learn-by-doing - GPSDO
Hello to the Group,
From a never having done a home-built GPSDO before... I would
think a home built GPSDO would be quite the involved software
undertaking. Forgive me if this has already been addressed and
I've obviously missed the previously posted info.
Just thinking out loud, unless the micro-controller considered and
uses all the important GPS information... would not your at sea
level 10 MHz zero start/crossing value be some value different from
my 6,000 Feet (1.8288 Kilometers) value. Precision is possible, but
accuracy would not... without including the differences in elevation
and probably other over looked (by me in this reply) items.
regards,
Skipp
Today's Topics: 1. Re: Looking for simple/beginner's design guide for
experimental > construction of GPSDO (1 PPS controlling 10 MHz).
From: chris elfpen.com chris@elfpen.com
Subject: [time-nuts] Re: Looking for simple/beginner's design guide
for experimental construction of GPSDO (1 PPS controlling 10 MHz)
I acquired the board and built the implementation from the attached
QEX article. It seems to work. I know it has been discussed on this
board quite a bit and has it's faults.
From: wb0gaz--- via time-nuts time-nuts@lists.febo.com
I'm curious to learn-by-doing about GPSDO architecture/implementation,
starting with a GPS receiver module (on hand already) that has 1 PPS
output signal.
I suspect this can be a complex undertaking and ill-advised ("why
would you build when you can buy?"), but this is about
experimentation for fun.
My starting point idea is this - a 10 MHz OCXO (which can be
frequency-trimmed by a small DC voltage) feeding a 24-bit counter
whose count is read by a microcontroller, then reset and re-armed
ready for the next the 1PPS signal from GPS module.
Looking for a pointer to a write-up (PDF or web page preferred,
I've never enjoyed trying to learn from youtube videos) that would
help guide my experimentation.
The outcome would certainly not need to be high-specification;
just something that implements 1 PPS leading to control over a
10 MHz oscillator using a microcontroller would suffice.
Thanks for any reading resource pointers,
73 Dave WB0GAZ wb0gaz@yahoo.com
Hi Skipp,
True, signals from satellites take less time to reach you at 6000 ft
elevation compared to someone closer to sea level. But since your GPS
receiver knows exactly where your antenna is, that time difference due
to elevation is taken into account by the receiver's firmware. As a
result the 1PPS output of every GPS receiver is synchronized to UTC and
the user has to do nothing extra to get both precision and accuracy.
This is why DIY GPSDO are easy to make: not much more than an adjustable
oscillator and a GPS board. In fact you don't even need a
microcontroller. One of the better performing DIY GPSDO uses an analog PLL:
description:
http://jrmiller.online/projects/ministd/frqstd0.htm
http://jrmiller.online/projects/ministd/frqstd.htm
performance:
http://leapsecond.com/pages/gpsdo/
https://www.febo.com/pages/oscillators/g3ruh_gpsdo/
/tvb
On 5/7/2025 10:19 PM, skipp isaham via time-nuts wrote:
Hello to the Group,
From a never having done a home-built GPSDO before... I would
think a home built GPSDO would be quite the involved software
undertaking. Forgive me if this has already been addressed and
I've obviously missed the previously posted info.
Just thinking out loud, unless the micro-controller considered and
uses all the important GPS information... would not your at sea
level 10 MHz zero start/crossing value be some value different from
my 6,000 Feet (1.8288 Kilometers) value. Precision is possible, but
accuracy would not... without including the differences in elevation
and probably other over looked (by me in this reply) items.
regards,
Skipp
Hello,
in practice, you do not need to thing about those things, because all
necessary corrections are done by the GPS receiver firmware.
Normally, you use the GPS receiver for GPSDO in "position fix" mode,
where longitude, lattitude and high above sea level are known. Modern
GPS receiver can switch to this mode automatically, you just need to
leave the GPS antenna at the same place for several hours or days. Of
course, you can switch to this mode manually (mostly via RS-232/USB port
on the HPS receiver) if your position is known - for example from Google
Maps.
In this mode, the 1PPS is always precise and accurate at the sea level
or at the top of Mount Everest. The precision and accuracy are of course
dependent on the precision of GPS receiver internal oscillator,
satellite status, ionosphere etc.
Best regards
PavelK
Dne 08.05.2025 v 7:19 skipp isaham via time-nuts napsal(a):
Hello to the Group,
From a never having done a home-built GPSDO before... I would
think a home built GPSDO would be quite the involved software
undertaking. Forgive me if this has already been addressed and
I've obviously missed the previously posted info.
