Discussion:
Most accurate small crystal
(too old to reply)
Jim Palfreyman
2010-05-27 04:22:18 UTC
Permalink
Hi All,

I have a RSA SecurID device which I use to log in to my work's VPN. For
those of you not familiar with these, they show you a 6 digit number that
you use (combined with a PIN) to log in. This number changes every minute.
The changing of this number lines up with the servers at the other end and,
as I understand it, they do take into account gradual drift. These devices
have a limited programmed lifetime of three years as well. So the internal
clock on one of these needs to be decent.

I have timed the accuracy of this internal clock and have found it to be
pretty good so far. 17 days ago it was ticking over at 21.8 sec past the
minute and a quick visual inspection today and it was still *very* close to
that. I will confirm it properly tonight.

Now this device travels around with me, like a wristwatch, but is not on my
person. It is usually in my laptop bag. So it is subject to quite varying
temperatures.

So it's been nearly 3 weeks and this device has barely changed 0.1 to 0.2
secs tops. Now that's not bad compared to digital watches.

I presume it has a quartz crystal in it, obviously no oven, but my question
is:

What is the best crystal you can get on the market today that would work in
a watch type device with very little power available? What's the best
accuracy that can be expected?

I know straight out of the factory they can be pretty spot on (as my Casio G
Shock was) but it has aged markedly and now gains a couple of seconds a
week. So what is the best we can expect long term too?

Regards,

Jim Palfreyman
_______________________________________________
time-nuts mailing list -- time-nuts-***@public.gmane.org
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Brooke Clarke
2010-05-27 05:30:36 UTC
Permalink
Hi Jim:

The last time I looked into this it was the DS3231 (now part of Maxim).
See the live demo:
http://www.maxim-ic.com/products/timers/DS3231_demo/

Have Fun,

Brooke Clarke
http://www.PRC68.com
Post by Jim Palfreyman
Hi All,
I have a RSA SecurID device which I use to log in to my work's VPN. For
those of you not familiar with these, they show you a 6 digit number that
you use (combined with a PIN) to log in. This number changes every minute.
The changing of this number lines up with the servers at the other end and,
as I understand it, they do take into account gradual drift. These devices
have a limited programmed lifetime of three years as well. So the internal
clock on one of these needs to be decent.
I have timed the accuracy of this internal clock and have found it to be
pretty good so far. 17 days ago it was ticking over at 21.8 sec past the
minute and a quick visual inspection today and it was still *very* close to
that. I will confirm it properly tonight.
Now this device travels around with me, like a wristwatch, but is not on my
person. It is usually in my laptop bag. So it is subject to quite varying
temperatures.
So it's been nearly 3 weeks and this device has barely changed 0.1 to 0.2
secs tops. Now that's not bad compared to digital watches.
I presume it has a quartz crystal in it, obviously no oven, but my question
What is the best crystal you can get on the market today that would work in
a watch type device with very little power available? What's the best
accuracy that can be expected?
I know straight out of the factory they can be pretty spot on (as my Casio G
Shock was) but it has aged markedly and now gains a couple of seconds a
week. So what is the best we can expect long term too?
Regards,
Jim Palfreyman
_______________________________________________
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
_______________________________________________
time-nuts mailing list -- time-nuts-***@public.gmane.org
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Poul-Henning Kamp
2010-05-27 06:28:08 UTC
Permalink
Post by Jim Palfreyman
I have timed the accuracy of this internal clock and have found it to be
pretty good so far. 17 days ago it was ticking over at 21.8 sec past the
minute and a quick visual inspection today and it was still *very* close to
that. I will confirm it properly tonight.
Being able to model the clock of each device precisely is a very important
part of the security of these devices.

The important part there, is that the clock runs at a constant rate,
but the exact value of the rate is not important, as long as the
bounds are known.
--
Poul-Henning Kamp | UNIX since Zilog Zeus 3.20
phk-***@public.gmane.org | TCP/IP since RFC 956
FreeBSD committer | BSD since 4.3-tahoe
Never attribute to malice what can adequately be explained by incompetence.

_______________________________________________
time-nuts mailing list -- time-nuts-***@public.gmane.org
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
jimlux
2010-05-27 13:24:00 UTC
Permalink
Post by Poul-Henning Kamp
Post by Jim Palfreyman
I have timed the accuracy of this internal clock and have found it to be
pretty good so far. 17 days ago it was ticking over at 21.8 sec past the
minute and a quick visual inspection today and it was still *very* close to
that. I will confirm it properly tonight.
Being able to model the clock of each device precisely is a very important
part of the security of these devices.
The important part there, is that the clock runs at a constant rate,
but the exact value of the rate is not important, as long as the
bounds are known.
I don't know that it needs to be all that precise..

The way it works is that your fob steps through a list of (pseudo)random
numbers, one per minute. The server steps through the same list. If
you "call in" with a particular number, the server doesn't just check
the current step on the list, it checks some (configurable) range of
steps on either side. If it gets a match, it resets the server's
"current step" to the one that matched.

The approach is very similar to the "rolling code" used in garage door
openers and wireless key entry systems for cars.

See patent 4,885,778.

I don't know if they actually try to model the clock rate. I think if
you get too many "misses" you just get another fob.

The thing steps at once per minute. Say you allow up to 1 minutes error,
and that the XO has 10 ppm error (e.g. 1 second/day).. that's a minute
every 2 months.



_______________________________________________
time-nuts mailing list -- time-nuts-***@public.gmane.org
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Bob Camp
2010-05-27 11:43:31 UTC
Permalink
Hi

If you dig into the way they "certify" things like ships chronometers and wrist watches, they focus on averaged errors over a long period of time. That's pretty much what you do when you check your watch. You check it today and you check it a week from now. It's off by a second (or not) and that's what you pay attention to. It's rate could be off quite a bit over some portion of the period. That's fine as long as it averages out.

This sort of thing is *very* algorithm friendly. You can correct for a lot of things after the fact. The results will be a bit variable since it's a "how lucky did you get" sort of thing. They also will tend to degrade over time, as the data you initially loaded in on temperature performance, aging, and G sensitivity departs from reality.

Bob
Post by Jim Palfreyman
Hi All,
I have a RSA SecurID device which I use to log in to my work's VPN. For
those of you not familiar with these, they show you a 6 digit number that
you use (combined with a PIN) to log in. This number changes every minute.
The changing of this number lines up with the servers at the other end and,
as I understand it, they do take into account gradual drift. These devices
have a limited programmed lifetime of three years as well. So the internal
clock on one of these needs to be decent.
I have timed the accuracy of this internal clock and have found it to be
pretty good so far. 17 days ago it was ticking over at 21.8 sec past the
minute and a quick visual inspection today and it was still *very* close to
that. I will confirm it properly tonight.
Now this device travels around with me, like a wristwatch, but is not on my
person. It is usually in my laptop bag. So it is subject to quite varying
temperatures.
So it's been nearly 3 weeks and this device has barely changed 0.1 to 0.2
secs tops. Now that's not bad compared to digital watches.
I presume it has a quartz crystal in it, obviously no oven, but my question
What is the best crystal you can get on the market today that would work in
a watch type device with very little power available? What's the best
accuracy that can be expected?
I know straight out of the factory they can be pretty spot on (as my Casio G
Shock was) but it has aged markedly and now gains a couple of seconds a
week. So what is the best we can expect long term too?
Regards,
Jim Palfreyman
_______________________________________________
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
_______________________________________________
time-nuts mailing list -- time-nuts-***@public.gmane.org
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Steve Rooke
2010-05-27 14:05:11 UTC
Permalink
Bob,
Post by Bob Camp
This sort of thing is *very* algorithm friendly. You can correct for a lot of things after the fact. The results will be a bit variable since it's a "how lucky did you get" sort of thing. They also will tend to degrade over time, as the data you initially loaded in on temperature performance, aging, and G sensitivity departs from reality.
I'm very interested in this area, do you know of any sources of
studies on this topic?

Thanks,
Steve
Post by Bob Camp
Bob
--
Steve Rooke - ZL3TUV & G8KVD
A man with one clock knows what time it is;
A man with two clocks is never quite sure.

_______________________________________________
time-nuts mailing list -- time-nuts-***@public.gmane.org
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Bob Camp
2010-05-27 16:32:43 UTC
Permalink
Hi

There are a raft of papers on each of the sub portions of the fitting
process. Aging, retrace, temperature, and acceleration all have their own
issues and fit approaches.

The whole "how (and why that way) do they test a chronometer?" is something
there's a lot of papers on as well. Some of them date back into the 1600's.

Where do you want to start?

Bob

-----Original Message-----
From: time-nuts-bounces-***@public.gmane.org [mailto:time-nuts-bounces-***@public.gmane.org] On
Behalf Of Steve Rooke
Sent: Thursday, May 27, 2010 10:05 AM
To: Discussion of precise time and frequency measurement
Subject: Re: [time-nuts] Most accurate small crystal

Bob,
Post by Bob Camp
This sort of thing is *very* algorithm friendly. You can correct for a lot
of things after the fact. The results will be a bit variable since it's a
"how lucky did you get" sort of thing. They also will tend to degrade over
time, as the data you initially loaded in on temperature performance, aging,
and G sensitivity departs from reality.

I'm very interested in this area, do you know of any sources of
studies on this topic?

Thanks,
Steve
Post by Bob Camp
Bob
--
Steve Rooke - ZL3TUV & G8KVD
A man with one clock knows what time it is;
A man with two clocks is never quite sure.

_______________________________________________
time-nuts mailing list -- time-nuts-***@public.gmane.org
To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.




_______________________________________________
time-nuts mailing list -- time-nuts-***@public.gmane.org
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Poul-Henning Kamp
2010-05-27 20:53:50 UTC
Permalink
Post by Bob Camp
Where do you want to start?
Isn't that HP/Agilent App-Note with the three wrist-watches the
place to start ? 1289 or something ?

Poul-Henning
--
Poul-Henning Kamp | UNIX since Zilog Zeus 3.20
phk-***@public.gmane.org | TCP/IP since RFC 956
FreeBSD committer | BSD since 4.3-tahoe
Never attribute to malice what can adequately be explained by incompetence.

_______________________________________________
time-nuts mailing list -- time-nuts-***@public.gmane.org
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Bob Camp
2010-05-28 01:30:10 UTC
Permalink
Hi

There are lots of starting points. The smart clock stuff is another starting point.

