Hi

First question would be “why?” ( or more properly, what are you trying
to accomplish? / what are the project goals? )

You can buy any one of a number of GPSDO’s on eBay for very little money.
They come in a range of “flavors” from a range of suppliers. You can
get OCXO based units. You can get ones based on Rb’s.

Setting up the instrumentation to evaluate a GPSDO design is a bit
complicated. Things like atomic clocks and phase noise / short term
stability analyzers tend to creep into the lab as a result. That’s not a
bad thing. One does need something to fill up all that empty bench
space :) :)

I would suggest that this all puts a project with a goal of “I want to save
money” into the “unlikely” category. If you want higher performance than
what you can easily buy, that will require pretty good instrumentation.

Basic choices these days:

uBlox makes a variety of modules. Some are single band, some are multi
band. Multi band offers some performance advantages. It also costs more.

The Mosaic-T from Septentrio appears to be the top of the heap right now
in terms of mulit band / multi GNSS timing modules.  It has the advantage
of being able to work directly with a 10 MHz input and eliminate some of
the sawtooth correction “stuff”.

Best guess is that going from a from scratch “idea” to something that you
can fully demonstrate works very well is a couple of year sort of project.

None of this is to say don’t do it. The only point is to recognize what you
are signing up for before you dive down the rabbit hole.

Bob

Larry Gadallah lgadallah@gmail.com writes:

Read this paper

https://hamsci.org/sites/default/files/publications/2020_TAPR_DCC/N8UR_GPS_Evaluation_August2020.pdf

and this is a talk by the author (which I haven't seen, but surely is
worth watching).

https://www.youtube.com/watch?v=-M5ayYDh6zY

I'm not saying that will answer all your questions, but I think you'll
be glad you read it.

Greg Troxel writes:

The "multistep" jitter seen on the PPS plots is a deliberately induced high frequency jitter.

The effect is to "rough up" long hanging bridges so they become oscillator wander in GPSDOs.

Pretty analogous to Floyd-Steinberg dithering an image really.

I have tried to implement the same basic idea with both Oncore and
PRS10, by sending commands via the serial port to move the PPS pulse
around but got mediocre results.

--
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.

Hi Bob:

I had to LoL (laugh out loud) when I read your message. I definitely
get what you are saying, but it amused me to no end; When I got my
first ham license, one of my elmers was the regional distributor for
Drake, so I announced my intention to take my hard earned cash and
purchase a shiny new Drake receiver.

The local elders croaked out: "You don't want to do that. You won't
learn anything. You should build your own radio from scratch and gain
some valuable experience." So, here I am some 40 years later, getting
the exact opposite suggestion. I guess it's true that amateur radio in
many respects has become merely a tiny slice of the consumer
electronics market.

But back to the question at hand: From a purely pragmatic point of
view, I've been collecting parts for this project for a decade, and
I'm just thinking that I should either do it, or dispose of the stuff
and forget about it. You are probably right about needing some
sophisticated equipment, and we'll see where that leads.

Seems to me that this is another reality of DIY ham radio these days:
75 years ago, you just needed a (wooden) bread board, some brass
screws, and a tube. Now you need spectrum analyzers, SMT microscopes,
fantastic eyesight and hand-eye coordination, it just isn't as easy
anymore.

73,

On Tue, 15 Feb 2022 at 10:25, Bob kb8tq kb8tq@n1k.org wrote:

--
Larry Gadallah, lgadallah AT gmail DOT com
PGP Sig: 4E5A 6A79 0A8D 2C0D 8D20  609D 3681 AA38 09C0 8303

On Fri, 18 Feb 2022 12:19:23 -0800
Larry Gadallah lgadallah@gmail.com wrote:

Well, if you want to go the DIY way and build everything yourself,
then have a look at [1], where I collected most of the available
GPSDO designs (There are a few missing, but nothing substantial).
I would personally go the way that Nick Sayer took. It is closest
to how I would build a GPSDO. Maybe use a TIC7200 as interpolator
like Tobias Pluess did a couple of months/years ago.

There are a lot of discussion on the time-nut mailinglist on how
to do this or that. I would especially look for the discussions
with Tobias Pluess around 2019/2020. There were some good ideas
passed around in those. But there are plent of others as well.
Just look at the archives and search for GPSDO and you will find plenty.

		Attila Kinali

[1] https://attila.kinali.ch/blog/2016/02/07/gps-disciplined-oscillator

--
The driving force behind research is the question: "Why?"
There are things we don't understand and things we always
wonder about. And that's why we do research.
-- Kobayashi Makoto

HI

If you decide to build your ham receiver from scratch, you then head out to
buy a surplus signal generator for < $100 or so. You buy a couple other pieces
of gear, but none of them are as expensive. Back in the day, your TR4 cost way
more than the “bench gear” to do this or that.

