[time-nuts] WWVB Measurements

[Douglas H Reed]

I'll risk exposing my ignorance to the list....  (i.e. open my mouth
and prove it...)

I seem to recall a comment on the LOWFER list recently about the data
decodes being much worse the previous night and a reply saying it was
probably due to storms moving into the area between the LOWFER beacon
and the receiving station. And in general the LOWFER activity tends to
die off in the summer months due to thunder storm activity making it
more difficult to hear anything at all. If we apply the same comments
to WWVB, it is likely that the past few winter months are the prime
time for relatively undisturbed WWVB reception and the coming summer
months will be more difficult and a better measure of how good the
receiver design might be.

And IIRC, the best technique for use of LORAN was to make your
readings at about the same time of the morning every day, preferably
before sun-rise and the morning path disturbances took place. And I
seem to recall a discussion on this list recently about using the time
stamp info to verify which second you are measuring so you could lock
an oscillator against a signal that wasn't always available. You had
to keep a time stamp of the oscillator you are disciplining so you
could compare it to WWVB or GPS time stamps after a long outage and
verify you were measuring the same second and how far you had drifted.

By extending that idea you could compare time stamps, measure the WWVB
1PPS differences for an hour or two to verify the standard deviation
and average (median?) error and maybe watch the RF signal levels. This
might help to identify disturbed conditions that might be causing
errors in the received signal decode and timing. If the signal
variations or SD indicate conditions are disturbed, continue hold-over
of the oscillator, no adjustment. If signals are good, you have enough
long(ish) term data to make an adjustment to the oscillator and go
back to hold over until the next test period is complete. I'm assuming
there is nothing new in this procedure since it is basically what was
done manually with LORAN or WWVB.

I would expect that somebody has already done a study to identify the
best time of day and methods to use with WWVB for precision timing but
I haven't done a search to find it. There are probably quite a few
people on this reflector who would have the equipment and software to
capture this sort of data and collect it centrally for further
examination. It could be interesting if it hasn't been done already.

Finally, I expect that using DSP for recovering the WWVB carrier
information below the noise would be fairly successful. I seem to
recall that knowing the expected phase and frequency of the signal is
a primary advantage for using DSP to dig way under the noise. I would
expect good DSP to add 20dB to the margin of any simple analog WWVB
receiver. I wonder if this might be an application for crossed loops
and stereo sound card with 96KHz sample rate. Could leak a bit of
50KHz from your 10MHz oscillator so the software has a known signal
for comparison if needed. You could at least extract a lot of signal
quality info with the DSP and processing available in the computer.

I'm glad that someone else is working on this (not me) and I'll be
very happy to read your results when done. Now I think I'll go back to
hiding in the weeds.....

73, Doug.
====================================
Date: Sat, 26 Mar 2011 18:53:30 -0700 (PDT)
From: "J. Forster" <jfor at quik.com>
To: "Discussion of precise time and frequency measurement"
       <time-nuts at febo.com>
Subject: Re: [time-nuts] WWVB Measurements

Paul,

I did that years ago in a qualitative way.

I used a local reference of an HP 105B and the 117A w/ HP loop. I synced a
Tek 7854 to the local standard and watched and averaged the recovered 60
KHz. I no longer remember the results well, but sometimes the "seeing" was
<1uS P-P and on other days it was at least 10s of uS to hundreds of uS.

This may be hard for some to believe, but the seeing seemed to depend on
weather. A big front coming east would make the "seeing" much worse. This
was just a casual observation. I never spent much time studying it. If I
saw the seeing was horrid, I just left the stuff running and went away.