a glaring problem but they do add up.

2 - If I had to guess I would say you have a sensor/machine issue as your worst deviations coincide with the Z axis use but this would require a Service Tech to verify 100%. Especially when you measure the sphere in different orientations with the same size stylus and the values are 3-4 microns different.

Could you post a roundness plot of any of those circle scans? Notably the MasterProbe but include as many of the styli with larger deviation as you can also. Show roundness at a magnification that is discernable and post that value as well please.

Another way to troubleshoot if you have a squareness/machine error issue is to measure a ring gage twice in each of X/Y, Y/Z and X/Z orientations. Record their size and plot roundness. At each orientation, then rotate the ring 90° about its axis (spin it) and remeasure. Compare roundness plots for the 2 identical orientations (but ring rotated 90° between). If the roundness deviation direction didn't change and coincide with the rotation of the ring, the error lies in the mechanical system of the machine. If it follows the ring, this indicates the machine is working properly.

If you notice a jagged/sharp transition in the plot when the machine/sensor changes direction, this usually indicates an issue with an air-bearing dragging, or less often, a drive motor issue.

Also, change the Navigator Optimum to a setting of a known value. 5mm/s with maximum point concentration (minimum step width before speed drops below 5mm/s) for your machine. 380-400 degree scan.

[[Th...]]

Masterprobe and ref sphere are both Zeiss-supplied and original to the machine. Sphere calibration cert is dated end of October, 2015. All adapter plates and extensions are likewise Zeiss components. Most measuring stylus tips are not.

Roundness pics attached. Masterprobe value of 0.0012mm is within 0.1 micrometer of the X/Y value in the most recent calibration. In addition, I've measured with the 1, 1.5, and 2mm styli. Angle range 380, speed 5mm/s, step width auto-calculated at 0.01111 mm. I also set up, qualified, and measured a Zeiss 8x114.5mm stylus that shipped with our kit, P/N 600342-8024-000. Qualified radius of 3.99843mm and sigma of 0.00136mm. I accidentally left the mag settings at automatic for the 1mm scan, so it's at 10k rather than 20k like the others.

Unfortunately I don't yet have a ring gauge or fixturing to check squareness, but this is now at the top of my shopping list.

8mm round.PNG2mm round.PNG1-5mm round.PNG1mm round.PNGmprobe round.PNG

[[Da...]]

Please sign in to view this quote.

Jeffrey, are you saying that Master size is very important? I can't get it why, if the calibrated reference sphere gives size to all probes, including Master.

[[Ri...]]

Throwing this out there.
The last time I viewed maps like that (knowing you're at 2000:1) our Y axis motor had an issue.
I'm going by memory, but it definitely triggered a memory.

[[Je...]]

Please sign in to view this quote.

No, it is not important other than it does have to be within a range (that range is quite large). It was just odd that where the values reported vs current production standards. The equipment is 8+ years old though and things may have changed since then.

Thomas, Sorry that I missed your response postings of the roundess. It would appear that you have a drive or air-bearing issue on your X axis (assuming your plots are relative to the cmmSys). On each plot there is a clear change in vectoring and actuals as the machine changes directions. This is usually referred to as a "lag error" but it is quite small. I would advise requesting a HW Service tech review your machine. Admittedly I am more experienced in the SW side of our business and not a HW expert but I am quite familiar with general diagnosis. I would expect a Service Tech to review the draw of the axis motors and the efficiency of the air bearings. Of course this is all based on the plots you have provided only and there are better tools to analyze this from a HW diagnosis perspective. I advise you contact your local regions CIC for a HW Service ticket.

A squareness check using a proper CMM Square or even a Ball-Bar may reveal the issue. Another way to self-diagnose a drive/bearing issue is to run the scans at higher speeds to see if the jumps in form become even more pronounced at the axis change in direction.

[[Th...]]

We had a service tech out a couple of weeks ago. XTG confirmed bad (something failing in the Y axis, it was clicking during tuning) and replaced. Roundness plots are well within spec again. However, it seems that the head issue was coincidental, and that as you say, there's something else wrong. Worked with our local equipment supplier and ran some gage block checks. In X and Y, using 1/2/3/4 inch blocks, the master probe measured -3 micrometers at the 1 inch block, down to -1 micrometer at the 4 inch block. I ran the same tests with the 1.5mm stylus and got -20 to -23 micrometers. There doesn't appear to be much difference in values between the two axes.

We've opened a PO for a calibration service, but if this one isn't treated as a repair then it's going to be months before we get another tech to visit, and the CMM has already been inoperable for a month.