Just thinking out loud, unless the micro-controller considered and
uses all the important GPS information... would not your at sea
level 10 MHz zero start/crossing value be some value different from
my 6,000 Feet (1.8288 Kilometers) value. Precision is possible, but
accuracy would not... without including the differences in elevation
and probably other over looked (by me in this reply) items.
regards,
Skipp
Today's Topics: 1. Re: Looking for simple/beginner's design guide for
experimental > construction of GPSDO (1 PPS controlling 10 MHz).
From: chris elfpen.com chris@elfpen.com
Subject: [time-nuts] Re: Looking for simple/beginner's design guide
for experimental construction of GPSDO (1 PPS controlling 10 MHz)
I acquired the board and built the implementation from the attached
QEX article. It seems to work. I know it has been discussed on this
board quite a bit and has it's faults.
From: wb0gaz--- via time-nuts time-nuts@lists.febo.com
I'm curious to learn-by-doing about GPSDO architecture/implementation,
starting with a GPS receiver module (on hand already) that has 1 PPS
output signal.
I suspect this can be a complex undertaking and ill-advised ("why
would you build when you can buy?"), but this is about
experimentation for fun.
My starting point idea is this - a 10 MHz OCXO (which can be
frequency-trimmed by a small DC voltage) feeding a 24-bit counter
whose count is read by a microcontroller, then reset and re-armed
ready for the next the 1PPS signal from GPS module.
Looking for a pointer to a write-up (PDF or web page preferred,
I've never enjoyed trying to learn from youtube videos) that would
help guide my experimentation.
The outcome would certainly not need to be high-specification;
just something that implements 1 PPS leading to control over a
10 MHz oscillator using a microcontroller would suffice.
Thanks for any reading resource pointers,
73 Dave WB0GAZ wb0gaz@yahoo.com
time-nuts mailing list -- time-nuts@lists.febo.com
To unsubscribe send an email to time-nuts-leave@lists.febo.com
Hi
There are a ton of GPS modules out there that do not do the things you need to do
for an accurate GPSDO. Unless you are using a timing oriented product, what you get
on the output may be very different than what you expect to get.
Bob
On May 8, 2025, at 2:58 PM, Pavel Kořenský via time-nuts time-nuts@lists.febo.com wrote:
Hello,
in practice, you do not need to thing about those things, because all necessary corrections are done by the GPS receiver firmware.
Normally, you use the GPS receiver for GPSDO in "position fix" mode, where longitude, lattitude and high above sea level are known. Modern GPS receiver can switch to this mode automatically, you just need to leave the GPS antenna at the same place for several hours or days. Of course, you can switch to this mode manually (mostly via RS-232/USB port on the HPS receiver) if your position is known - for example from Google Maps.
In this mode, the 1PPS is always precise and accurate at the sea level or at the top of Mount Everest. The precision and accuracy are of course dependent on the precision of GPS receiver internal oscillator, satellite status, ionosphere etc.
Best regards
PavelK
Dne 08.05.2025 v 7:19 skipp isaham via time-nuts napsal(a):
Hello to the Group,
From a never having done a home-built GPSDO before... I would
think a home built GPSDO would be quite the involved software
undertaking. Forgive me if this has already been addressed and
I've obviously missed the previously posted info.
Just thinking out loud, unless the micro-controller considered and
uses all the important GPS information... would not your at sea
level 10 MHz zero start/crossing value be some value different from
my 6,000 Feet (1.8288 Kilometers) value. Precision is possible, but
accuracy would not... without including the differences in elevation
and probably other over looked (by me in this reply) items.
regards,
Skipp
Today's Topics: 1. Re: Looking for simple/beginner's design guide for
experimental > construction of GPSDO (1 PPS controlling 10 MHz).
From: chris elfpen.com chris@elfpen.com
Subject: [time-nuts] Re: Looking for simple/beginner's design guide
for experimental construction of GPSDO (1 PPS controlling 10 MHz)
I acquired the board and built the implementation from the attached
QEX article. It seems to work. I know it has been discussed on this
board quite a bit and has it's faults.
https://ve2zaz.net/GPS_Std/Downloads/VE2ZAZ_GPS_Derived_Std_QEX_09_10_2006.pdf
From: wb0gaz--- via time-nuts time-nuts@lists.febo.com
I'm curious to learn-by-doing about GPSDO architecture/implementation,
starting with a GPS receiver module (on hand already) that has 1 PPS
output signal.
I suspect this can be a complex undertaking and ill-advised ("why
would you build when you can buy?"), but this is about
experimentation for fun.
