[time-nuts] WWVB Dephaser Question

Sat Oct 10 00:03:38 UTC 2020

Direct IQ demodulation to zero-IF is indeed subject to problems from LO
leakage into the
mixers.  Said leakage will displace the center of the IQ plot away from the
origin, which
then requires correction later on, which is a pain to do well.

A pretty good way to deal with all this is to downconvert the 60 kHz to a
lower IF (say,
a few hundred Hz or maybe even a few kHz) in a conventional mixer, then
clean up
the garbage with an analog filter.  Next digitize the IF (now a quite slow
A/D will suffice),
then do quadrature demodulation to zero-IF in SW.  Once there, use the
ATAN2 function
to obtain phase in the time domain.  The catch is this: that 1st LO must be
right on, hence
phase locked, or else let it drift around a bit as it pleases and phase
lock the 2nd (SW)
LO to obtain the locked zero-IF.  This kind of stuff is pretty easy to do
if you are doing it
in post-processing from a recorded signal.  But I can see it getting rather
more involved
when one wants to do it in real time, something I've never tackled.  Good
luck!

Dana

On Fri, Oct 9, 2020 at 5:56 PM <rcbuck at atcelectronics.com> wrote:

> Paul,
> I was seeing a -10 dBm on the SA when they enter high power mode at
> night. At low power they are only a couple of dB above the nearby noise
> makers. But that is at the output of the last stage of my receiver. The
> WWVB signal was about 10-15 dB above the noise makers in my work shop. I
> don't know what the actual WWVB signal level is and have no way to
> measure it. I estimate the ferrite antenna to be approx 20k ohm which
> doesn't match the 50 ohm input of the SA. Maybe feed the rod antenna
> into a single stage op amp with a gain of 1 to feed the SA would work
> give me a rough idea?? I previously said I had 1 mV of daytime signal
> according to the WWVB maps. But that was wrong. In my mind I was
> thinking 1000 uV when the maps show the 100 uV coverage area. So receive
> level during the day should be around 0.1 mV. I have PCB's on order that
> hopefully will improve the receiver performance.
>
> Bob,
> I'm still trying to figure out if a software or hardware solution is
> better. I've looked at the Costas loop and a PLL angle. But I run into
> the problem of the LO being on the same frequency as WWVB. Plus as Paul
> said the math is intimidating to me. Locking WWVB to a HF oscillator and
> dividing down to 10 kHz or so for the control voltage and I/Q
> demodulation may be an answer but involves more hardware.
>
> For a software solution I don't know exactly how to approach the task. I
> don't have any DSP software experience so that option is out. I haven't
> made a decision on the CPU yet. I will just use whatever it takes for
> the job, probably a STM32 part of some sort. I have a couple of STM32
> development boards I can use when I get to that point.
>
> Ray
>
>
> -------- Original Message --------
> Subject: Re: [time-nuts] WWVB Dephaser Question
> From: Bob kb8tq <kb8tq at n1k.org>
> Date: Fri, October 09, 2020 12:23 pm
> To: Discussion of precise time and frequency measurement
> <time-nuts at lists.febo.com>
>
> Hi
>
> Since you need a MCU to decode the data, you might as well get
> things in there sooner rather than later. There are a number of MCU’s
> out there that have fast enough ADC’s to do the job. They do have
> limited
> dynamic range. You can go to one of the 24 bit converters and have
> a ton of dynamic range. It’s all up to you.
>
> Once it’s into the MCU, it’s just code :). Depending on how much cpu
>
> horsepower you have (and how much code you want to write) you can
> go more or less crazy …. PHK has some interesting tidbits on his web
> site. http://phk.freebsd.dk/AducLoran/AducLoran-0.0.pdf
> <http://phk.freebsd.dk/AducLoran/AducLoran-0.0.pdf>
>
> Bob
>
>
> > On Oct 9, 2020, at 1:38 PM, rcbuck at atcelectronics.com wrote:
> >
> > Paul, Bob,
> >
> > I am not using any commercial receiver. I am building everything from
> > scratch. The RF front end starts with a ferrite rod antenna feeding a
> > differential first op amp followed by 5 stages of op amp filtering and
> > amplification. When the last stage is fed to my spectrum analyzer
> > (through attenuators) the WWVB signal is clearly visible. I'm now trying
> > to figure out how to detect the phase shift so I can get the time data
> > for my CPU to process and send to a display.
> >
> > I already have a GPS based clock that I built so I thought the WWVB
> > phase clock would be an interesting project.
> >
> > Ray
> >
> > -------- Original Message --------
> > Subject: Re: [time-nuts] WWVB Dephaser Question
> > From: Bob kb8tq <kb8tq at n1k.org>
> > Date: Fri, October 09, 2020 7:35 am
> > To: Discussion of precise time and frequency measurement
> > <time-nuts at lists.febo.com>
> >
> > Hi
> >
> > At least to me, anything dimensioned in the 100’s of feet is
> > “massive” compared to
> > the rod antennas normally seen in WWVB use ….
> >
> > The other point being that if the antenna is some sort of large loop,
> > it’s going to be
> > a good long ways away from the receiver. You get both a larger signal
