[Elecraft] 10 MHz external reference - GPSDO

Edward R Cole kl7uw at acsalaska.net
Fri Mar 11 16:30:22 EST 2016


Joe and all interested in the EXREF:

OK ... but most will not measure the 49.380 TCXO 
directly*.    *though REF CAL displays this.

What I think most want to know is how accurate their operating 
frequency will be.  I performed rigorous measurements as beta tester 
of the EXREF (unfortunately my website was hacked a couple years ago 
and I did not recover the beta-testing data page).

First one needs to realize that the TCXO establishes the basic 
stability, so choosing the TCXO-3 over the TCXO-1 will provide better 
stability and accuracy (these are two different things).  I bought 
the TCXO-3 with my initial K3/10 purchase so the rest of what I say 
is with the better TCXO-3 installed (if frequency is important then 
that is first thing I recommend updating).

Mostly, I use 28-MHz as a testing frequency since it is my VHF 
transverter IF.  On 28-MHz with only the TCXO-3, I see 14-Hz 
inaccuracy.  This 14/28000000 = 0.5 ppm (I think this is better than 
Elecraft specs the TCXO-3).  Adding the EXREF I see 2-Hz which is 
"about" 0.1ppm.  Measurements on 50-MHz are proportional.  I see 3-4 
Hz accuracy using the EXREF.

So going lower to 10-MHz I measure 1-Hz or less inaccuracy.  Due to 
the slow drift of the TCXO's, stability is same as accuracy (or maybe 
better).  So all ham bands lower than 10-MHz show less than 1-Hz 
(basically I measure zero error).

Read below for optional background in my measurements-----------
I am measuring with a high-end mw counter able to count up to 26-GHz 
and has an internal TCXO reference or can take an external 10-MHz 
reference.  I also have a LPRO Rubidium source which is rated +/- 5 
E-11 stability.  I find measurements using either the counter 
internal TCXO or the Rb source are identical when measuring under 
50-MHz signals (speaks well for that TCXO).  I have to wait at least 
30-minutes for complete warm-up stabilizing of the Rb source so 
usually for normal frequency checks I just run the counter with internal TCXO.

MY station 10-MHz standard is a Russian surplus MV62 Morion OCXO spec 
+/-5 E-12 (note it is more stable than my Rb).  But the OCXO must be 
calibrated as it is voltage tuned.  I use the counter/Rb for 
that.  Long-term (measured in months) stability of my OCXO is only 
+/-5 E-5 so I figure that I have to check the OCXO against the Rb 
about twice/year (never been more than 1-Hz off in 3-years running).

The counter can only count to 1-Hz so measuring less than a Hz is not 
possible.  There are sw to make such but I have not gone there (yet).

I build transverters up to 10-GHz which use PLL LO's which actually 
lock to the 10-MHz reference.  I have noted that I can adjust my OCXO 
when measuring 10,224.000,000 MHz to bring it closer to 10-MHz 
because any sub-Hz error is multiplied in the mw LO multiplier 
chain.  Something I cannot do if directly measuring the 10-MHz 
reference.  Of course this is "nit picking" for setting the OCXO for 
use on HF/6m with the K3EXREF.

One caveat in using a Rubidium full time as reference.  They have a 
finite life-span and more expensive (and probably harder to find once 
one fails).  So I only run my Rb for high accuracy checks and "lock" 
for daily use to my $45 or $25 OCXO.  My OCXO runs 24/7 from a 17AH 
battery which is float-charged from a good 12v PS.

73, Ed - KL7UW

Date: Fri, 11 Mar 2016 13:25:54 -0500
From: "Joe Subich, W4TV" <lists at subich.com>
To: elecraft at mailman.qth.net
Subject: Re: [Elecraft] 10 MHz external reference - GPSDO
Message-ID: <56E30DB2.6080701 at subich.com>
Content-Type: text/plain; charset=utf-8; format=flowed


  > While locked to the external source, the K3?s reference
  > oscillator frequency is maintained within 2 Hz.

The K3's reference oscillator is 49.380 MHz (+/-) not 10 MHz.

The K3EXTREF uses the external 10 MHz signal to gate a counter
which counts the reference.  That count is then used to adjust
the reference frequency.  Maintaining the 49.380 MHz reference
+/- 2Hz is on the order
of 0.04 ppm.

73,

    ... Joe, W4TV



73, Ed - KL7UW
http://www.kl7uw.com
     "Kits made by KL7UW"
Dubus Mag business:
     dubususa at gmail.com



More information about the Elecraft mailing list