Bob
Post by Poul-Henning Kamp
Post by Bob Camp
Where do you want to start?
Isn't that HP/Agilent App-Note with the three wrist-watches the
place to start ? 1289 or something ?
Poul-Henning
--
Poul-Henning Kamp | UNIX since Zilog Zeus 3.20
FreeBSD committer | BSD since 4.3-tahoe
Never attribute to malice what can adequately be explained by incompetence.
_______________________________________________
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
_______________________________________________
time-nuts mailing list -- time-nuts-***@public.gmane.org
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Steve Rooke
2010-05-28 14:05:06 UTC
Permalink
Hi Bob,
Post by Bob Camp
There are a raft of papers on each of the sub portions of the fitting
process. Aging, retrace, temperature, and acceleration all have their own
issues and fit approaches.
The whole "how (and why that way) do they test a chronometer?" is something
there's a lot of papers on as well. Some of them date back into the 1600's.
Where do you want to start?
I'll probably start in the middle and figure out if I'm in the right
place before I move from there. What I'm looking for are any docs on
the various ways that xtal oscillators are affected, IE. drift over
short time, long time, temp, pressure, humidity, gravitational effects
of the Moon and the Earth, that sort of thing, things that have been
studied and can be modelled for a xtal. I'm interested in what has
been done by others to try and correct as much as possible, IE. ocxo
et al, and to predict changes in an undisciplined xtal. When I look at
the efc of my gpsdo I can see the effects of drift and temp changes
and I'd like to look at the predictability of that. I'm excluding
other forms of noise and xtal flips from this as they are outside any
form of predictable control.

Thanks,
Steve
--
Steve Rooke - ZL3TUV & G8KVD
A man with one clock knows what time it is;
A man with two clocks is never quite sure.

_______________________________________________
time-nuts mailing list -- time-nuts-***@public.gmane.org
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Bob Camp
2010-05-28 14:29:37 UTC
Permalink
Hi

There are whole series of papers on each of the effects you are asking about. They all have first, second, third .... order impacts. You could also add things like drive level, electric field, and (obviously) load onto your list. There are also papers on the cross correlation between issues. Things like load, drive, and temperature are indeed inter-related.

The gotcha is that the crystal they studied probably isn't the crystal you have. Simply put, a 2" diameter blank in a 3" glass holder is going to have different "major" issues than a 2x3mm ceramic package strip crystal.The only way to actually know the impact on the crystal you have is to measure it. Even crystals from the same batch will have different specific issues.

If you are looking at a bare crystal, the issues will be very different than if you are trying to model a modern OCXO.

Best starting place - buy the CD's of the back issues of the Frequency Control Symposium. Each issue has a couple dozen papers looking at the kinds of issues you are interested in. The symposium has been running for a lot of years, there are lots of papers to dig into.

Bob
Post by Steve Rooke
Hi Bob,
Post by Bob Camp
There are a raft of papers on each of the sub portions of the fitting
process. Aging, retrace, temperature, and acceleration all have their own
issues and fit approaches.
The whole "how (and why that way) do they test a chronometer?" is something
there's a lot of papers on as well. Some of them date back into the 1600's.
Where do you want to start?
I'll probably start in the middle and figure out if I'm in the right
place before I move from there. What I'm looking for are any docs on
the various ways that xtal oscillators are affected, IE. drift over
short time, long time, temp, pressure, humidity, gravitational effects
of the Moon and the Earth, that sort of thing, things that have been
studied and can be modelled for a xtal. I'm interested in what has
been done by others to try and correct as much as possible, IE. ocxo
et al, and to predict changes in an undisciplined xtal. When I look at
the efc of my gpsdo I can see the effects of drift and temp changes
and I'd like to look at the predictability of that. I'm excluding
other forms of noise and xtal flips from this as they are outside any
form of predictable control.
Thanks,
Steve
--
Steve Rooke - ZL3TUV & G8KVD
A man with one clock knows what time it is;
A man with two clocks is never quite sure.
_______________________________________________
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
_______________________________________________
time-nuts mailing list -- time-nuts-***@public.gmane.org
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Steve Rooke
2010-05-29 06:05:56 UTC
Permalink
Hi Bob,
Post by Bob Camp
Hi
Best starting place - buy the CD's of the back issues of the Frequency Control Symposium. Each issue has a couple dozen papers looking at the kinds of issues you are interested in. The symposium has been running for a lot of years, there are lots of papers to dig into.
Thanks for the pointer to the IEEE and although I've not found
reference to the CD's yet, I came across this base link to Frequency
Control, http://www.ieee-uffc.org/frequency_control/, and from there
I've found some useful documents. This link is particularly
interesting, http://www.ieee-uffc.org/frequency_control/teaching.asp.

Thanks,
Steve
--
Steve Rooke - ZL3TUV & G8KVD
A man with one clock knows what time it is;
A man with two clocks is never quite sure.

_______________________________________________
time-nuts mailing list -- time-nuts-***@public.gmane.org
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Magnus Danielson
2010-05-29 11:17:27 UTC
Permalink
Post by Steve Rooke
Hi Bob,
Post by Bob Camp
Hi
Best starting place - buy the CD's of the back issues of the Frequency Control Symposium. Each issue has a couple dozen papers looking at the kinds of issues you are interested in. The symposium has been running for a lot of years, there are lots of papers to dig into.
Thanks for the pointer to the IEEE and although I've not found
reference to the CD's yet, I came across this base link to Frequency
Control, http://www.ieee-uffc.org/frequency_control/, and from there
I've found some useful documents. This link is particularly
interesting, http://www.ieee-uffc.org/frequency_control/teaching.asp.
Becomming a member to UFFC enables you to get access to their digital
archive. Highly recommended.

It is unfortunate that related articles in IEEE Instrumentation and
Measurements isn't accessable by the same means, but you can't win them all.

Cheers,
Magnus

_______________________________________________
time-nuts mailing list -- time-nuts-***@public.gmane.org
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Steve Rooke
2010-05-30 13:57:44 UTC
Permalink
Post by Magnus Danielson
Becomming a member to UFFC enables you to get access to their digital
archive. Highly recommended.
It is unfortunate that related articles in IEEE Instrumentation and
Measurements isn't accessable by the same means, but you can't win them all.
Thanks Magnus.

Steve
--
Steve Rooke - ZL3TUV & G8KVD
A man with one clock knows what time it is;
A man with two clocks is never quite sure.

_______________________________________________
time-nuts mailing list -- time-nuts-***@public.gmane.org
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Magnus Danielson
2010-05-30 14:11:49 UTC
Permalink
Steve,
Post by Steve Rooke
Post by Magnus Danielson
Becomming a member to UFFC enables you to get access to their digital
archive. Highly recommended.
It is unfortunate that related articles in IEEE Instrumentation and
Measurements isn't accessable by the same means, but you can't win them all.
Thanks Magnus.
You are most welcome. The UFFC has given me many hours of good reading
as I have researched various things.

You have to request the access, but they fix it quickly enought if only
they are in the office. Essentially all you need to know is here:
http://www.ieee-uffc.org/main/membership.asp

I actually ordered the DVD set since it is obvious I would benefit from
having it.

Cheers,
Magnus

_______________________________________________
time-nuts mailing list -- time-nuts-***@public.gmane.org
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Murray Greenman
2010-05-27 18:50:58 UTC
Permalink
Jim,
Getting back to the original question, I'd suggest that the Epson-Toyocom FC-12M would be an appropriate device. See http://ndap3-net.ebz.epson.co.jp/w/www/PDFS/epdoc_qd.nsf/WE_khz_unit/737394E88F20BFBA4925736B002B4C07?OpenDocument. It is a surface mount device, very small, 2 x 1.2 x 0.6mm, and is a true tuning fork, but manufactured using QMEMs (Quartz micro-electro-mechanical, photolithographic techniques). The repeatability achieved by this manufacturing method gives a tightly specified parabolic frequency response (typical of tuning forks) with 25°C turnover temperature.

As others have mentioned, there are plenty of strategies around for managing the temperature control of the reference, most achieved by operating the crystal slightly high, and running a pulse swallower driven by a table or other algorithm. Clearly (as is done in the Dallas device) the closer you can associate the temperature sensor with the crystal, the lower the hysteresis and the better the accuracy. Of course a pulse swallower isn't something you'd use for frequency control, but is quite appropriate for time keeping over days and weeks.

I have tested the FC-12M in a low power oscillator (I used the Texas LVC1404 oscillator chip and an LM94022 silicon temperature sensor), and it gave very nice results, no dips or jumps over temperature, and reasonable phase noise (although this isn't an issue with RTCs). I modelled the temperature response in EXCEL and it closely followed the 2nd order model.

Regards,
Murray Greenman

-----Original Message-----
From: time-nuts-bounces-***@public.gmane.org [mailto:time-nuts-bounces-***@public.gmane.org] On Behalf Of time-nuts-request-***@public.gmane.org
Sent: Thursday, 27 May 2010 5:31 p.m.
To: time-nuts-***@public.gmane.org
Subject: time-nuts Digest, Vol 70, Issue 75

Send time-nuts mailing list submissions to
time-nuts-***@public.gmane.org

To subscribe or unsubscribe via the World Wide Web, visit
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
or, via email, send a message with subject or body 'help' to
time-nuts-request-***@public.gmane.org

You can reach the person managing the list at
time-nuts-owner-***@public.gmane.org

When replying, please edit your Subject line so it is more specific
than "Re: Contents of time-nuts digest..."


Today's Topics:

1. Re: For your museum only Loran-C monitor (paul swed)
2. Re: IRIG B (jimlux)
3. Re: Digital tight PLL method (Steve Rooke)
4. Re: Digital tight PLL method (J. L. Trantham)
5. Re: Digital tight PLL method (Bob Camp)
6. Re: IRIG B (Robert Atkinson)
7. Most accurate small crystal (Jim Palfreyman)
8. Re: Most accurate small crystal (Brooke Clarke)


----------------------------------------------------------------------

Message: 1
Date: Wed, 26 May 2010 08:44:16 -0400
From: paul swed <paulswedb-***@public.gmane.org>
Subject: Re: [time-nuts] For your museum only Loran-C monitor
To: Discussion of precise time and frequency measurement
<time-nuts-***@public.gmane.org>
Message-ID:
<AANLkTimW7V-LPhnXADNrj__nYuXAHB1AoBS-v4rEk0Up-JsoAwUIsXosN+***@public.gmane.org>
Content-Type: text/plain; charset=ISO-8859-1

Great ?
Have to respond later
Thats why I designed the loran c simulator. It works well. All I need to
do
is hook it to an antenna and away you go. Maybe a small power amp would
be
handy. Say 100KW? Antennas the real killer. I think its a zoning problem.
What's going to happen to that chunk of spectrum?
How much power/antenna would it take to make a signal that was useful out
to
1 mile? 100 miles?
What are the chances the FCC would let amateurs run timing and/or location
services on that band? 1/2 :), but there might be something interesting in
there.
--
These are my opinions, not necessarily my employer's. I hate spam.
_______________________________________________
To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
------------------------------

Message: 2
Date: Wed, 26 May 2010 06:16:49 -0700
From: jimlux <jimlux-***@public.gmane.org>
Subject: Re: [time-nuts] IRIG B
To: Discussion of precise time and frequency measurement
<time-nuts-***@public.gmane.org>
Message-ID: <4BFD1F41.2050900-***@public.gmane.org>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Hi
Strange as it seems, *stocking* the R's and C's can be an issue. There's also placement cost. Based on some of the numbers you see, the cross over point (IC to odd value R's and C's) is amazingly low. I'm not saying any of that's right, just that it's the way a lot of companies roll up the costs.
Bob
Not surprising.. the cost to stock, pick, place, solder is probably the
same for a small IC and a R or C. So, the only possible saving would be
if the IC is a LOT more expensive than a single or two Rs or Cs.