By far, the most likely thing to buy as part of your GPSDO design project is
an atomic clock. Not an Rb mind you , but a properly working Cs or something
more exotic. ( liike a maser ). The bill for them is way past (>10X)  the price of the
GPSDO.

Learning wise, you could go to any of a number of Ham Radio / QST / CQ articles
and pick up on the details of how to build you radio. You could even buy any of
a number of. books that would lead you through the process.

On a GPSDO, there are no books. There are no magazine articles. The folks who
design GPSDO’s don’t talk about what’s inside. It’s not so much the individuals,
it’s how the companies operate. IP matters and it matters a lot. Practical stuff gets
buried as a result. You are off on a “invent it from scratch” expedition (more or less).

Truth in lending: I spend decades doing this and managing folks doing this ….

Bob

But there is a magazine article....
https://www.qsl.net/n9zia/wireless/QST_GPS.pdf
By Brooks Shera, W5OJM.  QST July, 1998.

Following this article I built my unit in 1998 using his PCB and a Motorola UT+.
It has been running almost continuously since then - almost 25 years.
It has been thru several HP 10811 and 10844 oscillators during that time.

Even though Brooks is SK his GPSDO is still running.

Regards.
Bill, NL7F

On Fri, 18 Feb 2022 19:48:21 -0500, Bob kb8tq wrote:

Bill Beam
NL7F

Hi

Building a kit is very different than a from scratch design. A
Heathkit (to extend the years ago ham radio analogy) radio
was an option when you shopped for your Drake. Building it
would give you some experience.

The Heathkit design was done for you and all the complicated
higher level stuff was worked out / tested out / reworked / retested
by somebody else. As long as you put it together correctly, it
worked and was a good radio.

These days there aren’t a lot of folks selling kit GPSDO’s. Even
the kit radio business isn’t what it once was. “I’m going to do
a GPSDO” these days generally means a scratch design.

Bob

After poring over the several valuable references people have posted
here, it seems that even the common commercial GPSDOs have
implementation problems, so it seems like there's no (financially
viable) off the shelf solution.

It would indeed be convenient to have a kit option, but I'd image the
potential marked for a GPSDO kit is vanishingly small.

FWIW, I started with a VE2ZAZ board and an Isotemp OCXO and a Motorola
receiver of questionable provenance. Since then I've gathered an HP
10811D and I have an HP E1938A coming (I think), so I have only so far
gathered some essential components that I hope are competent.

On Sat, 19 Feb 2022 at 07:13, Bob kb8tq kb8tq@n1k.org wrote:

--
Larry Gadallah, lgadallah AT gmail DOT com
PGP Sig: 4E5A 6A79 0A8D 2C0D 8D20  609D 3681 AA38 09C0 8303

Hi

The “typical” commercial GPSDO ( regardless of who made it and for whom)
is the result of years of work and some very exotic verification gear. If indeed
the goal is to do “much better than” what is commercially available, just the
cost of the test bench will be pretty amazing ( like > $100K .. at “eBay” prices
…. yikes …. ).

Bob

You might consider the design by Lars Walenius posted and discussed at
length here
https://www.eevblog.com/forum/projects/lars-diy-gpsdo-with-arduino-and-1ns-resolution-tic/

I have built several GPSDOs based on this design, most recently
incorporating a TDC7200 TDC chip for the time to digital conversion. The
basic hardware design is very simple, but it provides good performance and
a great learning platform.

On Sat, Feb 19, 2022 at 7:10 PM Larry Gadallah lgadallah@gmail.com wrote:

--

--Jim Harman

On Fri, 18 Feb 2022 19:48:21 -0500
Bob kb8tq kb8tq@n1k.org wrote:

I would not be so negative about doing your own GPSDO.
While it is not something the unintiated can just simply conjure
out of thin air, it isn't difficult either...

Depends on what you are looking for. If you trust your GPS receiver to
do the right thing, than you only have to measure out to a day or so.
And for that, a decent Rb is good enough. Maybe get two or three for
a three/four cornered hat.

Sure, if I'd be thinking about selling a GPSDO as a commercial product,
then I'd have to measure it against a H-maser /and/ a Cs beam standard.
But for a hobbyist? A stable Rb vapor cell standard is good enough.

While the companies who build GPSDO mostly keep quiet about what they are
doing, the scientific community doesn't. There are plenty of publications
that deal with the inner working of GPSDO. But you will have to piece it
together on your own, as there is no step-by-step guide available.

