[time-nuts] timelab

[Azelio Boriani]

Well, so I have to have stable values... but if I'm measuring my clock I
wouldn't expect stable values (of course at most, say, 1nS apart one to the
other for every second). That is, assuming that my counter samples at the
stop, have I to sample exactly at 1 second using the reference as the stop
and the DUT as the start so that the uncertainty will affect the start
instant that has no influence on the sampling interval?

On Thu, Jul 5, 2012 at 12:44 PM, Magnus Danielson <
magnus at rubidium.dyndns.org> wrote:

> On 07/05/2012 10:21 AM, Azelio Boriani wrote:
>
>> Don,
>> what do you mean by "not getting them sequentially"? The stop signal
>> should
>> determine the rate of the samples and if the stop signal is the test
>> signal
>> then any non-uniformity in time is due to that signal. Maybe a good idea
>> to
>> feed the stop trigger with the reference to assure the best time
>> uniformity. Once the sample is ready then feeding it at the exact time
>> when
>> it is available or deferred should be irrelevant. A process is ergodic
>> when
>> more than one statistic give the same result but in this case I can't see
>> how to check for ergodicity. One move maybe to apply the average to groups
>> of samples and see if the average is the same for every group of samples.
>> We already know that this may not happen and the Allan statistic was
>> introduced exactly to account for that.
>>
>
> It the time of the sample he means, and as a delay factor it should
> experience flicker and white phase noise in which case the assumption is
> fair.
>
> The trouble is if he doesn't have stable values and in particular if it
> skips over measure. I tried to sketch that in the previous post.
>
>
> Cheers,
> Magnus
>
> ______________________________**_________________
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/**
> mailman/listinfo/time-nuts<https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts>
> and follow the instructions there.
>