There might be a power dissipation difference, or a temperature range
difference that would push you one way or another.



------------------------------

Message: 3
Date: Thu, 27 May 2010 02:50:42 +1200
From: Steve Rooke <sar10538-***@public.gmane.org>
Subject: Re: [time-nuts] Digital tight PLL method
To: Discussion of precise time and frequency measurement
<time-nuts-***@public.gmane.org>
Message-ID:
<AANLkTin0L2YbApgu0h5FcxxwjV-_67xd6AsFvY3c8xzm-JsoAwUIsXosN+***@public.gmane.org>
Content-Type: text/plain; charset=ISO-8859-1
Can't beat a simple analog version of ?NIST's "Tight Phase-Lock Loop Method
of measuring Freq stability".
http://tf.nist.gov/phase/Properties/one.htm#oneone ? ?Fig 1.7"
I can't seem to access that link, can anyone else?

Steve
--
Steve Rooke - ZL3TUV & G8KVD
A man with one clock knows what time it is;
A man with two clocks is never quite sure.



------------------------------

Message: 4
Date: Wed, 26 May 2010 10:06:57 -0500
From: "J. L. Trantham" <jltran-***@public.gmane.org>
Subject: Re: [time-nuts] Digital tight PLL method
To: "'Discussion of precise time and frequency measurement'"
<time-nuts-***@public.gmane.org>
Message-ID: <00c701cafce5$16fc8a40$44f59ec0$@net>
Content-Type: text/plain; charset="iso-8859-1"

Worked fine for me and very informative.

Joe

-----Original Message-----
From: time-nuts-bounces-***@public.gmane.org [mailto:time-nuts-bounces-***@public.gmane.org] On
Behalf Of Steve Rooke
Sent: Wednesday, May 26, 2010 9:51 AM
To: Discussion of precise time and frequency measurement
Subject: Re: [time-nuts] Digital tight PLL method
Can't beat a simple analog version of ?NIST's "Tight Phase-Lock Loop
Method
of measuring Freq stability".
http://tf.nist.gov/phase/Properties/one.htm#oneone ? ?Fig 1.7"
I can't seem to access that link, can anyone else?

Steve
--
Steve Rooke - ZL3TUV & G8KVD
A man with one clock knows what time it is;
A man with two clocks is never quite sure.

_______________________________________________
time-nuts mailing list -- time-nuts-***@public.gmane.org
To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.




------------------------------

Message: 5
Date: Wed, 26 May 2010 11:58:03 -0400
From: "Bob Camp" <lists-***@public.gmane.org>
Subject: Re: [time-nuts] Digital tight PLL method
To: "'Discussion of precise time and frequency measurement'"
<time-nuts-***@public.gmane.org>
Message-ID: <069FBEBD95394AA4971B3D08B59EDE1F-mxmB69U+***@public.gmane.org>
Content-Type: text/plain; charset="iso-8859-1"

Hi

Works from here. It's a very brief description of pretty much every way in
the world to measure frequency.

Bob

-----Original Message-----
From: time-nuts-bounces-***@public.gmane.org [mailto:time-nuts-bounces-***@public.gmane.org] On
Behalf Of Steve Rooke
Sent: Wednesday, May 26, 2010 10:51 AM
To: Discussion of precise time and frequency measurement
Subject: Re: [time-nuts] Digital tight PLL method
Can't beat a simple analog version of ?NIST's "Tight Phase-Lock Loop
Method
of measuring Freq stability".
http://tf.nist.gov/phase/Properties/one.htm#oneone ? ?Fig 1.7"
I can't seem to access that link, can anyone else?

Steve
--
Steve Rooke - ZL3TUV & G8KVD
A man with one clock knows what time it is;
A man with two clocks is never quite sure.

_______________________________________________
time-nuts mailing list -- time-nuts-***@public.gmane.org
To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.






------------------------------

Message: 6
Date: Wed, 26 May 2010 20:17:02 +0000 (GMT)
From: Robert Atkinson <robert8rpi-/E1597aS9LT10XsdtD+***@public.gmane.org>
Subject: Re: [time-nuts] IRIG B
To: Discussion of precise time and frequency measurement
<time-nuts-***@public.gmane.org>
Message-ID: <424356.9169.qm-ZxlQ8pIuIvXGRxTy+Q50vsz6deESKz/***@public.gmane.org>
Content-Type: text/plain; charset=iso-8859-1

A good example is the good old 555 timer. A cmos 55 is 41 cents and needs a timing cap and resistor. A pic10f200 is 31 cents and needs no support components. You can also do fancy timing with the pic. I hate to say it but there is really no contest once you have the tools (free complier and a programmer for <$50). Just a shame it's killing analog skills.
?
Robert G8RPI.?

--- On Wed, 26/5/10, jimlux <jimlux-***@public.gmane.org> wrote:


From: jimlux <jimlux-***@public.gmane.org>
Subject: Re: [time-nuts] IRIG B
To: "Discussion of precise time and frequency measurement" <time-***@febo.com>
Date: Wednesday, 26 May, 2010, 14:16
Hi
Strange as it seems, *stocking* the R's and C's can be an issue. There's also placement cost. Based on some of the numbers you see, the cross over point (IC to odd value R's and C's) is amazingly low. I'm not saying any of that's right, just that it's the way a lot of companies roll up the costs.
Bob
Not surprising.. the cost to stock, pick, place, solder is probably the same for a small IC and a R or C.? So, the only possible saving would be? if the IC is a LOT more expensive than a single or two Rs or Cs.

There might be a power dissipation difference, or a temperature range difference that would push you one way or another.

_______________________________________________
time-nuts mailing list -- time-nuts-***@public.gmane.org
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.





------------------------------

Message: 7
Date: Thu, 27 May 2010 14:22:18 +1000
From: Jim Palfreyman <jim77742-***@public.gmane.org>
Subject: [time-nuts] Most accurate small crystal
To: Discussion of precise time and frequency measurement
<time-nuts-***@public.gmane.org>
Message-ID:
<AANLkTimNBz5wwBJ8dR0_eNZrFPxcKWlklxFCa7cx6yV9-JsoAwUIsXosN+***@public.gmane.org>
Content-Type: text/plain; charset=ISO-8859-1

Hi All,

I have a RSA SecurID device which I use to log in to my work's VPN. For
those of you not familiar with these, they show you a 6 digit number that
you use (combined with a PIN) to log in. This number changes every minute.
The changing of this number lines up with the servers at the other end and,
as I understand it, they do take into account gradual drift. These devices
have a limited programmed lifetime of three years as well. So the internal
clock on one of these needs to be decent.

I have timed the accuracy of this internal clock and have found it to be
pretty good so far. 17 days ago it was ticking over at 21.8 sec past the
minute and a quick visual inspection today and it was still *very* close to
that. I will confirm it properly tonight.

Now this device travels around with me, like a wristwatch, but is not on my
person. It is usually in my laptop bag. So it is subject to quite varying
temperatures.

So it's been nearly 3 weeks and this device has barely changed 0.1 to 0.2
secs tops. Now that's not bad compared to digital watches.

I presume it has a quartz crystal in it, obviously no oven, but my question
is:

What is the best crystal you can get on the market today that would work in
a watch type device with very little power available? What's the best
accuracy that can be expected?

I know straight out of the factory they can be pretty spot on (as my Casio G
Shock was) but it has aged markedly and now gains a couple of seconds a
week. So what is the best we can expect long term too?

Regards,

Jim Palfreyman


------------------------------

Message: 8
Date: Wed, 26 May 2010 22:30:36 -0700
From: Brooke Clarke <brooke-***@public.gmane.org>
Subject: Re: [time-nuts] Most accurate small crystal
To: Discussion of precise time and frequency measurement
<time-nuts-***@public.gmane.org>
Message-ID: <4BFE037C.7060301-***@public.gmane.org>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed

Hi Jim:

The last time I looked into this it was the DS3231 (now part of Maxim).
See the live demo:
http://www.maxim-ic.com/products/timers/DS3231_demo/

Have Fun,

Brooke Clarke
http://www.PRC68.com
Hi All,
I have a RSA SecurID device which I use to log in to my work's VPN. For
those of you not familiar with these, they show you a 6 digit number that
you use (combined with a PIN) to log in. This number changes every minute.
The changing of this number lines up with the servers at the other end and,
as I understand it, they do take into account gradual drift. These devices
have a limited programmed lifetime of three years as well. So the internal
clock on one of these needs to be decent.
I have timed the accuracy of this internal clock and have found it to be
pretty good so far. 17 days ago it was ticking over at 21.8 sec past the
minute and a quick visual inspection today and it was still *very* close to
that. I will confirm it properly tonight.
Now this device travels around with me, like a wristwatch, but is not on my
person. It is usually in my laptop bag. So it is subject to quite varying
temperatures.
So it's been nearly 3 weeks and this device has barely changed 0.1 to 0.2
secs tops. Now that's not bad compared to digital watches.
I presume it has a quartz crystal in it, obviously no oven, but my question
What is the best crystal you can get on the market today that would work in
a watch type device with very little power available? What's the best
accuracy that can be expected?
I know straight out of the factory they can be pretty spot on (as my Casio G
Shock was) but it has aged markedly and now gains a couple of seconds a
week. So what is the best we can expect long term too?
Regards,
Jim Palfreyman
_______________________________________________
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
------------------------------

_______________________________________________
time-nuts mailing list
time-nuts-***@public.gmane.org
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts

End of time-nuts Digest, Vol 70, Issue 75
*****************************************

_______________________________________________
time-nuts mailing list -- time-nuts-***@public.gmane.org
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Luke.Hoffmann-o70QjBUuf8rQT0dZR+
2010-05-28 02:33:56 UTC
Permalink
Hi Jim,

If you are looking for low power (< 80 mW @ 24 degrees C), stable
(better that 1ppb per day) & accurate (within 5ppb of 1Hz from factory)
then have a look at these devices.