If you start with a commercial receiver module, like most hobbyist would,
then a GPSDO becomes a control theoretical problem. I.e., how to design
the control loop to get most out of the stability of the local oscillator,
while using the noisy signal of GPS receiver as a long term reference.
Early GPSDO and what most of the hobby GPSDO that I mentioned do, is a
simple PI or PID loop. If you are a little bit more fancy, you can do a
PII²D loop that compensates for the drift of the local oscillator.
A bit more fancy would be using some adaptive control system, like
the often misunderstood Kalman filter and compensate for the environmental
effects on the local oscillator.

So far, none of this that make a good GPSDO is any secret. It is even well
documented and there are plenty of textbook that explain how to do it.
But, it's not the stuff one ever sees unless one specializes into control
systems (or has a keen interest in them). Another difficulty is that one
needs to model the behaviour of an OCXO. Especially one needs to do
system identification on it and find the proper model to then build the
adaptive control loop on it. Again, the data needed for this is easily
available, but it is not something one casually stumbles upon unless
one specifically goes looking. And it's this combination of two areas
of specialized knowledge that make it difficult for the casual engineer
in a company to build a GPSDO that performs well.

But then, this is time-nuts. Not only do we have plenty of people who
know these things and can point someone interested to the right literature,
the average time-nuts are also the people who spend an exorbitant time in
reading up on all the intricate details of this and that (it's a hobby after
all!) and will be ableo to gather all the knowledge one needs to do it correctly.

So, my advice to anyone who wants to build a GPSDO is the following:

There are 3 key areas that are difficult in a GPSDO design:

1. The receiver and the measurement of the PPS pulse
2. The oscillator and controlling its freequency
3. The control loop binding the above together

For the receiver, I recommend using an available GNSS receiver module.
Don't go the route of running your own (unless you have lots of time).
Measuring the PPS pulse relative to the local oscillator is a bit
tricky, but not too difficult. You only need better than 0.5ns resolution,
which can be done easily in various ways. The most instructive would probably
be to build your own time-to-amplitude converter. Bruce Griffiths published
a nice design years ago that works well. The PICTIC-2 has another one that
gives decent results. And then there is Nick Sayer's super simple one.
(Challenge for those who are so inclined: Come up with a way to do what
Bruce's design does, but only using a single, low voltage power supply.
Yes, it is possible and quite easy once you see it. Hint: Use modern opamps
instead of discrete transistors)

The oscillator is, for our needs, usually an OCXO. Most people use it's EFC
to control the frequency. Which in turn requires something around 20-24bit
of DAC resolution. While not difficult, it's not easy either. The easiest
way to achieve this is to use a 16-20bit DAC and use a 2nd-4th order sigma-
delta modulator to get the remaining bits. I would not go higher than 4th
order unless you are confident in your skills in designing the modulator
correctly. I also recommend using either a DAC with a load data word pin
to have the correct timing or use a CMOS switch to achieve the same.
If you don't use a CMOS switch, make sure you use a DAC with low glitch power.
Alternatively, it's possible to fix the output of the OCXO (bind the EFC
to GND) and use DDS to control the frequency going into the system. There
are multiple ways to do that, starting fromt he obvious way to use the DDS
to generate the frequency, over the FE-405 like PLL based offset generator,
to a full fledged multi-stage direct offset generator like Rick Karlquist did.
Advantage of the latter approach is that you can use any frequency source,
even those that cannot be tuned, like an Rb standard.

For the control loop, I would start with a simple PI loop, extend it to a
PII², then start adding a model that predicts the oscillators drift and
environmental parameters. Start with an as simple model as you can think
of (e.g. linear temperature dependence) and refine from there. Hint: do
not add too many model paramters. The more you have, the harder they get
to estimate. And with that, the chances that the control loop becomes
unstable increase. Also verify after each change that the control loop
is still stable under varying conditions.

And that's about what you need to know about how to design a GPSDO.
Sounds easy, doesn't it? ;-)

		Attila Kinali

--
Science is made up of so many things that appear obvious
after they are explained. -- Pardot Kynes

HI

Here’s the basic problem:

If you don’t already know that everything you are doing is perfect, you need
something better than what you GPSDO does to compare to. If you are happy
with the performance of an Rb then … why not just use the Rb? The grubby issues
are all out past a number of hours and stop at however long you decide to look.

Why “how ever long?”. There are a ton of corner cases. They all pop up eventually.
For a design that you can trust ( one up or whatever ), you need to handle whatever
comes along. Yes, this assumes that a design that works “most of the time” is not
adequate for the task.

Since this is a software rather than a hardware project, debugging does matter. If
you write bug free code every time you do anything, good for you. For the other
99.999% of the group, there will be “stuff” to track down. The problems are not all
from some weird GPS issue. There will indeed be self inflicted wounds to be dealt
with. Some of them are very hard to spot ….

No that is not to suggest in any way that you can ignore the “shorter term” stuff.
You most certainly can not. It’s simply to say that the things that happen quicker
are more easily troubleshot. You still need to spot them, the debug is easier.

