Brooks Shera's GPS standard or HP Z3801A??
Tom Van Baak wrote:
This issue comes up a lot. If weren't for the glut of
surplus Z3801A I'm sure the situation would be very
different.
Agreed.
Brooks Shera did a fine job with his project
and reached a wide audience with its QST publication.
I was always under the impression it was published in QEX, the ARRL's
more technical journal, but I see you are right, it was QST. One ham
once said to me a typical QST article was "How to build a 12V 1A power
supply using a 7812 voltage regulator". Clearly this was a bit different.
I would agree with you the article is excellent.
I don't get QST now, as I found it too childish, but that article, like
many that appeared years ago, was very good.
It's a shame ham radio has been spoilt with 99% of people just spending
money and not understanding anything.
If your goal is a turn-key 10 MHz frequency standard
with short-term stability under 1e-11 then a Z3801A
is a good solution.
If your goal is to build something yourself, to use parts
you have already, to learn as you go, and achieve
similar performance then use the Brooks Shera design.
Working out what my goal is not an easy task.
It started with the desire to stabilise an 18th century pendulum clock
with a crystal!! I realised I needed a decent crystal, and some way to
measure its frequency, and it sort of escalated from there.
The route I take might depend on the outcome of this eBay transaction of
mine for the 18011A.
If you're a time-nut you'll end up with both anyway so
that you can compare them; the only question is which
to get first. ;-)
/tvb
Perhaps.
BTW, there is a lady at work who is quite into clocks. I sent her an
email today with your wristwatch, but she replied she did not think it
would be too popular.
Dr. David Kirkby.
----- Original Message -----
From: "Dr. David Kirkby" drkirkby@medphys.ucl.ac.uk
To: "Discussion of precise time and frequency measurement"
time-nuts@febo.com
Sent: Monday, April 25, 2005 02:35
Subject: [time-nuts] Brooks Shera's GPS standard or HP Z3801A??
I was going to build the GPS reference by Brooks Shera
http://www.rt66.com/~shera/index_fs.htm
and have bought the PCB and started populating it (not much yet).
But now I am wondering if buying a used Z3801A might not be a better idea.
My thoughts are:
-
Cost - the Z3801A will probably cost less than buying all the parts
for Brook's Standard. This is especially so, with me in the UK, where
shipping charges on lots of small items from the USA soon add up. -
Medium term stability - the double oven in the Z3801A might be more
stable than an HP 18011A. -
Flexability - Brook's standard can be customised to whatever you want.
-
Long term servicability - if one builds the standard oneself, one is
far more likely to be able to debug it if things go wrong. One can buy a
few of the chips, for spares, just in case one dies and becomes hard to
get. The chips on the board are not that expensive. -
Jitter on GPS.
-
If one needs to buy a manual on the Z3801A, that will add to its cost.
Any other thoughts ????
--
Dr. David Kirkby PhD CEng MIEE,
Senior Research Fellow,
Department of Medical Physics,
Mallet Place Engineering Building,
Gower St,
University College London,
London WC1E 6BT.
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
--
Dr. David Kirkby,
G8WRB
Please check out http://www.g8wrb.org/
of if you live in Essex http://www.southminster-branch-line.org.uk/
Hi David:
The following web page is highley recomended if you're into precision
pendulum clocks. It's titled Electric CLocks to differentiate from
spring wound clocks. By all means get the latest Macromedia Shockwave
player and spend some time viewing the animations with the sound turned
on. In most cases the pendulum must swing for a number of seconds
before the electric circuit is actuated and gives it a "kick".
http://mypage.bluewin.ch/electric-clocks/
http://www.macromedia.com/shockwave/download/
Have Fun,
Brooke Clarke
--
w/Java http://www.PRC68.com
w/o Java http://www.pacificsites.com/~brooke/PRC68COM.shtml
http://www.precisionclock.com
David Kirkby wrote:
. . .
Working out what my goal is not an easy task.
It started with the desire to stabilise an 18th century pendulum clock
with a crystal!! I realised I needed a decent crystal, and some way to
measure its frequency, and it sort of escalated from there.
The route I take might depend on the outcome of this eBay transaction
of mine for the 18011A.