They also have an input pin that takes a GPS 1pps ref used to discipline
them. (Used by the device to recalibrate itself.)

www.l-3com.com/nautronix/products/hptr.html
www.l-3com.com/nautronix/pdfs/HPTR_Rev_2.1.pdf

Regards
Luke Hoffmann




-----Original Message-----
From: time-nuts-bounces-***@public.gmane.org [mailto:time-nuts-bounces-***@public.gmane.org] On
Behalf Of Jim Palfreyman
Sent: Thursday, 27 May 2010 12:22 PM
To: Discussion of precise time and frequency measurement
Subject: [time-nuts] Most accurate small crystal

Hi All,

I have a RSA SecurID device which I use to log in to my work's VPN. For
those of you not familiar with these, they show you a 6 digit number
that
you use (combined with a PIN) to log in. This number changes every
minute.
The changing of this number lines up with the servers at the other end
and,
as I understand it, they do take into account gradual drift. These
devices
have a limited programmed lifetime of three years as well. So the
internal
clock on one of these needs to be decent.

I have timed the accuracy of this internal clock and have found it to be
pretty good so far. 17 days ago it was ticking over at 21.8 sec past the
minute and a quick visual inspection today and it was still *very* close
to
that. I will confirm it properly tonight.

Now this device travels around with me, like a wristwatch, but is not on
my
person. It is usually in my laptop bag. So it is subject to quite
varying
temperatures.

So it's been nearly 3 weeks and this device has barely changed 0.1 to
0.2
secs tops. Now that's not bad compared to digital watches.

I presume it has a quartz crystal in it, obviously no oven, but my
question
is:

What is the best crystal you can get on the market today that would work
in
a watch type device with very little power available? What's the best
accuracy that can be expected?

I know straight out of the factory they can be pretty spot on (as my
Casio G
Shock was) but it has aged markedly and now gains a couple of seconds a
week. So what is the best we can expect long term too?

Regards,

Jim Palfreyman
_______________________________________________
time-nuts mailing list -- time-nuts-***@public.gmane.org
To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

_______________________________________________
time-nuts mailing list -- time-nuts-***@public.gmane.org
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Don Latham
2010-05-28 05:46:24 UTC
Permalink
As usual, no prices. If you have to ask, you can't afford one :-)
Don

----- Original Message -----
From: <Luke.Hoffmann-o70QjBUuf8rQT0dZR+***@public.gmane.org>
To: "Discussion of precise time and frequency measurement"
<time-nuts-***@public.gmane.org>
Sent: Thursday, May 27, 2010 8:33 PM
Subject: Re: [time-nuts] Most accurate small crystal
Post by Luke.Hoffmann-o70QjBUuf8rQT0dZR+
Hi Jim,
(better that 1ppb per day) & accurate (within 5ppb of 1Hz from factory)
then have a look at these devices.
They also have an input pin that takes a GPS 1pps ref used to discipline
them. (Used by the device to recalibrate itself.)
www.l-3com.com/nautronix/products/hptr.html
www.l-3com.com/nautronix/pdfs/HPTR_Rev_2.1.pdf
Regards
Luke Hoffmann
-----Original Message-----
Behalf Of Jim Palfreyman
Sent: Thursday, 27 May 2010 12:22 PM
To: Discussion of precise time and frequency measurement
Subject: [time-nuts] Most accurate small crystal
Hi All,
I have a RSA SecurID device which I use to log in to my work's VPN. For
those of you not familiar with these, they show you a 6 digit number that
you use (combined with a PIN) to log in. This number changes every minute.
The changing of this number lines up with the servers at the other end and,
as I understand it, they do take into account gradual drift. These devices
have a limited programmed lifetime of three years as well. So the internal
clock on one of these needs to be decent.
I have timed the accuracy of this internal clock and have found it to be
pretty good so far. 17 days ago it was ticking over at 21.8 sec past the
minute and a quick visual inspection today and it was still *very* close to
that. I will confirm it properly tonight.
Now this device travels around with me, like a wristwatch, but is not on my
person. It is usually in my laptop bag. So it is subject to quite varying
temperatures.
So it's been nearly 3 weeks and this device has barely changed 0.1 to 0.2
secs tops. Now that's not bad compared to digital watches.
I presume it has a quartz crystal in it, obviously no oven, but my question
What is the best crystal you can get on the market today that would work in
a watch type device with very little power available? What's the best
accuracy that can be expected?
I know straight out of the factory they can be pretty spot on (as my Casio G
Shock was) but it has aged markedly and now gains a couple of seconds a
week. So what is the best we can expect long term too?
Regards,
Jim Palfreyman
_______________________________________________
To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
_______________________________________________
To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
_______________________________________________
time-nuts mailing list -- time-nuts-***@public.gmane.org
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Hal Murray
2010-05-28 06:32:58 UTC
Permalink
Post by Bob Camp
Where do you want to start?
Isn't that HP/Agilent App-Note with the three wrist-watches the place to
start ? 1289 or something ?
If anybody finds it, please let me/us know.
--
These are my opinions, not necessarily my employer's. I hate spam.




_______________________________________________
time-nuts mailing list -- time-nuts-***@public.gmane.org
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
mike cook
2010-05-28 07:07:43 UTC
Permalink
This I think
http://www.allanstime.com/Publications/DWA/Science_Timekeeping/TheScienceOfTimekeeping.pdf
Post by Hal Murray
Post by Bob Camp
Where do you want to start?
Isn't that HP/Agilent App-Note with the three wrist-watches the place to
start ? 1289 or something ?
If anybody finds it, please let me/us know.
_______________________________________________
time-nuts mailing list -- time-nuts-***@public.gmane.org
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Steve Rooke
2010-05-28 13:52:21 UTC
Permalink
Post by mike cook
This I think
http://www.allanstime.com/Publications/DWA/Science_Timekeeping/TheScienceOfTimekeeping.pdf
Thanks for the link Mike.

Steve
--
Steve Rooke - ZL3TUV & G8KVD
A man with one clock knows what time it is;
A man with two clocks is never quite sure.

_______________________________________________
time-nuts mailing list -- time-nuts-***@public.gmane.org
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Hal Murray
2010-11-12 11:04:13 UTC
Permalink
This was one of the things that I wondered about: How large currents are
used ?
Can't be too much because that would lead to self-heating...
I don't know about temperature, but I've worked with load cells. The data
sheet I just found on the web said 350 ohms at 10 V, so 28 mA, or 14 each
side. That seems like a reasonable ballpark.

That's 1/3 watt, but they are mounted on a blob of steel the size of my fist.
--
These are my opinions, not necessarily my employer's. I hate spam.
Hal Murray
2011-01-04 20:29:45 UTC
Permalink
[From a couple of weeks ago...]
Opto-isolater? Why not just use fiber cable between cards.
I have only ever seen point-to-point versions, I needed 7 cards connected.
You can make a multi-drop setup with 2n unidirectional optical links. The
trick is to send all the links to a common node and OR the signal together
there.

Consider the old coax Ethernet... The coax was an OR gate.

There were rules about the length (timing) of the long coax and the short
drop cables between the host and the transceiver. You can use longer drop
cables and still meet the rules if you use a shorter coax. Push that all the
way and you get close to the typical setup that used twisted pair phone
wiring to a hub (not switch).

The early transceivers used 3 signal pairs, one for collision, but you can
encode that on the data link.

You can replace the copper link in that setup with fiber.

One disadvantage of this approach is that the hub needs power.
--
These are my opinions, not necessarily my employer's. I hate spam.
Hal Murray
2011-01-16 22:14:29 UTC
Permalink
Where to find the instruments is obviously something that should be
abstracted once more people starts to use this, I'm open to both suggestions
and patches in that respect.
How about a config file?

I think it needs something like:
A symbolic name (used by top level PYLT code)
The type of instrument (from the set of supported ones)
The type of access hardware
/dev/whatever
Possibly more info like
GPIB Address

I'm thinking of something like a line for each instrument.

There is overlap between the type of access hardware and the /dev/whatever
If you use symbolic links you can derive one from the other.
--
These are my opinions, not necessarily my employer's. I hate spam.
Poul-Henning Kamp
2011-01-16 22:25:31 UTC
Permalink
Post by Hal Murray
How about a config file?
Yeah, that bit is obvious, the trouble is all the "something like" :-)

I'm a firm beliver in Gettys principles for software development,
number 3 of which says:

3.The only thing worse than generalizing from one example
is generalizing from no examples at all.

But rather than write a lot of code to deal with config file reading,
we should simply use python:

Create a pylt_site.py containing something like:

import hp34401a
import hp33120a

def hp34401a():
return hp34401a.hp34401a(name="/dev/ttyfoo", adr=11)

def hp33120a():
return hp33120a.hp33120a(name="/dev/ttybar", adr=5)

and then use that in your programs:

include pylt_site

g=pylt_site.hp33120a()
m=pylt_site.hp34401a()

# do stuff.

Problem solved ?
--
Poul-Henning Kamp | UNIX since Zilog Zeus 3.20
phk-***@public.gmane.org | TCP/IP since RFC 956
FreeBSD committer | BSD since 4.3-tahoe
Never attribute to malice what can adequately be explained by incompetence.
Chuck Harris
2011-01-16 22:49:23 UTC
Permalink
That would be my preferred method.

-Chuck Harris

Poul-Henning Kamp wrote:
...
Post by Poul-Henning Kamp
import hp34401a
import hp33120a
return hp34401a.hp34401a(name="/dev/ttyfoo", adr=11)
return hp33120a.hp33120a(name="/dev/ttybar", adr=5)
include pylt_site
g=pylt_site.hp33120a()
m=pylt_site.hp34401a()
# do stuff.
Problem solved ?
Hal Murray
2011-03-09 08:29:26 UTC
Permalink
The reason LORAN-C got killed was that DoD/DHS couldn't jam it if a
terrorist with a suitcase-nuke was trying to find his way to Congress.
Why can't they jam it?

