Calibration Of Stop Watch using the NIST Clock

[Calibrator24]

Is it acceptable to use the NIST Clock or a web-synchronized stop watch software as the reference to claibrate stop watch or clocks?

How can we demonstrate the traceability if this is acceptable?

 

[BradM]

Re: Calibration Of Stop Watch

Yes, you can. Here is a N.I.S.T. document that can give you some guidance:

http://tf.boulder.nist.gov/general/pdf/2281.pdf

In essence, it will boil down to what tolerance/ uncertainty you are trying to test against. Notice the uncertainties listed for calling time using a land line. You will need to assure adequate ratios.

NOTE: Calling N.I.S.T. times listed in the above document are standards.

 

[Calibrator24]

Re: Calibration Of Stop Watch

Yes, you can. Here is a N.I.S.T. document that can give you some guidance:


In essence, it will boil down to what tolerance/ uncertainty you are trying to test against. Notice the uncertainties listed for calling time using a land line. You will need to assure adequate ratios.

NOTE: Calling N.I.S.T. times listed in the above document are standards.

We don't need a highly accurate reference standard for our stopwatch. We even use a ±2 seconds tolerance when we compared them with our reference timers.

Normally, what would be the uncertainty of the NIST sync stop watches in the net?

 

[BradM]

Re: Calibration Of Stop Watch

We don't need a highly accurate reference standard for our stopwatch. We even use a ±2 seconds tolerance when we compared them with our reference timers.

Normally, what would be the uncertainty of the NIST sync stop watches in the net?

Hello there!

What I would do is just call on the telephone for verification. The best uncertainty N.I.S.T. from their testing is 500ms (listed in the link). So that you would give a 4 to 1 T.U.R., which should be sufficient.

 

[Betty Murphy]

Hello! Is there a relatively quick way to calibrate stopwatches that does not require special equipment? My team of observers will be timing various air traffic controller activities (one timer per controller, unfortunately), and we know about the uncertainty introduced by variations in human response time. For purposes of evaluating new software, we want to time to the nearest second, and not carry our reported times out to decimal places. Any advice would be appreciated. Thanks!