[time-nuts] Some examples of HF frequency smearing`

John R. Ackermann N8UR jra at febo.com
Tue Nov 16 20:41:08 UTC 2004


Tom Van Baak wrote:
> That's a nice set of plots. Very cool tool. Too bad there's
> no Windows version. Or is there a x86 Windows-based
> emulator for Linux executables that would help?
> 
> Did you see the dozen mystery signal examples in:
> 
> http://www.baudline.com/mystery_signal/

There are some cool ones there.

> 
> I believe the baudline tool would be fantastic for analyzing
> short-term stability of oscillators.

Well, I've been talking with the author and I think I have him convinced 
to add a logging output that would give you timetagged frequency values 
using high-resolution measurement tools he recently added.

A problem with this is that you're at the mercy of the sample rate in 
your sound card; with a decent card the offset and/or drift is small, 
but it's still an undisciplined oscillator.  One answer is to use a card 
with a "wordclock" input that allows you to provide the clock signal. 
Another answer is to monitor a precise reference tone and compensate for 
offset or drift based on that.  I'm also working on getting the author 
to include something like this.

What's cool is that the author (Erik Olson) is extremely open to 
suggestions like this; the current version of the program includes 
several features I've asked for specifically for frequency measurement 
stuff.

> John, your plots bring up an interesting issue. Calling it
> a frequency shift over a half hour may be mathematically
> correct but we know this is actually a received carrier
> phase shift over the sampling time. A good analogy of
> how a phase microstepper works.

Good point.  We really need to do some comparison plots during times of 
more stable propagation.  On Sunday, both Daun and I saw significant 
times when the WWV carrier actually appeared shifted by as much as 0.5Hz 
for many minutes at a time.  If that's the result of Doppler as the 
ionosphere was moving up or down, it's much longer-lasting than I would 
have supposed.

This is where having a log of frequency vs. time would be interesting; 
we could correlate the frequency offset with local and remote sunset time.

By the way, in the plots a frequency of ~1849.577 Hz represents "on 
frequency".  It should be 1850.000 but the crystal oscillator used as 
the BFO in the 3586C are not stabilized and are a bit off frequency (I'm 
planning to rectify that by replacing them with a synthesizer driven 
from the reference oscillator).

John
----
> 
> /tvb
> 
> ----- Original Message ----- 
> From: "John Ackermann N8UR" <jra at febo.com>
> To: <time-nuts at febo.com>
> Sent: Monday, November 15, 2004 18:15
> Subject: [time-nuts] Some examples of HF frequency smearing`
> 
> 
> 
>>In preparation for the ARRL Frequency Measuring Test, yesterday Daun 
>>(N8ASB) and I set up our gear for a practice session.  We were working 
>>just after sunset here in Dayton, but before sunset at the WWV site in 
>>Boulder.
>>
>>I took some screenshots of the WWV carrier showing how much frequency 
>>variation took place due to propagation effects.  There was almost a 
>>Hertz of frequency shift over a half hour or so period.
>>
>>I put the screenshots at http://www.febo.com/time-freq/hf/index.html.
>>
>>John
>>
>>_______________________________________________
>>time-nuts mailing list
>>time-nuts at febo.com
>>https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>
> 
> 
> 
> 
> _______________________________________________
> time-nuts mailing list
> time-nuts at febo.com
> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> 
> 




More information about the Time-nuts_lists.febo.com mailing list