[time-nuts] Compairing two GPSDOs Oscillators

[Perry Sandeen]

List,

wrote: > I'm looking for some advice about ways I can compare the frequency of two gpsdo's.

> Any thoughts as to what could be obtained used for less than 1K that would be suited for this type of measurement ?

OK, Here’s what I use and why.

For a GPS receiver: the Lucent RFTG-M-XO GPS KS-24019 L106A.  Google it through Sports Lineup.  $100 to $125 plus shipping all day long.  Sometimes even less.  I got two over a period of time for $100 each including shipping.

Why I use them.  At the time I bought them the HP Z series were $300 to $500.  I didn’t want a Trimble from China and they require a PC type power supply.  If I had known about lady heather at the time I might have bought one.  The Lucent uses a single 24 volt power supply.  Also the Trimble oscillators seem to be of a lower quality than the Efratom in the Lucent.  That said, a U.S. supplier is offering the latest Trimble W/PS for $150.

The downside to the Lucent is you have to use the 15 MHz output or hack it to get the original 10 MHz before it’s up-converted.  For me this hasn’t been a problem.

For frequency comparisons I have two HP 5370B TIC’s.  It will resolve time differences of two frequencies down to 10 pico-seconds.  As a counter it reads 16 digits plus sign. It was the last and most complex counter HP made.  I paid $450 for one and $100 for the other. (A steal).

Downsides.  They run extremely hot.  One should add some additional fan(s).  Very heavy.  Very large.  The 10811 internal oscillator DOES NOT have EFC.  The best manual setting I could get was within 20 milli-Hz at 15 MHz.  It took a lot of time.  Although both pass self-test, the 100 pico-second differential test is off by 880 pico-seconds on one and 670 pico-seconds on the other.  This would require some serious calibration procedures.  However the error is constant so all one has to do is to remember to add or subtract it depending how you’re measuring.  For 25+ year old equipment I don’t consider that a bad defect.

I have 3 tested Lucent Rubidium RFG-M-RB 15 MHz/10 MHz units. One is slightly different but they all work.  I had to run them on the bench for four to eight weeks before their offsets stabilized.  I paid about $100 each with shipping.  All lucent output connectors are SMA so one needs SMA to BNC female adapters.

Wrote:> All are in the "sub $300" range on the normal sites. Some are sub $100. All are available with GPIB for logging.

I guess I’ve been looking in all the wrong places.  I watch Ebay prices all the time for HP 5370A and B prices.  I’ve never seen one for less that $500.  If there are cheaper places I’d sure like to go looking!

SR 620  This is made by Stanford Research.  I agree that it is the best.  It can do Allan variations as well.  A current Ebay price: Stanford Research SR620-01 Time Interval Counter $3,250.

It’s like the old saying:  How fast can my car go?  Answer:  How much money do you have?

IHTH

Perrier