Dave wrote:

No, you could not perform ANY traceable calibration with the 3458A
itself, much less with any instrument you had calibrated with the
3458A, because you are not accredited (i.e., your laboratory
procedures are not reviewed and audited by a competent third-party to
establish their reliability and, therefore, to create the link of
traceability between your USE of the traceable 3458A and a primary
voltage standard).  Thus, the chain of traceability is broken at your
USE of the 3458A.  You would have a tool with a traceable calibration
(the 3458A), but you could not perform traceable calibrations with it
unless you obtained accrditation for your home lab.

Equipment dealers and even some so-called "calibration labs" ignore
this fact and act as if using the traceable DMM to calibrate another
instrument can result in a traceable calibration, notwithstanding the
fact that the person/lab doing that calibration is not accredited
(this appears to be universal on ebay, but is common among used
equipment dealers everywhere).  That is simply not
true.  Traceability exists only if there is an unbroken chain of
accredited measurements between the calibrated instrument and a
primary standard.

Calibration is one thing.  Traceable calibration is another thing
entirely, and virtually nothing you find on ebay is traceably
calibrated regardless of what the seller says (or thinks).

Best regards,

Charles

ps.  For most products, Agilent uses different equipment to do the
different levels of calibrations.  (I cannot speak specifically to
their VNA calibrations.)

On 8/12/2013 12:21 PM, Charles Steinmetz wrote:

That's simply not true. Some organizations may require a lab to be
accredited in order to accept their services, but it is not a
requirement in order to legitimately claim traceability.

The NIST Traceability Policy is found here:
http://www.nist.gov/traceability/nist_traceability_policy_external.cfm

Hi Charles,
I have to disagree on one point, You CAN do a TRACABLE calibration without any approval. What you can't do is  ACCREDITED  Calibration. Many labs are accredited but also offer un-accredited, tracable calibration at lower cost. An example is that production test equipment could be tracable but qualification test accredited. Accredition is normally driven by legislation or self-regenerating "quality" systems.
Of course if you cal a 4.5 digit meter against a tracable standard, the highest level you could reasonably sub-calibrate would be 3.5 digit or possibly 3200 count. This assumes the 4.5 digit has suitable accuracy and stability specifications, just because it has more digits does not mean it's more accurate ;-)
Somethings don't need to be tracable, a Fluke 720 K-V divider or Caesium frequency standard spring to mind.

Robert G8RPI

From: Charles Steinmetz csteinmetz@yandex.com
To: Discussion of precise voltage measurement volt-nuts@febo.com
Sent: Monday, 12 August 2013, 17:21
Subject: Re: [volt-nuts] HP 3457A

Dave wrote:

No, you could not perform ANY traceable calibration with the 3458A
itself, much less with any instrument you had calibrated with the
3458A, because you are not accredited (i.e., your laboratory
procedures are not reviewed and audited by a competent third-party to
establish their reliability and, therefore, to create the link of
traceability between your USE of the traceable 3458A and a primary
voltage standard).  Thus, the chain of traceability is broken at your
USE of the 3458A.  You would have a tool with a traceable calibration
(the 3458A), but you could not perform traceable calibrations with it
unless you obtained accrditation for your home lab.

Equipment dealers and even some so-called "calibration labs" ignore
this fact and act as if using the traceable DMM to calibrate another
instrument can result in a traceable calibration, notwithstanding the
fact that the person/lab doing that calibration is not accredited
(this appears to be universal on ebay, but is common among used
equipment dealers everywhere).  That is simply not
true.  Traceability exists only if there is an unbroken chain of
accredited measurements between the calibrated instrument and a
primary standard.

Calibration is one thing.  Traceable calibration is another thing
entirely, and virtually nothing you find on ebay is traceably
calibrated regardless of what the seller says (or thinks).

Best regards,

Charles

ps.  For most products, Agilent uses different equipment to do the
different levels of calibrations.  (I cannot speak specifically to
their VNA calibrations.)

volt-nuts mailing list -- volt-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/volt-nuts
and follow the instructions there.

A calibration indicates that the unit under test is withing manufactures
specification. The equipment and procedure used has to be "good enough"
(bad words in a cal lab) to have a high probability (nothing is 100%) of
insuring the calibration documentation is valid. Things can  can be a
little looser if you are calibrating a 1% meter with a 10 ppm meter but it
does not work the other way around.

On Mon, Aug 12, 2013 at 9:43 AM, Mike S mikes@flatsurface.com wrote:

--
John Phillips

On 8/12/2013 2:23 PM, John Phillips wrote:

You can calibrate either way. You can't however, calibrate the 10 ppm
meter so it's in spec using a 1% meter. That's different. Calibration
merely means that it's documented how close it is to a reference, such
as NIST, not that it's within the manufacturer's spec. The 10 ppm meter
would end up with a 1%+ calibration - precise but not accurate. Not
particularly useful, but valid. A good cal lab would do a calibration to
specification, where the uncertainties place the 10 ppm meter within spec.

As I cited and someone else already quoted, calibration is the "property
of a measurement result whereby the result can be related to a reference
through a documented unbroken chain of calibrations, each
contributing to the measurement uncertainty." Nothing to do with making
a device meet its specifications.

That's why an eBay seller can claim they'll do a "calibration traceable
to NIST," because they're not claiming any particular accuracy. It's
really not worth anything, unless they give specific uncertainties or
claim calibration to manufacturer's specification.

On 12 August 2013 17:43, Mike S mikes@flatsurface.com wrote:

This is the conclusion I had reached. So it would appear to me quite
easy to legitimately claim traceability to NIST.

Others seem to disagree with this, and talk about accreditation and
being able to meet original manufacturers specifications. I don't
actually see this being a requirement myself, but I'm no expert.

Of course, I would never attempt to calibrate a 8.5 digit laboratory
multi-meter using a 4.5 digit handheld one, but there does not appear
to be anything to stop me doing that, and furthermore claiming
tractability to NIST.

It is a bit like the "engineer" in England - anyone can call
themselves an engineer, irrespective of how incompetent they are. I
believe that is not so in some other countries.

Dave

This all comes down to accreditation/Licensing. There are a lot of
situations that a co. rep is enough. Accreditation/Licensing costs money.
If you require them you will have to pay. Traceability can be had without
accreditation. Calibration has been around much longer that accreditation.
It all come down to where the interment is to be used and what controls the
calibration requirement. A lot of work can be done without calibration. If
the work is not critical calibration is not required even if it is
desirable.

On Mon, Aug 12, 2013 at 6:47 PM, Dr. David Kirkby drkirkby@gmail.comwrote:

--
John Phillips