Speaking of 7-segment clock displays and such:
As it happens, I am in the process of making yet another time-of-day clock,
and it involves a 7-segment display. The inspiration is:
In the last couple of years I've become very interested in "aerospace
telemetry" equipment and history. Somehow, my collection of vintage
telemetry receivers has now grown to seven receivers, plus some related HW.
Details are here: https://spyradios.com/intercept.html .
Then I realized that I need a "countdown clock", like what might have used
in the 1960's in a launch control room.

SO, I'm working on a 3D-print design for a clock display, in the
"+HH:MM:SS" format. The digits are about 2" high in a 3" background, with
two 3mm LEDs per segment.
I haven't decided on what the firmware will do, other than displaying
time-of-day. I'm thinking that periodically (hourly??) it will switch to a
'countdown' mode with "T minus zero" being the top of the hour. With that,
maybe I should also build a light/sound effect for a rocket launch... :-) .
I may use the old Sulzer 5A as the time base.

I'll be happy to share all the details after I've finalized things, in case
anyone wants to leverage it for their own project.

Pete

On Wed, Mar 26, 2025 at 6:00 AM Sebastien F4GRX via time-nuts <
time-nuts@lists.febo.com> wrote:

 
While the SDR sticks (e.g. RTL-SDR) can receive the signal, their clocks aren't so wonderful. One could generate a 28.8 MHz clock with high quality.  There's also not a direct path from ADC to USB - it goes through some sort of digital FIR filter and is potentially resampled. Whatever that transformation winds up being, at least it's deterministic, so once you figure it out, it should be the same every time. 

On Wed, 26 Mar 2025 08:50:59 -0000, "john.haine--- via time-nuts" time-nuts@lists.febo.com wrote:

Geoff, I'm not sure where you are located but in Europe there is of course the DCF77 signal from Mainflingen at 77.5 kHz and the MSF signal at 60 kHz from Anthorn (UK). I am fairly certain that there are still products around that use these to derive a stable frequency and time source, and there have been a number of amateur projects published. With the growing concerns over GNSS jamming and spoofing there's increasing interest in e-LORAN at 100 kHz for PNT, there is a station at Anthorn transmitting, possibly a network being set up over the UK, some I believe in Europe and (ironically) in Russia. With all these VLF standards working in similar frequency ranges a multi-standard receiver ought to be possible which could apply diversity.

Cellular base stations are extremely stable and GNSS locked, though fixed PNT networks are being deployed to remove this reliance. A standard frequency receiver based on cellular base station transmissions would be challenging but ought to be possible. After all, handsets have to lock on to the network frequency!

Digital TV broadcasts must also be very stable I would guess. These should be much easier to demod and decode as the transmissions are continuous. Receivable through cheap USB SDR sticks.

A few years back at Bristol University we had a PhD group project looking at using pulsars as a frequency source, inspired by work in Poland that actually resulted in a Pulsar Clock in a museum. Not such a mad idea, except that you have to keep switching to another pulsar as the earth rotates. Also not very portable!

-----Original Message-----
From: Geoff Van der Wagen via time-nuts
Sent: 26 March 2025 00:57
To: Sebastien F4GRX via time-nuts
Cc: Geoff Van der Wagen
Subject: [time-nuts] Re: Quartz crystal frequency jumps

Hi Sebastien,

I share a similar philosophy, and would also like to work on self-contained options.

Unfortunately, I have far too many ideas and far too little time/energy to do them!

My current wish is to make a VCOCXO and be able to discipline it in different ways. Various options here are broadcast AM, WWV, or possibly even picking up JJY. WWV will be problematic because of ionospheric height changes and intermittent propagation. JJY is apparently possible from Australia but no idea of what reliability would be like. It would be fun to observe these signals with some confidence. Before GNSS there used to be off-air frequency standards that would synchronise to LF services.

I also started daydreaming about truly human-independent sources like tracking pulsars but it rapidly becomes unachievable within reasonable time and energy/motivation constraints.