My starting point idea is this - a 10 MHz OCXO (which can be
frequency-trimmed by a small DC voltage) feeding a 24-bit counter
whose count is read by a microcontroller, then reset and re-armed
ready for the next the 1PPS signal from GPS module.
Looking for a pointer to a write-up (PDF or web page preferred,
I've never enjoyed trying to learn from youtube videos) that would
help guide my experimentation.
The outcome would certainly not need to be high-specification;
just something that implements 1 PPS leading to control over a
10 MHz oscillator using a microcontroller would suffice.
Thanks for any reading resource pointers,
73 Dave WB0GAZ wb0gaz@yahoo.com
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
After reading again https://febo.com/pipermail/time-nuts_lists.febo.com/2023-May/107744.html
I was wondering if anyone had some long term (> 1 month) record of time transfer over eLoRAN,
especially how day/night ionosphere impacts the 100 kHz delay, and how the eLORAN station
oscillators are disciplined.
Thanks, Jean-Michel
Hi
There’s a lot of data on Loran propagation from “back in the day”. Start digging in around 1960 or so and you should turn up a number of papers. Propagation for eLoran is no different than the good old Loran-C signals.
Simple answer: Back when we used it for timing, you picked a time of day. You looked at things “at that time” each day. The same “pick a time of day" was true for WWVB. The main advantage for Loran was more stations ment you might be able to find a shorter path from you to the station you used. Something around midnight or noon generally were popular choices for “that time”.
Bob
On Jul 31, 2025, at 10:28 AM, jeanmichel.friedt--- via time-nuts time-nuts@lists.febo.com wrote:
After reading again https://febo.com/pipermail/time-nuts_lists.febo.com/2023-May/107744.html
I was wondering if anyone had some long term (> 1 month) record of time transfer over eLoRAN,
especially how day/night ionosphere impacts the 100 kHz delay, and how the eLORAN station
oscillators are disciplined.
Thanks, Jean-Michel
time-nuts mailing list -- time-nuts@lists.febo.com
To unsubscribe send an email to time-nuts-leave@lists.febo.com
Jean I did see your post and not sure how to answer it.
If you're in the US there are 3 eLORAN sites operational in the western US.
Several of us have been monitoring the transmissions with Austron receivers
and automatically logging the details 24 X 7 Rodger (Fellow time-nut)has
done the most monitoring. We are both on the East Coast.
The eLORAN signals have been on the air for about 18-24 months now.
GRI 59900
Site references are HP5071 cesium with backup units hot switched if needed.
Here is why I haven't responded.
As far as time transfer goes there should be something in the eLORAN data
channel. But no one that I know is decoding that.
They are not doing the old loran C TOC. So I don't see a way to answer your
question.
With respect to the frequency and propagation, that is logged and our local
references are things like GPS, Rb, and CS references.
Various plots do show the propagation changes. An alternate consideration
is the vlf 60 KHz wwvb signal. It behaves the same pretty much.
Assuming you are in North America.
Hope this helps.
Paul
WB8TSL
On Fri, Aug 1, 2025 at 4:17 AM Bob Camp via time-nuts <
time-nuts@lists.febo.com> wrote:
Hi
There’s a lot of data on Loran propagation from “back in the day”. Start
digging in around 1960 or so and you should turn up a number of papers.
Propagation for eLoran is no different than the good old Loran-C signals.
Simple answer: Back when we used it for timing, you picked a time of day.
You looked at things “at that time” each day. The same “pick a time of day"
was true for WWVB. The main advantage for Loran was more stations ment you
might be able to find a shorter path from you to the station you used.
Something around midnight or noon generally were popular choices for “that
time”.
Bob
On Jul 31, 2025, at 10:28 AM, jeanmichel.friedt--- via time-nuts <
time-nuts@lists.febo.com> wrote:
After reading again
I was wondering if anyone had some long term (> 1 month) record of time
transfer over eLoRAN,
especially how day/night ionosphere impacts the 100 kHz delay, and how
the eLORAN station
oscillators are disciplined.
Thanks, Jean-Michel
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
Regarding the current eLoran rollout for timing, it seems like UrsaNav has the only hardware that interacts with the timing signal.
They had one of their boxes on display next to a BPS (Broadcast Positioning System) display at NAB earlier this year, getting time over the eLoran signal and comparing it to BPS (which itself was being received by an area TV station's ATSC3 transmission, which I believe had time set via GPS at the transmitter).
I wrote up a bit about it (with a link to UrsaNav's info) in a recent blog post: https://www.jeffgeerling.com/blog/2025/bps-gps-alternative-nobodys-heard
-Jeff Geerling
On Aug 2, 2025, at 10:48 PM, paul swed via time-nuts time-nuts@lists.febo.com wrote:
Jean I did see your post and not sure how to answer it.