Brooke Clarke wrote:
Hi David:
The following web page is highley recomended if you're into precision
pendulum clocks. It's titled Electric CLocks to differentiate from
spring wound clocks. By all means get the latest Macromedia Shockwave
player and spend some time viewing the animations with the sound turned
on. In most cases the pendulum must swing for a number of seconds
before the electric circuit is actuated and gives it a "kick".
http://mypage.bluewin.ch/electric-clocks/
http://www.macromedia.com/shockwave/download/
Have Fun,
Brooke Clarke
Thanks. The only problem is I'll have to leave my trusty old Sun and go
to a PeeCee running Windoze!!!!
--
Dr. David Kirkby,
G8WRB
Please check out http://www.g8wrb.org/
of if you live in Essex http://www.southminster-branch-line.org.uk/
David Kirkby wrote:
Brooke Clarke wrote:
Hi David:
The following web page is highley recomended if you're into precision
pendulum clocks. It's titled Electric CLocks to differentiate from
spring wound clocks. By all means get the latest Macromedia Shockwave
player and spend some time viewing the animations with the sound
turned on. In most cases the pendulum must swing for a number of
seconds before the electric circuit is actuated and gives it a "kick".
http://mypage.bluewin.ch/electric-clocks/
http://www.macromedia.com/shockwave/download/
Have Fun,
Brooke Clarke
Thanks. The only problem is I'll have to leave my trusty old Sun and go
to a PeeCee running Windoze!!!!
I correct myself - it seems there is a Shockwave plugin for Solaris!
--
Dr. David Kirkby,
G8WRB
Please check out http://www.g8wrb.org/
of if you live in Essex http://www.southminster-branch-line.org.uk/
Poul-Henning Kamp wrote:
In message 000801c549ca$7c106d80$b20ff204@computer, "Tom Van Baak" writes:
Don't forget that a PRS10 Rb from eBay and a Oncore M12+T from
synergy-gps.com would be both cheaper and more stable.
Why would it be cheaper than building Brook Shera's unit? You still need
all the same electronics, but a more expensive oscillator (I assume the
PRS10 is going to cost more than an HP 18011A.)
I can understand that it is probably likely the peformance will be
better than that of the HP Z3801A, as I guess you can use a long time
constant to keep the Rubidium on frequency, which would not be possible
if using a crystal.
Good point. Has someone posted a Z3801A vs.
PRS10/M12+ comparison? I have a lot of data
on the PRS10 but haven't put an M12+ to it yet.
How does the short term phase noise of the PRS10's compare with the HP
10811A? I guess in the long term, they will both be the same if you link
them to a GPS source, but short term you have said will be dominated by
the crystal oscillator.
I guess a current unit is likely to be better than a 20-30 year old one.
I should add that with an UT+ the sawtooth has a tendency to
throw the PRS10s PLL off kilter, even with the 256 second
averager (because the sawtooth does not necessairly average
out to zero in that timeinterval).
Even then, the PRS10+M12+T is by far the best bang for the
buck I can think off.
I am confused how this can be cheaper although I accept its performance
should be excellent.
--
Dr. David Kirkby PhD CEng MIEE,
Senior Research Fellow,
Department of Medical Physics,
Mallet Place Engineering Building,
Gower St,
University College London,
London WC1E 6BT.
In message 426E2C26.5090809@medphys.ucl.ac.uk, "Dr. David Kirkby" writes:
Poul-Henning Kamp wrote:
In message 000801c549ca$7c106d80$b20ff204@computer, "Tom Van Baak" writes:
Don't forget that a PRS10 Rb from eBay and a Oncore M12+T from
synergy-gps.com would be both cheaper and more stable.
Why would it be cheaper than building Brook Shera's unit? You still need
all the same electronics, but a more expensive oscillator (I assume the
PRS10 is going to cost more than an HP 18011A.)
The PRS10 has the pll built in, all you need is a wire and a power-supply.
--
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.
Why would it be cheaper than building Brook Shera's unit? You still need
all the same electronics, but a more expensive oscillator (I assume the
PRS10 is going to cost more than an HP 18011A.)
The data sheet for the PRS10 is at:
http://www.thinksrs.com/products/PRS10.htm
http://www.thinksrs.com/downloads/PDFs/Manuals/PRS10m.pdf
How does the short term phase noise of the PRS10's compare with the HP
10811A? I guess in the long term, they will both be the same if you link
them to a GPS source, but short term you have said will be dominated by
the crystal oscillator.