Doing the comparison requires that you have something good enough to measure
what’s going on. There are a ton of post from folks who really don’t understand
just how difficult this actually is. Without something like a TimePod, this is going
to be tough.

Can you build this or that from scratch? Sure you can. Being sure that it does indeed
work correctly .. not so easy.

Bob

Larry, Take a look at Brooks Shera's article in QST. you might know of it.

In my opinion, this has been the 'best ever' description of rolling your
own.

The July, 1998 issueBrooks Shera, W5OJM, titled, "A GPS-Based Frequency
Standard. https://www.qsl.net/n9zia/wireless/QST_GPS.pdf"

https://www.qsl.net/n9zia/wireless/QST_GPS.pdf

I build several of these and they made "good" GPSDOs

much has been written about this, and emulating it for other hardware.

https://k6jca.blogspot.com/2018/12/simulating-brooks-shera-w5ojm-gpsdo.html

https://github.com/k6jca/Brooks-Shera-GPSDO

u can still buy the original PCB PCB

http://www.a-aengineering.com/gps.htm

On 20/02/2022 10:59 am, Larry Gadallah wrote:

--
Glen English
RF Communications and Electronics Engineer

CORTEX RF

Pacific Media Technologies Pty Ltd trading as Cortex RF

ABN 40 075 532 008

PO Box 5231 Lyneham ACT 2602, Australia.
au mobile : +61 (0)418 975077

Very interesting to read about this good old design!

I have still a board for the Brooks Shera GPSDO and I would like to
complete it, but I cannot find the described 24 MHz/ 6 MHz Oscillator.
So I have some questions to it.

Is it neccessary that these frequencies are exact and running synchronous?
Are there special tcxo available which would fit?
Could I run this board as well with other (e.g. higher) frequencies and
of independent oscillators?
Are there improvements in meantime?

regards,

Arnold, DK2WT

Am 20.02.2022 um 03:11 schrieb glen english LIST:

Hi Arnold

The 24M and 6M must be synchronous to avoid adding an edge uncertainty .

But that isnt too hard to manage, just start with a  5V  24 MHz
oscillator (anything, or say a 74AC14 and a 24 MHz crystal)  and a fast
, say 74AC74 dual D type flip flop for 6 MHz (AC to keep the div 4 prop
delay down and more consistent with temp).

The speed this will run at is very much dependent on the logic family .
this board is built with 74HC and an accompanied old slow PIC.  You'd
need to change the logic and the processor software to use other higher
freqs.

There was nothing really special about this design except it was a
simple technique done well and used some novel ideas to permit
relatively low performance computing hardwaree to measure with a high
precision.

regards,

Glen.

On 21/02/2022 3:50 am, Arnold Tibus wrote:

--
Glen English
RF Communications and Electronics Engineer

CORTEX RF

Pacific Media Technologies Pty Ltd trading as Cortex RF

ABN 40 075 532 008

PO Box 5231 Lyneham ACT 2602, Australia.
au mobile : +61 (0)418 975077

For those of you interested in rolling your own a-la the Brooks Shera
style, which is essentially adigital readout phase detector , these days
, with almost a single chip....

with modern microcontrollers you can do better than the Shera hardware,
I describe how I do it :

The modern micros like 168 MHz $10 STM32F4 have 16 and 32 bit counters
running at full core speed of 168 MHz or core div2 at 84 MHz depending
on what you buy.

The core clock pin  receives it input clock from the 10 MHz (or whatever
you like) oscillator you wish to disclipine. Internally this 10 MHz is
routed to the clock of a  counter. The microcontroller has a PLL that
multiplies this up to 160 MHz ish.

If you are willing to accept a little extra noise from the internal PLL
multiplier , the 160 MHz core clock also sent to the counter. (much
higher PD resolution)

The 1PPS is connected to the external input counter gate ON OFF- all
hardware gates internally - so the 1PPS will gate on and gate off the
timer(s). There are several gate modes to get either emulated
phase/frequency detectors. Now you have the Shera hardware emulated.

The processor has a fast 1MHz 12 bit DAC and internal bandgap DAC
reference. nice !.... The 12 Bit DAC is oversampled up to 18 bits, There
will be a limit due to the ADEV of references etc etc but its all pretty
good.

I use the LVDS DS90LV....  family of differential receivers, non
internally terminated inputs,  for my squaring of input sine waves into
nice sharp logic inputs for the processor. Some processors have schmitt
and glitch removal  inputs already.

The C code is nice an easy to write and debug , fast processor, and
plenty of UARTs to drive serial outputs for status strings .

-glen

On 20/02/2022 12:36 pm, Bob kb8tq wrote: