[time-nuts] Z3805 two frequency maxima

[Volker Esper]

I agree. Please, Paul, start a new thread, and we'll be there.

Volker

Am 28.11.2012 08:10, schrieb Hal Murray:
>
> You might get better answers by starting a new thread rather than hiding your
> question in an existing thread.  (Use your New message button rather than
> Reply, and cut-paste the To address from an old message, then type in the new
> Subject.)
>
> Using a useful Subject also helps people find things in the archives.
>
>
> paul.destefano at willamettealumni.com said:
>> We have an application in which we plan to travel 7 hours with an SR620
>> turned off, then turn it on for 1 hour and take measurements we expect to
>> be accurate to less than 1ns.  I would appreciate more information about
>> how to get reliable data out of this instrument.
>
> What do you mean by "accurate to less than 1ns"?  What are you trying to measure?
>
> The internal clock in the SR620 may be off a bit.  The specs should be in the users manual.  In general, you have to wait a while for it to warm up and it will depend on which options you have.  (Some units have good crystals.  Some have low cost crystals to save $$ if they will normally run off an in-house reference clock.)
>
> Suppose it's worst case 1 part in 1E9, just to pick a handy number to work with.
>
> If you measure the width of a 1 microsecond pulse, the error from the clock will be 1E-15 seconds or 1E-6 ns.
>
> If you measure the time between 2 consecutive PPS pulses, the error from the clock will be 1E-9 seconds or 1 ns.
>
>
>