The ADEV for a PRS10 at 1 to 10 seconds is on the
order of 1e-11 making it 10x worse than the 10811
inside a Z3801A. So the choice depends on what
you want to use the output of your GPSDO for.
I guess a current unit is likely to be better than a 20-30 year old one.
Not necessarily true. Some of the best oscillators in
the world were made 30 to 40 years ago. On the other
hand, you can't beat a PRS10 for features. Check out
the manual for the list of commands.
/tvb
Hi Tom:
Don't forget that SRS has the free PC program Rbmon that talks to the
PRS10, although it's listed with a different Rb product.
73,
Brooke
Tom Van Baak wrote:
Why would it be cheaper than building Brook Shera's unit? You still need
all the same electronics, but a more expensive oscillator (I assume the
PRS10 is going to cost more than an HP 18011A.)
The data sheet for the PRS10 is at:
http://www.thinksrs.com/products/PRS10.htm
http://www.thinksrs.com/downloads/PDFs/Manuals/PRS10m.pdf
How does the short term phase noise of the PRS10's compare with the HP
10811A? I guess in the long term, they will both be the same if you link
them to a GPS source, but short term you have said will be dominated by
the crystal oscillator.
The ADEV for a PRS10 at 1 to 10 seconds is on the
order of 1e-11 making it 10x worse than the 10811
inside a Z3801A. So the choice depends on what
you want to use the output of your GPSDO for.
I guess a current unit is likely to be better than a 20-30 year old one.
Not necessarily true. Some of the best oscillators in
the world were made 30 to 40 years ago. On the other
hand, you can't beat a PRS10 for features. Check out
the manual for the list of commands.
/tvb
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
The SRS is actually looking more and more attractive I must say, mainly
because of its size (the Z3801A is not small), and the long holdover
period if I can't get a GPS signal - which is quite likely.
I'm tempted to take this in stages, using a reasonably large box, giving
room for a number of items. My thoughts are:
- Buy the SRS PRS10 and assume that is right for the purpose of the
inital aim, which was to sync my pendulum clock. Errors in the rubidium
source should be small.
I'll leave the GPS initially - I have a lot of expenses at the minute.
-
Put some sealed lead acid batteries for backup - I suffer a lot of
power failures at home. A friend has offered me a charger for 24V
batteries. -
Sync the pendulum clock to a crystal (either TCXO or perhaps an OCXO)
- the latter being a bit over the top. This is going to be the hard
part, especially as I will need a PIC to get 1 point something Hz, and I
have not used PICs before.
This will go in its own box.
-
Add an HP 10811A + Brooks Shera board to get low phase noise,
syncing not to GPS, but to the PRS10. I have all the bits for the
Brooks Shera board (apart from a few cheap ICs), so the cost in doing
this is small. -
Finally add a GPS, which should give me the low phase noise of the
10811A, with a decent holdover time from the PRS10 if there is no GPS
signal.
I suspect I could fit that lot (apart from 3 which I want sepparte) in a
3U rack.
I'd need to synchonise the power-up of the two ovens as the power supply
I have (24V, 2.4A) would not be capable of starting both ovens at the
same time. The PRS10 take 2.2A on startup, but only 0.6A when running.
Hence the 2.4A power supply should be okay in running both the PRS10 and
the HP 10811A, but not starting them together.
Any thoughts on that sort of idea above?
Any obvious flaws?? Apart from the fact it is getting more and more
expensive, and less and less related to the initial aim of making a
couple of hundred year old pendulum clock more accurate!!!
Dr. David Kirkby
Brooke Clarke wrote:
Hi Tom:
Don't forget that SRS has the free PC program Rbmon that talks to the
PRS10, although it's listed with a different Rb product.
73,
Brooke
Tom Van Baak wrote:
Why would it be cheaper than building Brook Shera's unit? You still
need all the same electronics, but a more expensive oscillator (I
assume the PRS10 is going to cost more than an HP 18011A.)
The data sheet for the PRS10 is at:
http://www.thinksrs.com/products/PRS10.htm
http://www.thinksrs.com/downloads/PDFs/Manuals/PRS10m.pdf
How does the short term phase noise of the PRS10's compare with the
HP 10811A? I guess in the long term, they will both be the same if
you link them to a GPS source, but short term you have said will be
dominated by the crystal oscillator.
