[time-nuts] How to test a super Rb standard

[ws at Yahoo]

on Wed, Jan 18, 2017 Chris wrote:
 
>But BEFORE you try and improve the Rb performance you need to have some way
>to MEASURE its performance.  This is likely much harder.
 
 
Testing most RB's mid & long term performance can be done "on the cheap" by
using a Tbolt that's been set up to accept it as an external Oscillator.
With discipline turned off, it uses the GPS directly as the Freq ref to plot
& record the Rb's Phase error over time. 
Can then use TimeLab to do the ADEV etc. plotting from the large saved
LadyHeather's  xx.Log data file.
 
The limit using GPS as the Reference to test a "super RB" is the GPS's
accuracy which I hear is around +- 10ns & ~1e-13 per day when using a good
antenna setup.
 
Attached is a LH and TimeLab plot of a super LPRO RB with ADEV of <1e-13 at
1 day and a pretty consistent freq offset of ~7e-14 over 4 days. 
(This LPRO showed an ageing rate of 2e-15/day over a selected 12 days) 
 
Ws

 

 

-------------- next part --------------
A non-text attachment was scrubbed...
Name: LPRO Nov2011d.gif
Type: image/gif
Size: 50219 bytes
Desc: not available
URL: <http://www.febo.com/pipermail/time-nuts/attachments/20170119/7ac0f2ba/attachment-0001.gif>