Do you think people would like to have a DDS VFO whose frequency and
phase were both direcly related to a common time/frequency standard
like WWV? That is, two people anywhere in the world using the VFO
could make a signal having exactly the same phase as measured at the
transmitter. I would adjust for lightspeed lag using a GPS receiver.

I am thinking this could be useful for coherent CW work.

Opinions?

The Eternal Squire

Do you think people would like to have a DDS
VFO whose frequency and phase were both
direcly related to a common time/frequency
standard like WWV?

Well, no existing communication standards require independent phase
coherence between transmitter and receiver, for the simple reason that
the path length is randomly fluctuating (and there is also multipath).

As an example, if it's 3000 miles from me to you, the actual path for
HF communications is going to be somewhere between 3100 miles (a single
bounce) and 3300 miles (several bounces and a little sideways jog) or
more. With wavelengths like 20meters or 40meters, that 200 miles
represents many many wavelengths.

Tim.

wrote in message
ups.com...
Do you think people would like to have a DDS VFO whose frequency and
phase were both direcly related to a common time/frequency standard
like WWV? That is, two people anywhere in the world using the VFO
could make a signal having exactly the same phase as measured at the
transmitter. I would adjust for lightspeed lag using a GPS receiver.

If you already have a GPS receiver, you can lock to their time code -- this is
already commonly done in, e.g., cell phone base stations to keep everyone in
sync as you move between cell sites. It's also a popular way to get a cheap,
high precision 10MHz reference clock for spectrum analyzers and other lab
gear -- TAPR has an inexpensive board that does this.

---Joel Kolstad

Well actually, the project I had in mind was to take the 60 khz WWVB
signal and use diode harmonic multipliers to create a clock in the
gigahertz range. I would then use this clock as the input to a DDS
system. I could use the GPS to calculate lightspeed delay between
the VFO and the WWVB to determine compensating phase lag.

This could be good for synchronizing the transmit and receive ends of a
digital communication, improving the signal to noise ratio.

The Eternal Squire

It would be very useful. Keep in mind that all oscillators in the system
would need to be coherent so a heterodyne conversion transceiver design
would need to have any other LO's synched as well.

An idea that I have had is the prize for transatlantic 2 meter DX. If
multiple stations stateside had same coherent carrier and were keyed (or
modulated) coherently, there would be space diversity in the transmitted
signal.

Joe K4SAT

wrote:

Do you think people would like to have a DDS VFO whose frequency and
phase were both direcly related to a common time/frequency standard
like WWV? That is, two people anywhere in the world using the VFO
could make a signal having exactly the same phase as measured at the
transmitter. I would adjust for lightspeed lag using a GPS receiver.

I am thinking this could be useful for coherent CW work.

Opinions?

The Eternal Squire




--
Joe Leikhim K4SAT
"The RFI-EMI-GUY"

"All Righty Then"

I think you'll have to find a way to synchronize the startup ('reset')of all
the DDS's. If you don't, they'll all start clocking at different times. The
output doesn't come for some number of clock cycles later, depending on
what's loaded into the counters. Because of this, you'll have "phase lock"
but not "phase coincidence". Synchronizing DDS's is described by Analog
Devices in an appp-note on generating quadrature signals using 2 DDS chips.

Joe
W3JDR


wrote in message
oups.com...
Well actually, the project I had in mind was to take the 60 khz WWVB
signal and use diode harmonic multipliers to create a clock in the
gigahertz range. I would then use this clock as the input to a DDS
system. I could use the GPS to calculate lightspeed delay between
the VFO and the WWVB to determine compensating phase lag.

This could be good for synchronizing the transmit and receive ends of a
digital communication, improving the signal to noise ratio.

The Eternal Squire

wrote:
Well actually, the project I had in mind was to take the 60 khz WWVB
signal and use diode harmonic multipliers to create a clock in the
gigahertz range. I would then use this clock as the input to a DDS
system. I could use the GPS to calculate lightspeed delay between
the VFO and the WWVB to determine compensating phase lag.

This could be good for synchronizing the transmit and receive ends of a
digital communication, improving the signal to noise ratio.

Wow... multiplying a very noisy 60kHz WWVB signal up to the gigahertz
using diode multipliers. Astonishing. Unless you live in the shadow
of Boulder CO, WWVB will be so noisy that after multiplication
you just have a huge broad hash of frequencies. It's buried in the
hash of noise most of the time anyway. And here on the East
Coast, WWVB reception is only barely usable for a few hours in the very
early AM (and not even that if there's a broadband LF noise source
anywhere in the vicinity.)

Receiving WWV (2.5, 5, 10, 15, and 20MHz) or WWVH is not quite so
problematic (usually in the continental US one of those frequencies
is receivable) but still not easy. And in say Asia or Europe or
South America most of those frequencies will not be open most of the
time, and there are local time/freq stations on those same frequencies
often. But multiplying what is usually a noisy carrier up to the
GHz via diode multipliers is still completely unthinkable.

What *is* done is to lock a high-stability OCXO to a time/freq
standard. This is often used for extremely-narrow-bandwidth moonbounce
communications for example. In the past decade almost all of this
is done by locking to GPS satellites, but if WWV or WWVB is receivable
at both ends this has been done too.

You do have the right spirit (low-bandwidth-for-low-noise requires
an accurate frequency standard) but multiplying WWVB into the GHz
using diode multipliers is decidedly *not* the way to do it. OCXO's
are cheap and GPS receivers are cheap and GPS is receivable everywhere
around the world and PLL's are so easy to do, so that's what
we use now. Before that PLLing a crystal oscillator to WWV/WWVB (
even if the PLL was a guy's ear listening for zero-beat and a
screwdriver
to tweak the oscillator fine-tuning) was
reasonable but there you're up against propogation and varying
path-length much of the time.

Then how come wall clocks based on WWVB work so well anywhere in CONUS?

wrote in message
oups.com...
Then how come wall clocks based on WWVB work so well anywhere in CONUS?

They don't -- there are plenty of places in the U.S. where it takes a fair
amount of experimentation to get them to sync at all. Most of them 'listen'
in the wee hours of the morning when propagation tends to be better too. And
finally, even if they do only manage to sync up 'once in a while' (say every
week or two), they still perform much better than a typical wall clock.

wrote in message
oups.com...
Well actually, the project I had in mind was to take the 60 khz WWVB
signal and use diode harmonic multipliers to create a clock in the
gigahertz range. I would then use this clock as the input to a DDS
system. I could use the GPS to calculate lightspeed delay between
the VFO and the WWVB to determine compensating phase lag.

This could be good for synchronizing the transmit and receive ends of a
digital communication, improving the signal to noise ratio.

The Eternal Squire



I think you'd have a difficult problem with the variable phase due to
multipath causing shifting path length between both ends of the digital
communication.

That is why systems may use time based coherancy for data start and stop
times, and so on, but do not bother to try for radio carrier frequency phase
coherant systems.


Jim
N6BIU