Reference Sphere

[[Me...]]

Quick question for everyone: Does everyone have their reference spheres on a calibration cycle by itself?

[[Ca...]]

I do. It came up during some audit, so now we calibrate them.

Don't get me started on how to answer: "Why does the calibration cert list the sphere size as 0.0002mm smaller than the cert the sphere came with?"

It's a good opportunity to work on your "blank stare" poker face.

[[To...]]

Please sign in to view this quote.

Where did you get your Ref Sphere calibrated? Price? Is cheaper to spend $450 - $625 to purchase a new one every year. Just curious. Having a new one waiting also means no down time.

[[De...]]

My understanding has always been that the Calibration Sphere that is used during the calibration of the machine should be considered 'calibrated' when the machine is calibrated. The calibration cycle should be the length of time the machine calibration is good for.

This is because the sphere is used during the calibration and therefore the machine calibration is dependent on using that reference sphere with that machine. The measurements taken during the re-verification of the machine for the As-Left reflect the measurements results obtained with that reference sphere for that machine.

Other reference spheres that were not used in the calibration are not used in the calibration of the machine and are therefore subject to the same requirements for calibration as any other master.

[[Ca...]]

Cert says Martin Calibration. I think I'd heard $225 (maybe?) each.

[[Ca...]]

Derek,

That was my argument, but in the end, we (they) chose to have them calibrated on a yearly cycle. I now have 2 of each, 30mm and 8mm.

[[Ge...]]

Recently, I talked to a Zeiss mechanic about this subject and got the same answer as explained by Derek.
Calibration of the reference sphere is included in the regular calibration of the machine.

[[Me...]]

Thank you all for the response. I thought as well that the sphere was calibrated with the machine during the calibration. That is what I remembered from years ago but needed some kind of validation, 🤠

[[De...]]

I am not saying its wrong to calibrate the reference spheres, however it does seem like its a particularly deep rabbit hole. The idea of calibrating reference spheres brings up so many questions for me.

What is the uncertainty of the reference sphere calibration? How is it included in the machine calibration since that is outside of the scope of a standard Zeiss calibration?

What do you do with the machine when the Sphere is out for calibration? I assume this is why you say you have two of each, and you use a second sphere during that time?

If you are using two spheres, how do you verify that the second sphere and the first are giving the same numbers? Do you have a re-verification done with both spheres? How do you account for the differences between the spheres and the uncertainty of measurement in the calibrations of the spheres?

If your original calibration cert and the 'new' calibration numbers differ from one another, does that mean you modify the numbers in the machine to match your 'new' calibration each time it is done?

If you do modify the numbers each time then it would seem that your machine would need to be re-verified with the 'new' numbers in order for the calibration to have any validity with regards to the results from the machine.

If the machine is not re-verified with the new numbers then it seems like as soon as you updated the numbers for your sphere in the machine, then the calibration of the machine would be void until you did.

If the numbers in the machine are not being updated to match the new calibration, then how does having the sphere calibrated do anything with regard to the machine?

[[Ca...]]

Derek,

Those are all really good questions, and I asked most of them, and got no clear answers. The end result was our current system, of sending spheres out for calibration. I (still) believe it's completely unnecessary.

[[Me...]]

I put a call into Zeiss back in June. I had posted this very same question for the very same reason then on this forum.
I believe they said the sphere calibration is part of your calibration report. The software re-verifies that your sphere is still accurate.
So it should be in the "paperwork".
I've called again and I'm going to have them help me find it.

[[Me...]]

I looked back. My bad. I didn't ask the question but it was for the this reason. Our I.S.O. guy wanted to see traceability for all our spheres. I believe Zeiss says they take care of that. But I did needed to send one out from our manual machine. As i said before I've called and will have them help me dig it up.

[[Me...]]

Just got off the phone with Zeiss. DO NOT CALIBRATE YOUR SPHERE. It's for reference only but falls under the calibration of the machine. Hence it is "calibrated". I've been given a document for internal use only. They've asked me not to post this but you may call to receive you own.

[[Ro...]]

If only there was a way to square a Zeiss Qual sphere to the table so that the ball is always at exactly at 45º from X&Y.
I wish somebody would come up with a way to ensure the angle is always right, for about the amount of a 1mm probe............

[[Aa...]]

Unless you are using it for more than qualifying styli on your machine, sending the ref sphere out for calibration is a waste of time and money. If, for whatever reason you do send it out, do not enter the new value into Calypso. Doing so effectively nullifies the calibration of the machine.

