CCWN 75:22
                                                   
WAYS OF SETTING THE PHASE OF CCW KEYING & FILTERING Chas. Woodson

          The improved effectiveness of CCW is due to the
advantage of  knowing the frequency, pulse length and pulse phase
of the signal one  desires to detect.             Even without
phase lock, the stability of the frequency standards  we use will
keep CCW filters in phase for a long time. Four our .1  second
pulses, and an accuracy of 10-7 for the standard (which 
corresponds to 1 cycle error between the stations' standards in
their  adjustment to the 10 mc WWV signal), it would take hours
for a 108  pulse phase error to develop between the two stations.
           The Petit filter does not actually "know" the timing
of incoming  signals. In other words, it does not "know" when a
pulse will start,  although it "knows" the frequency of the
desired signal and the length of  pulses. Dependable reception is
obtained by adjustment of phase during  reception. Due to the
high precision of the frequency standard, once  adjusted, phase
relationships hold for a relatively long time. Although we  don't
yet know from experience, I am concerned that under weak signal 
conditions this adjustment may be very difficult to obtain.       
    Several years ago I built a system in which the pulse phase,
as  well as frequency and pulse length was accurately known. It
was based  upon synchronization with WWV and JJY. Both characters
and pulses were  adjusted to start at the beginning of a minute.
The bit length was 10  seconds, which is 100 times the length
used by our current CCW system,  but the technique used might
still be applicable to our current CCW  system.            
Applying this logic to current CCW efforts, the second timing 
pulses from WWV might be used. An oscillator operating at 10 cps,
could  be divided by ten and phase locked on the second pulses of
WWV. This  would give us our desired .1 second pulses, but would
also have the  desired property that operators throughout the
world would have their  signals in phase. This appears feasible
but phase lock loops at this low  frequency may not be easy to
construct. Interference from transmitter  during transmissions
may also be a problem.                                   an ,     
                     This suggests an interesting operating
possibility. Suppose each  station, when beginning operation for
the day, fed a mix of the wwv audio  and their keying (e.g., a
string of dots) into the vertical input of a  scope with the
horizontal sweep frequency set to lock at, say, 10  cycles. The
phase of the keyed signal could then be adjusted to begin a 
pulse at the same time a second begins. Once adjusted at the
start of  operation, phase adjustment should be accurate for
several hours.  Following such a procedure, all CCW stations in
the world would have the  same phase, and the need for adjustment
on reception would be  eliminated.