[time-nuts] A Rb with 2e-15 per day aging rate

[ws at Yahoo]

The sensitivity of that EFC input is so low on my LPRO unit, that the PS and 
DAC noises are not a problem at all.
Should add a 1K resistor in series with the Dac out before connecting it to 
anything outside of the TBolt.
Attached is an expanded 4 day plot showing part of the same LPRO run where 
the Dac is not changing at all.
Using just the Phase data, The freq offset calculates to 7e-14 and the 
ageing rate is under 1e-14

ws

*******************

Azelio Boriani azelio.boriani at screen.it  posted:

The frequency accuracy is measured based on the EFC variation.
I have recently received my TBolts, not yet fired up, actually preparing the
power
supply. Where is the DAC? I can't find... the LT1014 is not a DAC, the SO8
2105 is not an LT2105 nor an LTC2105. The DAC is 20bit as I can read in the
manual and for a 10-volt range (-5..+5) the resolution is 9.54uV. I think I
have to take great care in the power supply...

******************
On Tue, Nov 22, 2011 at 11:00 PM, WarrenS <warrensjmail-one at
yahoo.com>wrote:

> To TBolt nuts
>
> A Tbolt that has been modified to use an external 10 MHz osc is all that
> is needed to be able to measure freq and aging of an Rb osc to better than
> 1e-14 using a week long test run.
> Attached is a plot showing a recent 13 day long LPRO Rb test I did.
> This modified LPRO showed an unbelievable average ageing rate of only
> 2e-15/day over the first 12 days.
> Of course anytime you get much below 1e-13 for a non Cs osc, it has some
> luck evolved.
> But still interesting that using ONLY a Tbolt and LadyHeather one is able
> to measure freq that accurately.
> (Can do even better using Dual Tbolts. Used to cancel the shorter term GPS
> phase noise)
>
> ws
-------------- next part --------------
A non-text attachment was scrubbed...
Name: ws-LPRO-11-22b1.gif
Type: image/gif
Size: 35122 bytes
Desc: not available
URL: <http://www.febo.com/pipermail/time-nuts/attachments/20111122/ae7d7b53/attachment-0001.gif>