The reference sphere is not, in its own right, a link in the chain of traceability. For that matter, the CMM without the reference sphere is not a link in the chain of traceability. The two, together as a system (along with the size the software thinks the ref sphere is), form a link. Zeiss calibrates/verifies the machine/ref sphere combination. Ergo, they are ascertaining within their own uncertainty budget that when you use that system, specific error and uncertainty sources will not exceed stated limits. If you change any part of that system, you break the chain of traceability.

On Zeiss Micuras, for example, the calibrated (non-accredited) size uncertainty coming from Zeiss for the ref sphere is over half the base value spec for size probing error. And it does not matter one iota, because if the size of the ref sphere were off enough to significantly affect the size results on CMM measurements (statistically, this happens), the Zeiss tech could simply make software adjustments to compensate, and that would likely happen at the factory calibration, not the on-site calibration. But if one then sends out the sphere for calibration with a lower uncertainty and get a different result, changing the value of the ref sphere in Calypso woul not only breaks that link in the traceability chain, it will make the CMM results less accurate, not more accurate. (If this is not true, either the Zeiss tech was not worth his salt, or the service he was hired to do did not include making any adjustments.)

[[Ca...]]

It's really nice to hear people echo my opinion.

When I get a moment, I'll be calling Zeiss and asking for a copy of the internal document that Kevin Sipola spoke of. Maybe I'll get these sphere's out of the pointless calibration cycle....maybe.

I definitely DON'T update the sphere size btw, that was a no brainer.

[[Ri...]]

I went down this rabbit hole in my previous life.

It took some good explaining, to get internal and external people off my back about it.

One thing you could do to help your case is verify a probe with a class XXX ring.

[[Me...]]

O.K. I've had a couple of private requests to forward the doc I got from Zeiss.
For the record I promised I would not do that and I'm keeping my promise.
Sorry but my standing with Zeiss is more important than risking it.
Minnesota, which is where I called, seems more than willing to forward this to whoever calls
and will explain how they can back you up in an I.S.O. issue.

To Roberto my advantage to sphere alignment is I've got a TE-CO plate so I'm able to use the threaded holes in aligning the sphere. Haven't missed .........yet.

[[Ja...]]

I’ve been programming Zeiss CMMs since 2006. I’ve been present for every yearly calibration for every machine. I’ve never once witnessed a Zeiss tech include the machines’ artifact in the calibration. I highly doubt it’s coincidence that I’ve missed it every time. So how am I supposed to know if the artifacts are wearing if it’s never checked by Zeiss or sent out for a calibration?

[[Aa...]]

Please sign in to view this quote.

Do you mean to say they calibrate your machine without qualifying the styli they used to perform the 10360-2 and 10360-4 tests? That's Incredible!

[[De...]]

Per ISO 10360, they calibrate the probe they are using to do their linear and spatial runs for MPE E0 on the reference sphere for the machine.

They then do the linear and spatial runs using a step gage with a traceable calibration.

The measured results must come out within the specified limits for the machine based on the nominals provided at the time of the calibration of the step gage.

If the results are in the acceptable range then the sphere is correct, or more precisely it is correct enough. How do we know it is right, we know it is right because it was used to calibrate the probe that was used to correctly measure a known value.

If your not using the reference sphere to measure anything directly, then the 'calibration' of the sphere is not necessary because it is used as an integral part of the calibration that checks the sphere and the CMM as a system that is tied together.

[[Cr...]]

Zeiss document saying that periodic sphere calibration was not necessary confirmed. However, I was not asked to keep the document to myself. I have attached it to this post.

ref spheres.pdf

[[To...]]

Please sign in to view this quote.

I think he was saying they didn't use the reference sphere that came with the machine. Normally, they use their own Master Probe and Ref Sphere.

[[Aa...]]

I don't think I've witnessed as many as Jack, but I've never seen one bring their own reference sphere. 10360-2 says, "Set up and qualify the probing system in accordance with the manufacturer's normal procedures." That would mean using the machine's reference sphere.

[[De...]]

Per Section 6.2.3 of ISO 10360-5 which pertains to Single and Multiple Stylus Probing systems and complements ISO 10360-2 a calibrated reference sphere other than the one supplied with the CMM must be used.

It is most likely this test that you will see them using their own reference sphere on.

This test checks single stylus form error and since the reference sphere is calibrated for form and size gives a method to independently verify the machine specific reference sphere is giving comparable results.

Once this test is completed, the checks for ISO 10360-2 are completed to check linear error of the CMM, these checks should be completed with the Reference Sphere that is specific to the machine.

If a tech uses their own Reference Sphere to do the ISO 10360-2 checks then I would question that because if that is the case then the results in that case would not validate the calibration of the CMM and the reference sphere as a combined unit.