[volt-nuts] EDC VS330 Voltage Standard

[Marv @ Home]

You could have a vote if all instruments being queried were equal in 
all respects, but at the metrology level every measurement has a 
confidence level attached to it, and between 3 meters the most 
recently cal'd one is probability wise, to be right than the older 
ones.  OTAH, if you have data on older machines showing their rate of 
drift and direction, you may have confidence in them compared to a 
younger undefined machine.  A lot depends on the data you have on 
comparable meters to give you more or less confidence in the reading.

This data is the difference between the costs of the calibration 
written earlier:

ISO 9001 = $50
Z540 = $175
ISO 17025 = $275

The costs vary by the amount of data provided _and_ the amount of 
secondary checking of your device.

In order to value this extra cost, you need to understand and track 
the data of your meter for its lifetime. If you have no need of more 
stringent data set, you can opt for the ISO9001, the basic procedure 
level calibration; it may not include adjustment.

If all your instruments are off their recommended cal date, and the 
VS330 is received in an unknown state stored or treated in unknown 
conditions, you really can't say its a better measure.  You need a 
verification from a reference source, be it a cal lab or someone's in 
cal DMM, and given its your most accurate device, assuming nothing is 
wrong with it, it should be the one calibrated to serve as the model 
for the rest of your lab.

Finally, the cal lab itself.  There are primary labs that hold their 
own JJ standards, and there are secondary labs that do not; primary 
labs are the best but far less numerous and maybe more expensive.

e.g. Agilent and NIST labs are primary labs.

An inbetween option are the VLAP labs, NIST accredited to follow 
procedures far stricter than ISO standards.




At 08:24 PM 7/28/2013, J. L. Trantham wrote:
>Same issue as with the time-nuts and the man with one watch versus more.  No
>way to tell except to 'vote with the majority'.
>
>So, the answer is 'probably'.
>
>Joe
>
>-----Original Message-----
>From: volt-nuts-bounces at febo.com [mailto:volt-nuts-bounces at febo.com] On
>Behalf Of Joseph Gray
>Sent: Sunday, July 28, 2013 6:37 PM
>To: Discussion of precise voltage measurement
>Subject: Re: [volt-nuts] EDC VS330 Voltage Standard
>
>When I get the VS330, I think I'll first try setting it for 5 V and
>comparing the 3478A reading with what I got with the MAX6350. If I get about
>the same reading of 5.0026 V, is it safe to assume that the 3478A is in
>error?
>
>
>Joe Gray
>W5JG