Once I have a 10MHz reference I'd use that to run a hardware digital clock with 7-segment displays maybe down to the millisecond, just for fun (you'd need a high speed camera to see what time it is, but the act of reading the time that way totally defeats the purpose which I find amusing). Synchronising such a clock beyond NTP or GNSS becomes a whole exercise in itself. Realistically, GNSS is the best way to get accurate time. But buying a cheap GPSDO and plugging it in isn't very rewarding, nor do I learn anything useful.

I was impressed by the ability of WSJT-x to run a frequency calibration using various sources. Using it, I found out that some of the local AM stations are well-disciplined and others are free-running (a couple were considerably off-frequency, 2.5ppm or so while my FT-897D's TCXO ended up being around 1ppm off).

Cheers

Geoff VK2WA

On 13/3/25 20:24, Sebastien F4GRX via time-nuts wrote:

time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe send an email to time-nuts-leave@lists.febo.com

time-nuts mailing list -- time-nuts@lists.febo.com
To unsubscribe send an email to time-nuts-leave@lists.febo.com
 

In the 1980's when I worked for HP Santa Clara Division, they had three
dedicated 5061A high performance option cesiums that were steered by
Loran C signals.  My boss would brag that this was the most accurate
clock on the West Coast.  This was called the "House Standard" and was
distributed throughout the plant.  They had a setup to measure the
chroma carrier of various local TV stations, and they could tell them by
how much their Rb clock was in error.  (I believe the TV stations moved
on from that paradigm many years ago, so you can't use your TV for a
frequency reference any more).

Rick Karlquist
N6RK

On 2025-03-26 01:50, john.haine--- via time-nuts wrote:

If for no other reason than broadcast TV is now all digital, so there's no color burst or subcarrier to measure.
The transmitted frequency might be fairly accurate (so they can do multiple transmitters on the same frequency - to a receiver, it just looks like multipath, and it uses a RAKE receiver to deal with it).
But that probably wouldn't give you "time".
 

On Wed, 26 Mar 2025 11:17:32 -0700, Richard Karlquist via time-nuts time-nuts@lists.febo.com wrote:

In the 1980's when I worked for HP Santa Clara Division, they had three
dedicated 5061A high performance option cesiums that were steered by
Loran C signals. My boss would brag that this was the most accurate
clock on the West Coast. This was called the "House Standard" and was
distributed throughout the plant. They had a setup to measure the
chroma carrier of various local TV stations, and they could tell them by
how much their Rb clock was in error. (I believe the TV stations moved
on from that paradigm many years ago, so you can't use your TV for a
frequency reference any more).

Rick Karlquist
N6RK

On 2025-03-26 01:50, john.haine--- via time-nuts wrote:

time-nuts mailing list -- time-nuts@lists.febo.com
To unsubscribe send an email to time-nuts-leave@lists.febo.com
 

Correction to Jim's statement that DTV signals do not
have subcarriers:

ATSC 1.0:  There is a continuous pilot carrier near the
bottom end of the overall spectrum.

ATSC 3.0:  There are pilot tones all over the place.
While most would be very difficult to use,
two of them (one at the bottom and one
at the top) are easy to use.  Each is
interrupted very briefly at a rate of 4X per
second, and a clean uninterrupted tone
can be obtained by merely selecting that
clean carrier with a BPF narrow enough to
block the sidebands.  Caveat: these tones
track variations in the transmitting station's
RF frequency.  For the two stations within
indoor dipole receiving range of my home,
I see frequency variations in the neighborhood
of 50 mHz (that's milli-Hz) P-P.  Both stations
(KMYS & KPXL) operate at roughly 500 MHz
regime, and had roughly comparable stability,
although their detailed waveforms differed.  I
think this could offer insight into the nature of
their frequency references.