It's just a matter of signal strength. If want to jam a small region, they
can use a local transmitter. If they want to disable a large region, they
can turn off the transmitters or have them transmit garbage.
--
These are my opinions, not necessarily my employer's. I hate spam.
Poul-Henning Kamp
2011-03-09 08:33:52 UTC
Permalink
Post by Hal Murray
The reason LORAN-C got killed was that DoD/DHS couldn't jam it if a
terrorist with a suitcase-nuke was trying to find his way to Congress.
Why can't they jam it?
They can, All the need is 3 truck mounted generators, a truck with
the electronics and then drape a superconducting loopantenna over
a couple of skyskrapers...

DHS was concerned this would not be "immediately available" in an
emergency, whereas pretty much any army or air-force base can deliver
GPS jamming on minuted notice.
--
Poul-Henning Kamp | UNIX since Zilog Zeus 3.20
phk-***@public.gmane.org | TCP/IP since RFC 956
FreeBSD committer | BSD since 4.3-tahoe
Never attribute to malice what can adequately be explained by incompetence.
Hal Murray
2011-07-15 07:10:04 UTC
Permalink
m(pop,china) = 1.5e9 * 50kg = 7.5e10 kg
m(earth) = 5.97e24 kg
A ratio of roughly 8e13...
Neat. Thanks.

I think you missed another big factor. The height of a chair is tiny
relative to the radius of the Earth.

Lets call a chair 1 meter. The radius of the Earth is 6,3xx km or 6E6
meters. So that makes a ratio of more than 1E20.

1E14 we might be able to notice. 1E20 will be lost in the noise.

Does anybody have a good graph for summer vs winter? I'd expect snow loading
might be big enough to show up.

--------

PS: You need to round down the earth's moment of inertia by 2/5 because it's
a uniform density sphere rather than the mass being concentrated at the
radius on the equator. (mass next to the axis doesn't help)

I'm assuming all the people in China are on the equator. There is a cos
latitude fudge factor to correct for that.
--
These are my opinions, not necessarily my employer's. I hate spam.
Tom Van Baak (lab/iPad)
2011-07-15 07:48:50 UTC
Permalink
Post by Hal Murray
1E14 we might be able to notice
Hal,

No. Look at the adev of the earth (earlier posting). The length of earth day varies in the *milli*second range, day to day. VLBI measurements are under 0.1 millisecond, which comes to about 1e-9 resolution.

Realize that none of the NASA "earthquake may have shortened" press releases are about real measurements of rotation. They are just impressive models of changes in momentum. The predictions are in the *micro*second range. The press does not always distinguish between milli and micro.

/tvb
Jim Lux
2011-07-15 13:14:00 UTC
Permalink
Post by Tom Van Baak (lab/iPad)
Post by Hal Murray
1E14 we might be able to notice
Hal,
No. Look at the adev of the earth (earlier posting). The length of earth day varies in the *milli*second range, day to day. VLBI measurements are under 0.1 millisecond, which comes to about 1e-9 resolution.
Realize that none of the NASA "earthquake may have shortened" press releases are about real measurements of rotation. They are just impressive models of changes in momentum. The predictions are in the *micro*second range. The press does not always distinguish between milli and micro.
And, there's a somewhat non-noise-free-channel from the guys doing the
calculations to the public affairs officer to the media.

This kind of thing is actually sort of interesting in a planetary sense.
While earth is pretty stable, there are places where there are a lot
more earthquakes and internal tidal forces (Jupiter's moons) and changes
in rotation rate of the moons might be detectable by radar.

Is Io gradually slowing? There's also coupling to Jupiter, of course
(e.g. our Moon having a rotation rate synced to orbital period)

What about Mercury?
Jim Lux
2011-07-15 13:10:28 UTC
Permalink
Post by Hal Murray
Does anybody have a good graph for summer vs winter? I'd expect snow loading
might be big enough to show up.
This is the kind of thing that Richard Gross at JPL fools with. As I
recall, atmospheric drag changes on a cyclical basis too. And
solid/liquid tides

Among other sources, data comes from all those geodetic GPS receiving
stations feeding into gipsy
Hal Murray
2011-07-15 09:02:07 UTC
Permalink
Post by Tom Van Baak (lab/iPad)
No. Look at the adev of the earth (earlier posting). The length of earth day
varies in the *milli*second range, day to day. VLBI measurements are under
0.1 millisecond, which comes to about 1e-9 resolution.
Realize that none of the NASA "earthquake may have shortened" press releases
are about real measurements of rotation. They are just impressive models of
changes in momentum. The predictions are in the *micro*second range. The
press does not always distinguish between milli and micro.
Ouch. Thanks for the correction/heads-up.

[Here is the graph:
Loading Image...
and background
http://www.leapsecond.com/museum/earth/ ]


Do you have any guesses on the short term stability? (where short means left
of your graph)

Suppose all those Chinese people hopped up and down on their chair in
synchrony. What would be the ideal timing? I think I'm fishing for
something like a chopper amplifier, but I'm not sure how I would explain a
chopper to somebody who didn't know about them.
--
These are my opinions, not necessarily my employer's. I hate spam.
Hal Murray
2012-03-16 10:08:47 UTC
Permalink
Hmm... do you mean you want to store all samples of an hour and then
avarage over it?
That would be the ideal way to do it, since it would make one heck of a comb
filter and eliminate pretty much anything else.
That only works if your reference clock is stable enough over the collecting
period.

I'm far from a DSP wizard. Assume I have a signal at 100 kHz with a
bandwidth of 1 kHz. (or pick any numbers you like) How many samples per
second do I need? How stable does my sampling clock need to be if I collect
data for for N seconds?

Is "stable" the right term? What's the right way to think about the question
I'm trying to ask?
--
These are my opinions, not necessarily my employer's. I hate spam.
Hal Murray
2012-03-28 11:31:22 UTC
Permalink
Why do we need really accurate clocks?
Time or frequency?
"Because accurate clocks is the central technology that makes GPS, mobile
phones and the internet work."
What part of the internet depends upon accurate clocks?

Ethernet, for example, requires roughly similar frequencies but doesn't care
at all about time.
--
These are my opinions, not necessarily my employer's. I hate spam.
Poul-Henning Kamp
2012-03-28 13:25:16 UTC
Permalink
Post by Hal Murray
"Because accurate clocks is the central technology that makes GPS, mobile
phones and the internet work."
What part of the internet depends upon accurate clocks?
SONET at the bottom and eBay endings at the top and a lot of stuff
in the middle.
--
Poul-Henning Kamp | UNIX since Zilog Zeus 3.20
phk-***@public.gmane.org | TCP/IP since RFC 956
FreeBSD committer | BSD since 4.3-tahoe
Never attribute to malice what can adequately be explained by incompetence.
Hal Murray
2012-05-06 02:12:12 UTC
Permalink
If you cannot apply the negative sawtooth, you will get better results by
disciplining almost any random quartz xtal, ovenized or not to the GPS,
divide it down to PPS and then discipline the PRS10 to that.
I don't understand that. What am I missing?

In the simple case, the main parameter for a PLL is the time constant or
cutoff frequency on the low pass filter. Normally, if the time constant is
low enough the PLL will track the incoming wobbles. If the time constant it
long enough, the PLL will ignore the short term wobbles yet track the long
term wanders.

In this context, the Rb is very stable so the time constant can be very long.
Unfortunately, the time constant of hanging bridges can also be very long,
especially if your temperature regulation is good (which is probably a good
first step for anything like this).

What does putting an extra crystal between GPS and Rb do? What time
constants do I use on the PLL from GPS to xtal and xtal to Rb?
--
These are my opinions, not necessarily my employer's. I hate spam.
Poul-Henning Kamp
2012-05-06 06:47:01 UTC
Permalink
Post by Hal Murray
If you cannot apply the negative sawtooth, you will get better results by
disciplining almost any random quartz xtal, ovenized or not to the GPS,
divide it down to PPS and then discipline the PRS10 to that.
I don't understand that. What am I missing?
You are missing that the average of the 1PPS pulse only can be trusted
to be zero over a timescale of many hours.

This is an error-source distinct from GPS reception, caused by the
picking a preexisting flank nearest to the epoch, with no attempt
to keep the average of the resulting error zero.

Imagine that the GPS receivers clock happens to run on perfect
frequency for a while: That means that the flank used to generate
the PPS will have a fixed location relative to the epoch, for instance
always 12 nanoseconds early or late.

I belive that some GPS receivers have deliberately de-tuned Xtals
for this very reason, but unfortunately that is only a partial
fix, as the problem is a modulus-issue, so not only is perfect
frequency bad, but "perfect +/- n Hz" is equally bad.

The hanging bridges Tom has plots of on leapsecond.com, arises when
the frequency of the GPS xtal changes.

At one point I tried putting a 1W resister close to the xtal and
feed it with a very slow sine-wave to see if "jittering" it would
get me an average of zero of shorter timespans. My experiment
was inconclusive, but the idea is not unsound.
--
Poul-Henning Kamp | UNIX since Zilog Zeus 3.20
phk-***@public.gmane.org | TCP/IP since RFC 956
FreeBSD committer | BSD since 4.3-tahoe
Never attribute to malice what can adequately be explained by incompetence.
Magnus Danielson
2012-05-06 09:33:34 UTC
Permalink
Post by Poul-Henning Kamp
Post by Hal Murray
If you cannot apply the negative sawtooth, you will get better results by
disciplining almost any random quartz xtal, ovenized or not to the GPS,
divide it down to PPS and then discipline the PRS10 to that.
I don't understand that. What am I missing?
You are missing that the average of the 1PPS pulse only can be trusted
to be zero over a timescale of many hours.
This is an error-source distinct from GPS reception, caused by the
picking a preexisting flank nearest to the epoch, with no attempt
to keep the average of the resulting error zero.
Imagine that the GPS receivers clock happens to run on perfect
frequency for a while: That means that the flank used to generate
the PPS will have a fixed location relative to the epoch, for instance
always 12 nanoseconds early or late.
I belive that some GPS receivers have deliberately de-tuned Xtals
for this very reason, but unfortunately that is only a partial
fix, as the problem is a modulus-issue, so not only is perfect
frequency bad, but "perfect +/- n Hz" is equally bad.
Since GPSes typically have a TCXO, it is not as stable in relationship
to temperature as you would wish. Whenever the TCXO frequency is near an
integer multiple of the PPS, then it will select the same number of
clock cycles over a long run, and then averaging effects of alternating
between two nearby clock-cycles are gone.

It would be possible to do error accumulation in the GPS that would
steer the PPS to on average be more accurate. It should not be that many
lines of code to achieve it.

