[NLRS] 10 MHz frequency standards

[Donn - WA2VOI/0]

----- Original Message ----- 
From: <tosca005 at umn.edu>
To: "NLRS Reflector" <nlrs at mailman.qth.net>
Sent: Wednesday, September 16, 2009 1:11 PM
Subject: [NLRS] 10 MHz frequency standards


Which do you think would provide a superior 10 MHz frequency standard to
lock a 10368 MHz transverter using an Apollo board?
  -- rubidium oscillator
  -- Thunderbolt GPS disciplined OCXO

SNIP

Ideally, I am looking for BOTH precision and accuracy. It would be really
nice to tune to 10368.1 MHz and know that I was within 1 KHz of the actual
frequency so that QSO's wouldn't be lost by failing to tune around far
enough, as has happened in the past. And it would be really nice to have
the frequency stay stable from start to finish.

Any thoughts out there?
73 de WØJT

****
For that level of accuracy (1kHz), a simple TCXO in a box to keep the wind off 
of it is more than adequate.  The major error will be in the overall accuracy of 
the 10MHz.  (I'm using a 10MHz TCXO on 10GHz... its offset from 10MHz enough 
that an actual frequency of 10.368.100 reads 144.120 on my i.f. radio.  The 
20kHz offset is deliberate and very stable... less than the 1kHz you're looking 
for over the conditions we see during the day.

An OCXO is a step up and with proper alignment (have a GPS standard at home, and 
use Lissajous patterns for alignment) will put you well under 1kHz.  The 
advantages here are cheap, no GPS lock-up time, etc.  A (GPS) disciplined OCXO 
is a very good option IF THE OCXO CAN RUN WITHOUT being disciplined for "long" 
periods of time... say a few hours.  Not all of them can.  Some require constant 
corrections.

An Rb will only put you within a few cycles (unless it, itself, is GPS 
disciplined), and MAY increase the phase noise.  (We still don't know/understand 
how much phase noise is too much.  For the operation we do around here, it 
doessn't matter, but it may some day.)  Rbs are also more expensive and, as 
Bruce notes, DO have a finite life.  Those we get surplus could die tomorrow (or 
5 years from now... you don't know !).

The last issue is your i.f. radio's tuning accuracy.  It doesn't do much good to 
be on 10.368.100 000 +/-1 (Rb) if the i.f. radio is off by 150 Hz (~1ppm) or 75 
Hz (0.5ppm) and/or drifts like a bansee.  Most of us are using older 2m i.f. 
radios that are worse than 2-3ppm.  An IC-706 w/o the high accuracy reference 
is, all things considered,  +/-2.5ppm*; with the high accuracy ref, its spec'd 
at +/- 0.5ppm**.  The only solution is to ALSO discipline the i.f radio to the 
same level of accuracy... which is VERY hard to do with the stuff we're using.

Why spend time and money on 1ppb (1 Hz) at 10.368 and feed it to an i.f. thats 
100X worse ?

73 Donn
WA2VOI/0


* Less than ± 7ppm from 1 min. to 60 min. after power ON. After that rate of 
stability less than ± 1 ppm/hr. at +25 °C (+77°F). Temperature fluctuations 0°C 
to +50°C (+32°F to +122°F) less than ±5 ppm. (SOURCE: ICOM 706MKIIG 
specifications, Frequency Stability)

** SOURCE: IC706MKIIG Manual, pg 60, CR282 High stability crystal unit.