The ADEV for a PRS10 at 1 to 10 seconds is on the
order of 1e-11 making it 10x worse than the 10811
inside a Z3801A. So the choice depends on what
you want to use the output of your GPSDO for.
I guess a current unit is likely to be better than a 20-30 year old one.
Not necessarily true. Some of the best oscillators in
the world were made 30 to 40 years ago. On the other
hand, you can't beat a PRS10 for features. Check out
the manual for the list of commands.
/tvb
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
--
Dr. David Kirkby PhD CEng MIEE,
Senior Research Fellow,
Department of Medical Physics,
Mallet Place Engineering Building,
Gower St,
University College London,
London WC1E 6BT.
The pendulum clock was what got me started on this, but I do have other
test equipment, including a microwave counter, which could do with a
decent reference
Brooke Clarke wrote:
Hi David:
How accurate is the pendulum clock now? How accurate do you want it to be?
Have Fun,
Brooke
Dr. David Kirkby wrote:
The SRS is actually looking more and more attractive I must say,
mainly because of its size (the Z3801A is not small), and the long
holdover period if I can't get a GPS signal - which is quite likely.
I'm tempted to take this in stages, using a reasonably large box,
giving room for a number of items. My thoughts are:
- Buy the SRS PRS10 and assume that is right for the purpose of the
inital aim, which was to sync my pendulum clock. Errors in the
rubidium source should be small.
I'll leave the GPS initially - I have a lot of expenses at the minute.
-
Put some sealed lead acid batteries for backup - I suffer a lot of
power failures at home. A friend has offered me a charger for 24V
batteries. -
Sync the pendulum clock to a crystal (either TCXO or perhaps an
OCXO) - the latter being a bit over the top. This is going to be the
hard part, especially as I will need a PIC to get 1 point something
Hz, and I have not used PICs before.
This will go in its own box.
-
Add an HP 10811A + Brooks Shera board to get low phase noise,
syncing not to GPS, but to the PRS10. I have all the bits for the
Brooks Shera board (apart from a few cheap ICs), so the cost in doing
this is small. -
Finally add a GPS, which should give me the low phase noise of the
10811A, with a decent holdover time from the PRS10 if there is no GPS
signal.
I suspect I could fit that lot (apart from 3 which I want sepparte) in
a 3U rack.
I'd need to synchonise the power-up of the two ovens as the power
supply I have (24V, 2.4A) would not be capable of starting both ovens
at the same time. The PRS10 take 2.2A on startup, but only 0.6A when
running. Hence the 2.4A power supply should be okay in running both
the PRS10 and the HP 10811A, but not starting them together.
Any thoughts on that sort of idea above?
Any obvious flaws?? Apart from the fact it is getting more and more
expensive, and less and less related to the initial aim of making a
couple of hundred year old pendulum clock more accurate!!!
Dr. David Kirkby
Brooke Clarke wrote:
Hi Tom:
Don't forget that SRS has the free PC program Rbmon that talks to the
PRS10, although it's listed with a different Rb product.
73,
Brooke
Tom Van Baak wrote:
Why would it be cheaper than building Brook Shera's unit? You still
need all the same electronics, but a more expensive oscillator (I
assume the PRS10 is going to cost more than an HP 18011A.)
The data sheet for the PRS10 is at:
http://www.thinksrs.com/products/PRS10.htm
http://www.thinksrs.com/downloads/PDFs/Manuals/PRS10m.pdf
How does the short term phase noise of the PRS10's compare with the
HP 10811A? I guess in the long term, they will both be the same if
you link them to a GPS source, but short term you have said will be
dominated by the crystal oscillator.
The ADEV for a PRS10 at 1 to 10 seconds is on the
order of 1e-11 making it 10x worse than the 10811
inside a Z3801A. So the choice depends on what
you want to use the output of your GPSDO for.
I guess a current unit is likely to be better than a 20-30 year old
one.
Not necessarily true. Some of the best oscillators in
the world were made 30 to 40 years ago. On the other
hand, you can't beat a PRS10 for features. Check out
the manual for the list of commands.
/tvb
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
--
Dr. David Kirkby PhD CEng MIEE,
Senior Research Fellow,
Department of Medical Physics,
Mallet Place Engineering Building,
Gower St,
University College London,
London WC1E 6BT.