Oh, if we only had hooks so we could insert some code into the GPS
receivers...
Post by Poul-Henning Kamp
The hanging bridges Tom has plots of on leapsecond.com, arises when
the frequency of the GPS xtal changes.
At one point I tried putting a 1W resister close to the xtal and
feed it with a very slow sine-wave to see if "jittering" it would
get me an average of zero of shorter timespans. My experiment
was inconclusive, but the idea is not unsound.
Indeed. But you would not be "jittering" it, you would "wander" it. It
would be a good experiment. Another approach would be steer the XTAL
using the statistics and then control the heater.

Locking it up to .5 Hz (i.e. 50% of the transitions is short and the
others long) would be possible, but I don't think you gain much in
lowering the quantization floor, so I rather think a slow triangle
modulation (that you also remove from the TI measurements) is preferred.

Cheers,
Magnus
Bill Hawkins
2012-05-06 19:01:31 UTC
Permalink
Here is a different tactic for disciplining Rb from GPS/TXCO -

Consider the relatively (relative to a second) long stability of
an Rb oscillator and the not-so-good stability of GPS. Perhaps
using 1 PPS for a sampling period for stabilizing Rb is way too
short. Maybe 1000 seconds is better. That's way too long for an
analog integrator to do, so a microcomputer is required.

Count the Rb and GPS 1 PPS signals for 1000 counts of Rb 1 PPS.
You'll need to interpolate between 1 PPS GPS intervals to the
level of accuracy required, so maybe we count cycles of 10 MHz
from the GPS, using as many registers (cascaded integer counters)
as required for 1 E10 (or more) counts (2540BE400 hex). At the
end of 1000 seconds, use the difference between the lowest counter
reading and 0xEB400 (times an appropriate gain) to tweak the value
for the DAC doing the fine correction to the control voltage or
current. Use the upper counters for a sanity check on the reading.

As may be evident, I have never tried to discipline an Rb, but I
am well aware of the effect of sample time on the control of a
long time constant loop. True, the effect is strongest when dead
time dominates the time constant, but that is not the case here.
Still, I think there is value in using a long sampling time for
the control action.

Comments accepted with enthusiasm.

Bill Hawkins
Randall Prentice
2012-05-06 20:12:23 UTC
Permalink
The paper should really be titled "Another way to discipline your Rb to GPS".

Wouldn't it make more sense given the phase noise in the Rb to base your correction on a moving average diffence.

1) Reduce the Rb phase noise
2) Carefully make it a multiple of a day to cancel diurnal effects of the ionosphere.

Regards
Randall
-----Original Message-----
From: time-nuts-bounces-***@public.gmane.org [mailto:time-nuts-bounces-***@public.gmane.org] On Behalf Of Bill Hawkins
Sent: Monday, 7 May 2012 7:02 a.m.
To: 'Discussion of precise time and frequency measurement'
Subject: Re: [time-nuts] Interesting paper: Don't GPSD' your Rb...

Here is a different tactic for disciplining Rb from GPS/TXCO -

Consider the relatively (relative to a second) long stability of an Rb oscillator and the not-so-good stability of GPS. Perhaps using 1 PPS for a sampling period for stabilizing Rb is way too short. Maybe 1000 seconds is better. That's way too long for an analog integrator to do, so a microcomputer is required.

Count the Rb and GPS 1 PPS signals for 1000 counts of Rb 1 PPS.
You'll need to interpolate between 1 PPS GPS intervals to the level of accuracy required, so maybe we count cycles of 10 MHz from the GPS, using as many registers (cascaded integer counters) as required for 1 E10 (or more) counts (2540BE400 hex). At the end of 1000 seconds, use the difference between the lowest counter reading and 0xEB400 (times an appropriate gain) to tweak the value for the DAC doing the fine correction to the control voltage or current. Use the upper counters for a sanity check on the reading.

As may be evident, I have never tried to discipline an Rb, but I am well aware of the effect of sample time on the control of a long time constant loop. True, the effect is strongest when dead time dominates the time constant, but that is not the case here.
Still, I think there is value in using a long sampling time for the control action.

Comments accepted with enthusiasm.

Bill Hawkins


_______________________________________________
time-nuts mailing list -- time-nuts-***@public.gmane.org To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Magnus Danielson
2012-05-06 21:44:50 UTC
Permalink
Post by Bill Hawkins
Here is a different tactic for disciplining Rb from GPS/TXCO -
Consider the relatively (relative to a second) long stability of
an Rb oscillator and the not-so-good stability of GPS. Perhaps
using 1 PPS for a sampling period for stabilizing Rb is way too
short. Maybe 1000 seconds is better. That's way too long for an
analog integrator to do, so a microcomputer is required.
Using analog integrator for 1 PPS stuff is hardish and your performance
will most probably suffer. You want your integrator to be digital as it
is clearly a better memory over time.

However, this is not to be confused with the time-constant, which should
be much higher than 1000 s to use the stability of the rubidium where
the GPS noise is worse. See PRS-10 manual for good illustration.
Post by Bill Hawkins
Count the Rb and GPS 1 PPS signals for 1000 counts of Rb 1 PPS.
You'll need to interpolate between 1 PPS GPS intervals to the
level of accuracy required, so maybe we count cycles of 10 MHz
from the GPS, using as many registers (cascaded integer counters)
as required for 1 E10 (or more) counts (2540BE400 hex). At the
end of 1000 seconds, use the difference between the lowest counter
reading and 0xEB400 (times an appropriate gain) to tweak the value
for the DAC doing the fine correction to the control voltage or
current. Use the upper counters for a sanity check on the reading.
Why wait with the updates of the DAC? By incrementally average for each
second I think you get a smoother transitions.

You sure want time-constants in excess of 1000 s, but you can achieve
that by using 1 s updates. Again, read the PRS-10 manual for a fairly
good description.
Post by Bill Hawkins
As may be evident, I have never tried to discipline an Rb, but I
am well aware of the effect of sample time on the control of a
long time constant loop. True, the effect is strongest when dead
time dominates the time constant, but that is not the case here.
Still, I think there is value in using a long sampling time for
the control action.
I fail to see the benefit, and I have many times learned the hard way to
see the downside of too low update rate.
Post by Bill Hawkins
Comments accepted with enthusiasm.
:)

Cheers,
Magnus
Bill Hawkins
2012-05-07 03:12:08 UTC
Permalink
Thanks for the analysis, Magnus.

A few other time constants might be interesting -

When a step change is made to the control voltage or current,
how long does it take for the oscillator to settle down to a
new value? Is it instantaneous compared to a second?

Do different components in different oscillators affect the
settling time?

It is not useful to make the next change before the last one
is complete, at least for sampled systems. Using counters
filters the change rather than taking a sample.

Bill Hawkins


-----Original Message-----
From: Magnus Danielson
Sent: Sunday, May 06, 2012 4:45 PM
To: time-nuts-***@public.gmane.org
Subject: Re: [time-nuts] Interesting paper: Don't GPSD' your Rb...
Post by Bill Hawkins
Here is a different tactic for disciplining Rb from GPS/TXCO -
Consider the relatively (relative to a second) long stability of
an Rb oscillator and the not-so-good stability of GPS. Perhaps
using 1 PPS for a sampling period for stabilizing Rb is way too
short. Maybe 1000 seconds is better. That's way too long for an
analog integrator to do, so a microcomputer is required.
Using analog integrator for 1 PPS stuff is hardish and your performance
will most probably suffer. You want your integrator to be digital as it
is clearly a better memory over time.

However, this is not to be confused with the time-constant, which should
be much higher than 1000 s to use the stability of the rubidium where
the GPS noise is worse. See PRS-10 manual for good illustration.
Post by Bill Hawkins
Count the Rb and GPS 1 PPS signals for 1000 counts of Rb 1 PPS.
You'll need to interpolate between 1 PPS GPS intervals to the
level of accuracy required, so maybe we count cycles of 10 MHz
from the GPS, using as many registers (cascaded integer counters)
as required for 1 E10 (or more) counts (2540BE400 hex). At the
end of 1000 seconds, use the difference between the lowest counter
reading and 0xEB400 (times an appropriate gain) to tweak the value
for the DAC doing the fine correction to the control voltage or
current. Use the upper counters for a sanity check on the reading.
Why wait with the updates of the DAC? By incrementally average for each
second I think you get a smoother transitions.

You sure want time-constants in excess of 1000 s, but you can achieve
that by using 1 s updates. Again, read the PRS-10 manual for a fairly
good description.
Post by Bill Hawkins
As may be evident, I have never tried to discipline an Rb, but I
am well aware of the effect of sample time on the control of a
long time constant loop. True, the effect is strongest when dead
time dominates the time constant, but that is not the case here.
Still, I think there is value in using a long sampling time for
the control action.
I fail to see the benefit, and I have many times learned the hard way to
see the downside of too low update rate.
Post by Bill Hawkins
Comments accepted with enthusiasm.
:)

Cheers,
Magnus

_______________________________________________
time-nuts mailing list -- time-nuts-***@public.gmane.org
To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Magnus Danielson
2012-05-07 06:57:19 UTC
Permalink
Hi Bill,
Post by Bill Hawkins
Thanks for the analysis, Magnus.
Always happy to contribute.
Post by Bill Hawkins
A few other time constants might be interesting -
When a step change is made to the control voltage or current,
how long does it take for the oscillator to settle down to a
new value? Is it instantaneous compared to a second?
It depends on the rubidium FLL bandwidth. For OCXO it is much quicker.
Post by Bill Hawkins
Do different components in different oscillators affect the
settling time?
For rubidiums, yes. The FLL bandwidth will have such a "lag" effect
until the slaved OXXO is back on track.
Post by Bill Hawkins
It is not useful to make the next change before the last one
is complete, at least for sampled systems. Using counters
filters the change rather than taking a sample.
If you flip back and forth, then it makes sense because your phase
deviations will be less.

You should however consider it when doing the stability analysis.

Cheers,
Magnus
Azelio Boriani
2012-05-07 07:56:54 UTC
Permalink
Magnus,
Post by Magnus Danielson
If you flip back and forth, then it makes sense because your phase
deviations >will be less.