> > voltage and better
> > isolation …..
> >
> > Bob
> >
> >> On Oct 8, 2020, at 11:30 PM, John C. Westmoreland, P.E. <
> john at westmorelandengineering.com> wrote:
> >>
> >> Hello All,
> >>
> >> Are there any design details someplace regarding these massive antennas?
> >>
> >> Thanks,
> >> John
> >> AJ6BC
> >>
> >>
> >> On Thu, Oct 8, 2020, 19:27 paul swed <paulswedb at gmail.com> wrote:
> >>
> >>> Hello to the group.
> >>> Ray as Bob mentions you are taking a 10s of uv signal to a logic level
> of
> >>> maybe 4V.
> >>> If the loop is any place close to the divided down signal, it will
> >>> oscillate. It would take incredible shielding to protect the receiver.
> >>> Thats why you often see a solution that doubles to 120 KHz and
> modifies the
> >>> detectors to work at that frequency. That means hacking the radio
> >>> internally. Not fun. The other really annoy effect is that the doubling
> >>> slips phace due to noise and propagation. So if charting suddenly you
> get a
> >>> 180 degree flip. Thats messy.
> >>> The doubling solution can work. Search for carter and there are several
> >>> others.
> >>> But having tested and used all of the alternates and lots more on the
> east
> >>> coast decided they were too much trouble. You should see the box of
> boards
> >>> I have chuckle.
> >>> For me I am very happy with the d-psk-r. Though in being above board I
> >>> designed version 1 and Rodger and I did version 2. Its solid and no
> mods to
> >>> any receiver. Everything has always been released to the time-nuts
> group.
> >>> As they say have fun.
> >>> Regards
> >>> Paul.
> >>> WB8TSL
> >>>
> >>> On Thu, Oct 8, 2020 at 5:39 PM <rcbuck at atcelectronics.com> wrote:
> >>>
> >>>> Bob,
> >>>>
> >>>> I am using a ferrite rod antenna for the receiver. No outside antenna.
> >>>>
> >>>> Ray
> >>>>
> >>>> -------- Original Message --------
> >>>> Subject: Re: [time-nuts] WWVB Dephaser Question
> >>>> From: Bob kb8tq <kb8tq at n1k.org>
> >>>> Date: Thu, October 08, 2020 12:40 pm
> >>>> To: Discussion of precise time and frequency measurement
> >>>> <time-nuts at lists.febo.com>
> >>>>
> >>>> Hi
> >>>>
> >>>> A lot depends on your antenna setup. You can also swamp out the
> incoming
> >>>> WWVB signal…….
> >>>>
> >>>> Bob
> >>>>
> >>>>> On Oct 8, 2020, at 2:07 PM, <rcbuck at atcelectronics.com> <
> >>>> rcbuck at atcelectronics.com> wrote:
> >>>>>
> >>>>> I have read several different articles where the WWVB phase shift is
> >>>>> eliminated by doubling the signal to 120 kHz. Several members of the
> >>>>> list have built these units.
> >>>>>
> >>>>> Assume I build a circuit to double the incoming signal and use a
> >>> schmitt
> >>>>> trigger to get a 120 kHz square wave. If I then divide that signal
> back
> >>>>> down to 60 kHz will that signal be strong enough to swamp out the
> WWVB
> >>>>> signal? I'm guessing it will be since it is at the 5 volt level and
> >>>>> somewhere in the +25 dBm or greater range.
> >>>>>
> >>>>> Ray,
> >>>>> AB7HE
> >>>>>
> >>>>> _______________________________________________
> >>>>> time-nuts mailing list -- time-nuts at lists.febo.com
> >>>>> To unsubscribe, go to
> >>>> http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> >>>>> and follow the instructions there.
> >>>>
> >>>>
> >>>> _______________________________________________
> >>>> time-nuts mailing list -- time-nuts at lists.febo.com
> >>>> To unsubscribe, go to
> >>>> http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> >>>> and follow the instructions there.
> >>>>
> >>>> _______________________________________________
> >>>> time-nuts mailing list -- time-nuts at lists.febo.com
> >>>> To unsubscribe, go to
> >>>> http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> >>>> and follow the instructions there.
> >>>>
> >>> _______________________________________________
> >>> time-nuts mailing list -- time-nuts at lists.febo.com
> >>> To unsubscribe, go to
> >>> http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> >>> and follow the instructions there.
> >>>
> >> _______________________________________________
> >> time-nuts mailing list -- time-nuts at lists.febo.com
> >> To unsubscribe, go to
> http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> >> and follow the instructions there.
> >
> >
> > _______________________________________________
> > time-nuts mailing list -- time-nuts at lists.febo.com
> > To unsubscribe, go to
> > http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> > and follow the instructions there.
> >
> > _______________________________________________
> > time-nuts mailing list -- time-nuts at lists.febo.com
> > To unsubscribe, go to
> http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> > and follow the instructions there.
>
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe, go to
> http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> and follow the instructions there.
>
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe, go to
> http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> and follow the instructions there.
>