[time-nuts] Frequency standard

[Bob kb8tq]

Hi

What are the alternatives:

1) If you are doing it by ear and listening to WWVB beats and fading, something around 0.1 ppm (one hertz at 10 MHz) is doing pretty well. 
I suppose if you lived in Ft. Colins you could do better. 

2) A typical OCXO in a counter will drift and age. Depending a *lot* on the counter, drift (warmup drift, temperature drift ….) will be in the
0.01 ppm range. If you keep it powered up 24/7 you can eliminate most of that drift. You still get aging of maybe 0.01 ppm / month (maybe
more maybe less). Calibrate once a year and it all comes out around 0.1 ppm ( maybe … depending …. yada yada yada).

3) A Rubidium based atomic clock will hold 1 ppb (note we just switched units, that’s 0.001 ppm) pretty much forever. It will get to that level 
after maybe a day on power. Even while warming up / settling in, you still should be under 2 ppb. 

So, we now are at nine accurate digits on your frequency counter. We are at 1 Hz at 1 GHz after the Rb warms up. 

4) Calibrated against a cheap GPS module, an OCXO will probably get you below 1 ppb with a weekly check. 

5) Calibrate the Rb against a GPS module and you probably will be in the < 0.01 ppb range with monthly calibration. 

At this point somebody should pop up and mention that standards (all of them) have noise and we are well into the range that
noise does matter. How much depends on your counter setup. If you have something that will give you 13 digits a second. It does
matter quite a bit. 

6) Next up are OCXO based GPSDO’s. With a good antenna location (you want a view to the horizon pretty much from due east 
to due west) they just chug along. You run software only to get them setup and to see if they are still happy. The ones you see on 
eBay are going to get you into the 0.01 ppb range most of the time. Just how well they do depends on how you look at it and turns 
into a many messages in the thread sort of thing. 

7) Past that are fancier GPSDO’s, Cesium standards, and cool stuff like Hydrogen masers. Each of them could get at least a couple
paragraphs. Each has plusses and minuses. 

With the GPSDO what you are getting is convenience. Once you set it up and get it running, you forget about it. When your counter,
radio, or signal generator needs a reference, you just plug it in. What you have is good enough for any practical purpose other than 
building certain frequency standards. ( = the stuff in 7 above …).

In terms of “I want my radio to be on frequency”, we got past that up at number 3 on the list for any combo of modes and bands that 
I’ve heard of. Noise ( phase noise and spurs) would be a separate category to dig into for really exotic radio applications. Most Rb’s
are not terrific used stand alone for microwave radios. 

One thing that the GPSDO *also* provides you is an accurate 1 PPS signal. That may be of use. If so, it comes along with just about
any GPSDO. 

Bob

> On Jan 25, 2019, at 8:57 PM, Bob Albert via time-nuts <time-nuts at lists.febo.com> wrote:
> 
> Please excuse my ignorance but I have been trying to improve on my frequency calibration.
> I have always used WWV but when I try to get really close in frequency, the beat and the fading are difficult to separate.
> Someone suggested I get a GPSDO to refine my 10 MHz.  I looked around and have seen a few, mostly in the $100 range, but before I shell out fot something I know little about I wonder if someone can give me a short tutorial or at least some advice on what I need to do.
> I live in Los Angeles so I imagine signals would be reasonably strong.  But what hardware and software do I need?  I want 10 MHz to put into my counter or, at least, to calibrate the time base in the counter.  I prefer the latter, since that means I won't have to receive GPS signals in order to measure frequency.  Most of the time I don't need really accurate results.
> I could run an outdoor antenna if needed.  What receiver ought I look into, how much budget do I need, and all the other questions relating to this subject.
> And finally, how much frequency error can I expect?
> Bob
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> and follow the instructions there.