Can you further explain this? Thanks.
Post by Magnus Danielson
Hi Bill,
Post by Bill Hawkins
Thanks for the analysis, Magnus.
Always happy to contribute.
A few other time constants might be interesting -
Post by Bill Hawkins
When a step change is made to the control voltage or current,
how long does it take for the oscillator to settle down to a
new value? Is it instantaneous compared to a second?
It depends on the rubidium FLL bandwidth. For OCXO it is much quicker.
Do different components in different oscillators affect the
Post by Bill Hawkins
settling time?
For rubidiums, yes. The FLL bandwidth will have such a "lag" effect until
the slaved OXXO is back on track.
It is not useful to make the next change before the last one
Post by Bill Hawkins
is complete, at least for sampled systems. Using counters
filters the change rather than taking a sample.
If you flip back and forth, then it makes sense because your phase
deviations will be less.
You should however consider it when doing the stability analysis.
Cheers,
Magnus
______________________________**_________________
To unsubscribe, go to https://www.febo.com/cgi-bin/**
mailman/listinfo/time-nuts<https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts>
and follow the instructions there.
Magnus Danielson
2012-05-07 16:52:14 UTC
Permalink
Azelio,
Post by Azelio Boriani
Magnus,
Post by Magnus Danielson
If you flip back and forth, then it makes sense because your phase
deviations>will be less.
Can you further explain this? Thanks.
Certainly!

Consider that you flip back and forth between two levels, let's just say
50% "high" and 50% "low", the rate of flips considering the
time-constant of the adjustment will matter, since if you let the time
be so long that the alignment "ring out" you will get maxium frequency
and phase deviation, where as if the rate of flipping is higher, then it
drift just a little towards "high" when it gets a "low" and drift
towards that... which gives you much smaller phase and frequency
deviations... and you stay very close to the 50% level between "high"
and "low". Essentially, the time-constant of the low-pass filtering will
for a higher rate fairly well dampen the changes where as slow changes
is in the pass-band with almost no low-pass filtering effects.

It is thus nothing stranger than a low-pass filter and a square-wave of
different frequencies.

Also, if you update to slowly, you will build up larger errors before
you have to steer back since you measured an error. Too low comparator
rate in a PLL forces heavy filtering just to lower that comparator
sub-frequency. Using high enough rate, the filtering process will be
more continuous rather than very obviously sampled oriented.

I've been bitten by this before. I've seen what too low comparator
frequency does to create instability and modulations.

Cheers,
Magnus
Azelio Boriani
2012-05-07 20:13:32 UTC
Permalink
OK, got it. Yes, something like the dithering with a DAC to increase the
resolution.
Post by Magnus Danielson
Azelio,
Post by Azelio Boriani
Magnus,
If you flip back and forth, then it makes sense because your phase
deviations>will be less.
Can you further explain this? Thanks.
Certainly!
Consider that you flip back and forth between two levels, let's just say
50% "high" and 50% "low", the rate of flips considering the time-constant
of the adjustment will matter, since if you let the time be so long that
the alignment "ring out" you will get maxium frequency and phase deviation,
where as if the rate of flipping is higher, then it drift just a little
towards "high" when it gets a "low" and drift towards that... which gives
you much smaller phase and frequency deviations... and you stay very close
to the 50% level between "high" and "low". Essentially, the time-constant
of the low-pass filtering will for a higher rate fairly well dampen the
changes where as slow changes is in the pass-band with almost no low-pass
filtering effects.
It is thus nothing stranger than a low-pass filter and a square-wave of
different frequencies.
Also, if you update to slowly, you will build up larger errors before you
have to steer back since you measured an error. Too low comparator rate in
a PLL forces heavy filtering just to lower that comparator sub-frequency.
Using high enough rate, the filtering process will be more continuous
rather than very obviously sampled oriented.
I've been bitten by this before. I've seen what too low comparator
frequency does to create instability and modulations.
Cheers,
Magnus
_______________________________________________
To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Magnus Danielson
2012-05-07 20:33:45 UTC
Permalink
Hi Azelio,
Post by Azelio Boriani
OK, got it. Yes, something like the dithering with a DAC to increase the
resolution.
Indeed. Now, consider now that the variations can come from any form of
noise source.

Another thing I've learned is that the longer you wait with a
correction, the larger the deviation becomes and lower frequency it will
have... and both makes it harder to suppress by filtering. It's really
just the same thing form another angle.

Cheers,
Magnus
Azelio Boriani
2012-05-07 20:59:45 UTC
Permalink
the larger the deviation becomes and lower frequency it will have... and
both makes it >harder to suppress by filtering.
Filtering at what level? Lengthen the sampling time? The average build up?
That is, now I'm not aware and think that I have to correct as slowly as
possible because I think that the oscillator has to be disturbed to a
minimum. Then I see low frequency large deviations, so I think, OK, I have
to average longer to account for. Is this the filtering you are referring
to? So that one ends up increasing the slowness of the system getting only
very slow frequency very large deviations.
Thanks for the help

On Mon, May 7, 2012 at 10:33 PM, Magnus Danielson <
Hi Azelio,
Post by Azelio Boriani
OK, got it. Yes, something like the dithering with a DAC to increase the
resolution.
Indeed. Now, consider now that the variations can come from any form of
noise source.
Another thing I've learned is that the longer you wait with a correction,
the larger the deviation becomes and lower frequency it will have... and
both makes it harder to suppress by filtering. It's really just the same
thing form another angle.
Cheers,
Magnus
_______________________________________________
To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Magnus Danielson
2012-05-07 21:44:19 UTC
Permalink
the larger the deviation becomes and lower frequency it will have... and
both makes it>harder to suppress by filtering.
Filtering at what level? Lengthen the sampling time? The average build up?
That is, now I'm not aware and think that I have to correct as slowly as
possible because I think that the oscillator has to be disturbed to a
minimum. Then I see low frequency large deviations, so I think, OK, I have
to average longer to account for. Is this the filtering you are referring
to? So that one ends up increasing the slowness of the system getting only
very slow frequency very large deviations.
Thanks for the help
If you have a little frequency error the longer you wait to do any
adjustment the larger phase-deviation that frequency error will result
in. If you sample to seldom, then you rely on your DAC resolution and
stability inbetween your samples, the clock will essentially be in
hold-over. If this is 1 ms, 1 s or 1000 s will make a difference.

That relates to sampling rate, which puts a limit to the loop bandwidth
you can have.

But my main reaction was to the sample-rate vs. measure and adjust rate
(i.e. sample rate), and I wanted to point out that there is a merit in
sampling (much) faster. The modulation waveform is only to illustrate
the averaging behaviour.

Cheers,
Magnus
Azelio Boriani
2012-05-07 22:17:06 UTC
Permalink
Well, so given the goal (stability at tau, for example) find the best
measure and adjust rates (maybe they are not the same) given the
oscillator-to-be-disciplined characteristics.

On Mon, May 7, 2012 at 11:44 PM, Magnus Danielson <
Post by Azelio Boriani
the larger the deviation becomes and lower frequency it will have... and
both makes it>harder to suppress by filtering.
Filtering at what level? Lengthen the sampling time? The average build up?
That is, now I'm not aware and think that I have to correct as slowly as
possible because I think that the oscillator has to be disturbed to a
minimum. Then I see low frequency large deviations, so I think, OK, I have
to average longer to account for. Is this the filtering you are referring
to? So that one ends up increasing the slowness of the system getting only
very slow frequency very large deviations.
Thanks for the help
If you have a little frequency error the longer you wait to do any
adjustment the larger phase-deviation that frequency error will result in.
If you sample to seldom, then you rely on your DAC resolution and stability
inbetween your samples, the clock will essentially be in hold-over. If this
is 1 ms, 1 s or 1000 s will make a difference.
That relates to sampling rate, which puts a limit to the loop bandwidth
you can have.
But my main reaction was to the sample-rate vs. measure and adjust rate
(i.e. sample rate), and I wanted to point out that there is a merit in
sampling (much) faster. The modulation waveform is only to illustrate the
averaging behaviour.
Cheers,
Magnus
_______________________________________________
To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
ws at Yahoo
2012-05-06 18:02:12 UTC
Permalink
Poul

I agree, 1PPS from many GPS engines, if sawtooth correction is not used, can
add all sorts of errors like hanging bridges.
Some can sound really bad, like non zero bias over short or very long time
periods and 12 ns of early or late wonder.
It's interesting to talk about these problems, But from a practical point
you do not need to fix them, because all these errors are small compared to
what the errors are from the received GPS signal including it's short term
wonder and long term near 12 and 24 hour cycles due to satellite orbits
cycles.

If you want something without the1 PPS problems, can use an optimized tuned
Tbolt.

ws

*****************
From: "Poul-Henning Kamp"
Post by Poul-Henning Kamp
Post by Hal Murray
I don't understand that. What am I missing?
You are missing that the average of the 1PPS pulse only can be trusted
to be zero over a timescale of many hours.
This is an error-source distinct from GPS reception, caused by the
picking a preexisting flank nearest to the epoch, with no attempt
to keep the average of the resulting error zero.
Imagine that the GPS receivers clock happens to run on perfect
frequency for a while: That means that the flank used to generate
the PPS will have a fixed location relative to the epoch, for instance
always 12 nanoseconds early or late.
I belive that some GPS receivers have deliberately de-tuned Xtals
for this very reason, but unfortunately that is only a partial
fix, as the problem is a modulus-issue, so not only is perfect
frequency bad, but "perfect +/- n Hz" is equally bad.
The hanging bridges Tom has plots of on leapsecond.com, arises when
the frequency of the GPS xtal changes.
At one point I tried putting a 1W resister close to the xtal and
feed it with a very slow sine-wave to see if "jittering" it would
get me an average of zero of shorter timespans. My experiment
was inconclusive, but the idea is not unsound.
Poul-Henning Kamp
Hal Murray
2012-06-26 18:16:05 UTC
Permalink
I'm very south of the equator on a family vacation right now, away from my
lab, remote enough to miss the upcoming leap second. But here's a photo of
a sundial I made with driftwood and shell markers every 5 minutes.
Speaking of sundials, here is a neat one:
http://apod.nasa.gov/apod/ap120626.html

(more good info and pictures at some of the links on that page)
--
These are my opinions. I hate spam.
Hal Murray
2012-08-08 04:32:45 UTC
Permalink
That is _exactly_ why you should use a vectorformat like SVG: Raster format
is a waste of bytes for line graphics.
Except that it doesn't work that way, at least for my simple test case.

SVG is uncompressed text. PNG compresses well, at least for simple cases.

PNG, 8K:
Loading Image...

SVG, 46K:
http://www.megapathdsl.net/~hmurray/time-nuts/coax/Front-5ns-800x600.svg

(It doesn't work right in my browser. It shows up as type text/plain rather
than image/svg. I don't know if that's my screwup or a bug in my ISP's web
server.)