If you're in the US there are 3 eLORAN sites operational in the western US.
Several of us have been monitoring the transmissions with Austron receivers
and automatically logging the details 24 X 7 Rodger (Fellow time-nut)has
done the most monitoring. We are both on the East Coast.
The eLORAN signals have been on the air for about 18-24 months now.
GRI 59900
Site references are HP5071 cesium with backup units hot switched if needed.
Here is why I haven't responded.
As far as time transfer goes there should be something in the eLORAN data
channel. But no one that I know is decoding that.
They are not doing the old loran C TOC. So I don't see a way to answer your
question.
With respect to the frequency and propagation, that is logged and our local
references are things like GPS, Rb, and CS references.
Various plots do show the propagation changes. An alternate consideration
is the vlf 60 KHz wwvb signal. It behaves the same pretty much.
Assuming you are in North America.
Hope this helps.
Paul
WB8TSL
On Fri, Aug 1, 2025 at 4:17 AM Bob Camp via time-nuts <
time-nuts@lists.febo.com> wrote:
Hi
There’s a lot of data on Loran propagation from “back in the day”. Start
digging in around 1960 or so and you should turn up a number of papers.
Propagation for eLoran is no different than the good old Loran-C signals.
Simple answer: Back when we used it for timing, you picked a time of day.
You looked at things “at that time” each day. The same “pick a time of day"
was true for WWVB. The main advantage for Loran was more stations ment you
might be able to find a shorter path from you to the station you used.
Something around midnight or noon generally were popular choices for “that
time”.
Bob
On Jul 31, 2025, at 10:28 AM, jeanmichel.friedt--- via time-nuts <
time-nuts@lists.febo.com> wrote:
After reading again
I was wondering if anyone had some long term (> 1 month) record of time
transfer over eLoRAN,
especially how day/night ionosphere impacts the 100 kHz delay, and how
the eLORAN station
oscillators are disciplined.
Thanks, Jean-Michel
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
time-nuts mailing list -- time-nuts@lists.febo.com
To unsubscribe send an email to time-nuts-leave@lists.febo.com
paul swed via time-nuts writes:
As far as time transfer goes there should be something in the eLORAN data
channel. But no one that I know is decoding that.
The original eLORAN on the NELS chain used the modulation described in
ITU recommendation M.589:
https://www.itu.int/rec/R-REC-M.589/_page.print
The fundamental principle is to shift two pulses one microsecond in
opposite directions, so that the average remains constant.
The snapshot attached show the I+Q plot of that modulation, without
the eLORAN modulation, the dots would form a straight line.
I never managed to decode the data stream usefully, and years
later I learned that they transmitted "an evolution" of M.589,
without telling anybody, because they assumed they had the
only handful of receivers in the world.
I dont know if the current experiment in US/UK/FR uses that modulation.
--
Poul-Henning Kamp | UNIX since Zilog Zeus 3.20
phk@FreeBSD.ORG | TCP/IP since RFC 956
FreeBSD committer | BSD since 4.3-tahoe
Never attribute to malice what can adequately be explained by incompetence.
On Sun, Aug 03, 2025 at 12:02:03PM +0000, Poul-Henning Kamp via time-nuts wrote:
The original eLORAN on the NELS chain used the modulation described in
ITU recommendation M.589:
https://www.itu.int/rec/R-REC-M.589/_page.print
I never managed to decode the data stream usefully, and years
later I learned that they transmitted "an evolution" of M.589,
without telling anybody, because they assumed they had the
only handful of receivers in the world.
I don't think that "evolution" can have made much of a difference.
I wrote a decoder based on ITU M.589-3 back in 2019, and it has been happily
decoding Anthorn's eLoran (Eurofix) transmissions ever since:
1754601250.572 CRC_ok 16277523140529197b5d UTC: 2025:218_21:14:10.62990
1754601252.591 CRC_ok 0c3c5f2e7f1c29656421 message, type 12 caa4e7f5d7de0c
1754601254.611 CRC_ok 113501407d7f40315526 DGPS correction, satellite 0
[...]
1754601274.804 CRC_ok 54440e45094b027f472a Station: ref=549 health=6 eLoran Y secondary GPS ref. Lon. -3.28764
[...]
1754601335.383 CRC_ok 54440e0b6f592e10012a Station: ref=549 health=6 eLoran Y secondary GPS ref. Lat. 54.9114
(But as previously noted, the US stations may well use a different format.)
Regards,
Pieter-Tjerk, PA3FWM