[R-390] WWV Frequency Standard

[Bill Hawkins]

Roger, you've written some excellent stuff for this list, but I'm a
time-nut so I have to disagree with your first sentence.
 
The relative time between cycles is equal to 1 divided by the frequency
of the cycles. Time and frequency are exactly related. You did say
absolute time, and that's a different story. Absolute time was
completely independent of frequency until the atomic clock came along.
 
 
Absolute time was related to an image of a star passing through the
cross-hairs of a telescope. It had nothing to do with frequency. This
produced a pulse that marked the time of day in terms of rotation of the
Earth. The pulse was distributed by telegraph and then radio. In between
star crossings, counters were used to measure elapsed time from the last
crossing. The oscillators that made the counters count were adjusted so
that there was negligible error between the counter and each crossing.
Of course, it wasn't that simple because the Earth goes around the sun.
 
Since 1972, when UTC replaced GMT, absolute time of day is reckoned from
the oscillation of Caesium atoms (cesium for you CPS fans) because there
was too much noise in the star crossing detection for modern science.
You don't have to wait for star crossings. You can listen to WWV or a
set of GPS satellites and adjust your clock's oscillator to match as
closely as the laws of physics and probability allow. It's possible to
match within 1 part in 10 to the -15 power (14 zeroes before the 1), but
that gets expensive.
 
Matching the frequency is necessary for long term accuracy, but the
counter has to start from some number that is derived from counting the
9 GHz frequency of very cold Caesium atoms since 1972 - not a trivial
task. In fact, there is no single time standard. Absolute time is
derived from an average of 50 or more atomic clocks. Leap seconds are
introduced because the Moon slows the rotation of the Earth with tidal
forces. Fortunately, WWV and GPS also broadcast the correct time.
 
You can do this at home with a GPS receiver that sends data to a
computer to decode the data stream. I've got an HP Z3801A GPS receiver
and the SatCom program running under Windows 98, which synchronizes the
other computers on my network. I've got some 1970 vintage time code
receivers that could be used for station clocks, but aren't.
 
 
Ah, but this is more than most people want to know. Roger is right -
R-390 class sets don't need that kind of accuracy. The run-out on a
carefully adjusted PTO may be 0.1 KHz, which is a tenth of the digital
dial resolution. You can find WWV on the dial and then switch to CAL and
adjust the local crystal to match. Now CAL can be used to find 100 KHz
points on the dial. This gets you close enough for all practical
porpoises.
 
 
Also, I get 2 hits on Google for kebelsa (Perry, not Roger) and 75,000
for kielbasa sausage. I don't know how to pronounce kielbasa in Polish
(or French). Maybe they sound the same.
 
Bill Hawkins

  _____  

From: Flowertime01 at wmconnect.com
Sent: Saturday, December 06, 2008 10:33 AM
To: r-390 at mailman.qth.net
Subject: Re: [R-390] WWV Frequency Standard


Fellows, 

There is a difference between absolute time and frequency. 

Count time ticks from WWV and time ticks from your local standard. 

After six months or so review the difference on your two counters. 
Then adjust your local time standard up or down a bit. 

You frequency standard is within one 1 second time tick over six months.

I suggest that is close enough for government work. 

In 1975 in Okinawa we would trigger two counters and let it run 24
hours. 
We expected zero difference. We perforned this test as a monthly
maintenance procedure. This was just for a frequency standard. 

For time we used a dual trace scope and an off set time delay number
provided by the military to us. There was a document with lots of
locations and expected time delay. 

Again we measured the delay with a dual trace scope and averaged the
delay over lots of time (weeks). The "clock" had an offset in it. The
frequency standard tick edge was any where it wanted to be. The clock
would let us have any number of steps up to 1,000,000 between the
frequency edge and the time edge. Then you could set the time display to
tick any second from the time edge. 

In the Army with lots of techs working 24 X 7 you could put 6 or 8 guys
on the job of guarding the "maintenance procedure" for 3 or 4 days each
month. You did not watch it all shift but you set up the procedure and
passed it from shift to shift. 

In the Army you just did not clown around with some activity going on.
Hundreds of guys and girls worked in the radio rooms and just left
things alone for ever and ever. 


You can have a standard in your shop that is closer than a RCH. 

And then I ask you what for? 

The American Military has run planet wide RF communications and
reportedly off planet communications with R390's, TS505s, TS352s, TEK
505s, and AN/URM25's since the 1950s. 

What are you doing in your hobby that needs better resolution than what
you can get with that equipment? 

You cannot prepare kebelsa sausage with better resolution. 

Roger AI4NI 







-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.qth.net/pipermail/r-390/attachments/20081206/d09ec8db/attachment-0001.htm