Dana
Kerrville, TX

                variations of

On Thu, Mar 27, 2025 at 11:59 AM Jim Lux via time-nuts <
time-nuts@lists.febo.com> wrote:

The Tx frequency (at least for the DVB-T2 standard) is specified to be within 1e-9 for single frequency networks.  I think that it's highly likely that most DVB transmitters will use GNSS to provide a local frequency reference just like cellular does, so this requirement is probably easily met.  The modulation is COFDM so multipath handling doesn't require a RAKE.  The receiver will need to accurately recover a reference clock locked to the incoming signal for the COFDM demod to work, so I think that it would be possible to get "frequency" from the modem.  "Time" signals are at least included in the data multiplex in a standardised way I believe.

Digital audio broadcasting should also be a potential source for time and frequency.  It also uses (in Europe) OFDM and single frequency networks are more common I think.

-----Original Message-----
From: Jim Lux via time-nuts time-nuts@lists.febo.com
Sent: 27 March 2025 16:36
To: time-nuts@lists.febo.com
Cc: Jim Lux jim@luxfamily.com
Subject: [time-nuts] Re: Quartz crystal frequency jumps

If for no other reason than broadcast TV is now all digital, so there's no color burst or subcarrier to measure.
The transmitted frequency might be fairly accurate (so they can do multiple transmitters on the same frequency - to a receiver, it just looks like multipath, and it uses a RAKE receiver to deal with it).
But that probably wouldn't give you "time".

On Wed, 26 Mar 2025 11:17:32 -0700, Richard Karlquist via time-nuts time-nuts@lists.febo.com wrote:

In the 1980's when I worked for HP Santa Clara Division, they had three dedicated 5061A high performance option cesiums that were steered by Loran C signals. My boss would brag that this was the most accurate clock on the West Coast. This was called the "House Standard" and was distributed throughout the plant. They had a setup to measure the chroma carrier of various local TV stations, and they could tell them by how much their Rb clock was in error. (I believe the TV stations moved on from that paradigm many years ago, so you can't use your TV for a frequency reference any more).

Rick Karlquist
N6RK

On 2025-03-26 01:50, john.haine--- via time-nuts wrote:

time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe send an email to time-nuts-leave@lists.febo.com

time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe send an email to time-nuts-leave@lists.febo.com

On the DVB

In the case of local 'translators' or 'gap fillers' as we used to call
them... Those transmitters may well be a local OCXO, and not tied to
anything, although, where phase noise is not the strictest
consideration, but more so frequency accuracy , I have found that its
cheaper to put in a $20 XO and and $12 GPS receiver chip .... rather
than a $200 XO or Rb.

-glen

On 28/03/2025 05:29, john.haine--- via time-nuts wrote:

There is a proposal to distribute time over ATSC 3.0 with some
modifications in the signal.

Measurements have been done by NIST and it's gives ns stability. You
need to compensate for offset delay, but you can do that through several
methods.

It will solve some, but not all needs.

Cheers,
Magnus

On 2025-03-27 18:28, Dana Whitlow via time-nuts wrote:

Hi,

This is a topic right up my alley, as I've worked with time-transfer to
DVB-T and DAB SFN networks.

The 1E-9 specification is more technical than anything else, it needs to
be that stable at the least for proper signal quality. This does not
reflect in any way the accuracy you will get, as it will in practice be
much better. The original design assumed GPS receivers (they had not
taken the shift over to denote it GNSS), but in practice many countries
operating SFN mode made their network GNSS-independent, because they
treat TV signals as a critical infrastructure that needs to function.

Similar goes for DAB, which when run in SFN also goes to COFDM from the
usual OFDM. You have the same background timing network typically. Some
DAB transmitters may operate independent of national resources thought,
so it may be not as robust.

Both these is typically operated with 1-5 us allowance of transmitter
time error. In practice within 1 us is often achieved.

Cheers,
Magnus

On 2025-03-27 19:29, john.haine--- via time-nuts wrote:

However, when the "gap filler" is in an SFN region, it is not a
translator and needs to obey SFN rules. This can be done either through
a simplified retransmission or a full SFN DVB-T modulator. Then the gap
filler locally improves the signal without using frequency space, which
is a very compelling reason to do SFN as gap fillers can be used more
generously as needed.

Cheers,
Magnus

On 2025-03-28 02:07, glen english LIST via time-nuts wrote: