[time-nuts] Measuring 10 MHz

Tom Van Baak tvb at LeapSecond.com
Thu May 8 15:37:06 EDT 2008


> Hello,
> 
> Thanks to all for your replies and suggesting especially Brooke, Bruce and 
> Pete.
> 
> I am already using the Stanford technique with the 1 kHz output.  That's how 
> I get 2 x 10E-12/sec accuracy.

Martyn,

Just to make sure -- when you do this with your SR620 (i.e.,
1000 externally gated TI readings per second), what sort of
wiggling do you see in the last digit of your display?
Also, how much did the mean change over a minute or two?
Finally, if you display jitter instead of mean, approximately
what numbers are you seeing with your setup?

> Using this technique, the accuracy should be 7.9 uHz at 10 MHz (7.9E-13). 
> But I don't reliably  achieve this.  I reliably get 20 uHz, sometimes 10 uHz 
> and  I do have resolution to parts in 10E-13, but the last digit isn't 
> reliable.

Where did you get the 7.9 number?

/tvb




More information about the time-nuts mailing list