It works OK if I download it and view it with Firefox from my disk.
--
These are my opinions. I hate spam.
gary
2012-08-08 04:42:00 UTC
Permalink
You can save the file then view it in inkscape.
http://inkscape.org/
That worked for me. Firefox is supposed to read SVG, but all I got was text.
Chris Albertson
2012-08-08 05:13:32 UTC
Permalink
Post by Murray Greenman
That is _exactly_ why you should use a vectorformat like SVG: Raster
format
is a waste of bytes for line graphics.
Except that it doesn't work that way, at least for my simple test case.
SVG is uncompressed text. PNG compresses well, at least for simple cases.
http://www.megapathdsl.net/~hmurray/time-nuts/coax/Front-5ns-800x600.png
http://www.megapathdsl.net/~hmurray/time-nuts/coax/Front-5ns-800x600.svg
So you are saying that SVG can't work because one example of it is broken.
Also, there are other vector formats, like Postscript and PDF.
--
Chris Albertson
Redondo Beach, California
Sylvain Munaut
2012-08-08 12:19:04 UTC
Permalink
Hi,
Post by Hal Murray
SVG is uncompressed text. PNG compresses well, at least for simple cases.
Decently configured web servers will compress SVG on the fly during
transport, wich yields a 9k transfer size.
(and your server is definitely not properly configured for SVG, it
doesn't compress and serves it as text/plain ...)

Cheers,

Sylvain
Jim Lux
2012-08-13 13:26:49 UTC
Permalink
Post by Sylvain Munaut
Hi,
Post by Hal Murray
SVG is uncompressed text. PNG compresses well, at least for simple cases.
Decently configured web servers will compress SVG on the fly during
transport, wich yields a 9k transfer size.
(and your server is definitely not properly configured for SVG, it
doesn't compress and serves it as text/plain ...)
"decently configured" probably excludes my application, where I have a
very limited function server (an Arduino)

Web services are a convenient, nearly universal, scheme to communicate
between boxes. It's not always in the context of a traditional "fat
server/thin client" sort of model.

In fact, I'd say that the client end (the web browser) typically has
more computing and display horsepower (on a per connection basis) than
the server in most cases. Even a fairly big server serving 100s of
pages per second to iPhone type clients has a more horsepower on the
receiving end than the sending end.
Hal Murray
2013-04-21 21:14:39 UTC
Permalink
The OCXO's Vref has 18 microVolt RMS noise
I used my HP3458A to measure the stability of the Vref output from an
Neat. Thanks.

What's the spectrum look like and what's the bandwidth of the filter inside
the OCXO? How much of that noise is going to cause trouble?
--
These are my opinions. I hate spam.
Hal Murray
2013-05-05 04:53:00 UTC
Permalink
Rule of thumb: quartz is best short term, Rb or H-maser mid-term, and Cs by
far the best long-term.
What is short, medium, and long?

Radio astronomers use H-masers. Can I assume that they are mid-term and that
H-masers are better than Rb (at mid-term)?

Does the classic ADEV graph contain all the information, or is it making an
assumption that is valid in most cases that allows it to compress/hide lots
of information that is interesting for only a few obscure types of
applications?
--
These are my opinions. I hate spam.
Bob Camp
2013-05-05 12:08:07 UTC
Permalink
Hi

All the data is in an adev plot. In this case short is < 100 seconds, and long is > 10,000 seconds. Those are rough numbers, since a really good Rb (like Corby's) may cross over a bit earlier. A really crummy Cs (low beam current) might not cross over for a couple of days against a well stabilized Rb or Maser. A good BVA OCXO will give the Rb a bit more of a run for it's money ….

The cross overs will happen. Where is going to depend entirely on the specific individual standards you happen to have. If you are making decisions about which of your boxes to use, you have to measure them.

Bob
Post by Hal Murray
Rule of thumb: quartz is best short term, Rb or H-maser mid-term, and Cs by
far the best long-term.
What is short, medium, and long?
Radio astronomers use H-masers. Can I assume that they are mid-term and that
H-masers are better than Rb (at mid-term)?
Does the classic ADEV graph contain all the information, or is it making an
assumption that is valid in most cases that allows it to compress/hide lots
of information that is interesting for only a few obscure types of
applications?
--
These are my opinions. I hate spam.
_______________________________________________
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
WarrenS
2013-05-05 16:40:21 UTC
Permalink
"All the data is in an adev plot... The cross overs will happen... you have
to measure them."
True, but then what do you do?
It is not quite as simple or easy as it may sound.

Although it is a good place to start,
for best results in a GPSDO you can not just compare the ADEV crossover
points of the two frequency sources.
The problem I've seen is that long term ADEV plots generally show a turn-up,
often around the 1000 sec range.
The turn-up in the plot, more often than not, is caused by systematic
errors, not random noise,
so the turn-up may be scaled incorrectly by the ADEV plot.

The things I've seen that can cause 'premature' turn-up on an ADEV plot are:
Not allowing enough time for the osc to stabilize after turn on,
room temperature variation, outliers and fixed rate ageing.
With careful attention to many details, the turn-up can often be
significantly reduced.
The effect that each of these errors types have on various disciplined
control loops varies greatly.

The problem is when the effect that each of these errors has on a
disciplined control
loop such as a GPSDO is not the same as the effect that they have on the
ADEV plot,
you can not just use the crossover point of the two plots.

The most extreme example is **fixed** ageing rate of the frequency source
that is to be disciplined.
A fixed ageing rate drift causes a slope of one turn-up on an ADEV plot
(but has little or no effect on a Hadamard plot).
On the other hand a fixed ageing rate error, which is often the major error
of a good DOCXO,
has no effect on frequency stability in a basic fixed time constant
disciplined control loop such as used in a TBolt.
It does cause a constant fixed phase error that is a function of the control
loop's time constant and damping settings,
but that can be removed completely if desired by just changing the control
loop's cable length setting.
On other types of disciplined control loops, the effect of a fixed ageing
rate error may vary and depends on the type of advanced control loop used.

The effect of temperature variation on a disciplined control loop is another
big variation that can effect ADEV plots and disciplined control loops
differently.
In the case of a TBolt, delta temperature correction is only applied when
the unit is in Holdover,
so its effect has to be considered when setting up the GPSDO.
This is why the best way to fix that error source is to not let the
temperature change or to use a external DOCXO.
Advanced control loops can greatly reduce the effect of changing temperature
with feed-forward control,
so they may not be nearly as sensitive to temperature variation.

ws

*******************
Hi
All the data is in an adev plot. In this case short is < 100 seconds, and
long is > 10,000 seconds. Those are rough numbers, since a really good Rb
(like Corby's) may cross over a bit earlier. A really crummy Cs (low beam
current) might not cross over for a couple of days against a well
stabilized
Rb or Maser. A good BVA OCXO will give the Rb a bit more of a run for it's
money ..
The cross overs will happen. Where is going to depend entirely on the
specific individual standards you happen to have. If you are making
decisions about which of your boxes to use, you have to measure them.
Bob
*************************************
Rule of thumb: quartz is best short term, Rb or H-maser mid-term, and Cs
by far the best long-term.
What is short, medium, and long?
Radio astronomers use H-masers. Can I assume that they are mid-term and
that H-masers are better than Rb (at mid-term)?
Does the classic ADEV graph contain all the information, or is it making
an assumption that is valid in most cases that allows it to compress/hide
lots of information that is interesting for only a few obscure types of
applications?
Hal Murray
2013-08-22 10:14:51 UTC
Permalink
Loading Image...
Neat. Thanks.

I remember that one from ages ago.

Does anybody have a URL for the original/official cartoon?
Or even just the name of that cartoon strip?
--
These are my opinions. I hate spam.
Poul-Henning Kamp
2013-08-22 10:19:59 UTC
Permalink
Post by Hal Murray
http://faculty.ucc.edu/business-greenbaum/images/NanosecZits.jpg
Does anybody have a URL for the original/official cartoon?
Or even just the name of that cartoon strip?
The strip is simply called "Zits" I'm surprised how anybody can
live on this planet and not know it :-)

http://zitscomics.com/

I don't belive their on-line archive stretches that far back...
--
Poul-Henning Kamp | UNIX since Zilog Zeus 3.20
phk-***@public.gmane.org | TCP/IP since RFC 956
FreeBSD committer | BSD since 4.3-tahoe
Never attribute to malice what can adequately be explained by incompetence.
Hal Murray
2014-03-24 06:29:21 UTC
Permalink
Does anybody have data on the drift of Rs or Cs?
...

Ah, thanks. But by "Rs and Cs" I meant plural or R/resistor and C/capacitor.
I know they change a lot with temperature, but how much do they drift if the
temperature is constant?
--
These are my opinions. I hate spam.
Bob Camp
2014-03-24 11:12:19 UTC
Permalink
Hi

Which R’s and C’s did you buy? There are a *lot* of different types and they each have their issues. A very common issue is - “they came from the store”. Any ability to trace them back to a manufacturer and a process has been lost.

Bob
Post by Hal Murray
Does anybody have data on the drift of Rs or Cs?
...
Ah, thanks. But by "Rs and Cs" I meant plural or R/resistor and C/capacitor.
I know they change a lot with temperature, but how much do they drift if the
temperature is constant?
--
These are my opinions. I hate spam.
_______________________________________________
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Hal Murray
2014-05-31 06:43:31 UTC
Permalink
Always locate your DUTs physically orthogonal to each other.
Unless you have 3 clocks. (and everybody knows what happens if you only have
2)
From an old time-nuts message (Mar, 2009)
Allied to this discussion is the Loomis effect, discovered by the
American millionaire who had three Shortt clocks running in his
basement. They synchronised unless aligned at 120 degrees to each
other. I wonder weather they were shaking the bedrock, or maybe the
gravitational attraction between the 10 kg pendulums may have
synchronised them. (See "Tuxedo Park" by Jennet Conant) He qualified
as the first time nut.
It's on page 67-68.

Google for >Tuxedo-Park Shortt< gets a hit in books.google.com at page 65
which is the start of the coverage of Shortt clocks. That was long before
eBay. :)
--
These are my opinions. I hate spam.
Hal Murray
2015-12-20 04:27:24 UTC
Permalink
The main reason Ethernet went balanced was actually for fault isolation
(star-topology vs. bus) and signal quality (IT people were horrible at
"sharking" and crimping coax.)
The reason Ethernet switched to a star topology was to take advantage of the
wires that were already installed in most buildings.

The electronics at each end were cheap compared to the cost of installing a
new wire in the ceiling.
--
These are my opinions. I hate spam.



_______________________________________________
time-nuts mailing list -- time-***@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Poul-Henning Kamp
2015-12-20 09:03:46 UTC
Permalink
--------
Post by Hal Murray
The main reason Ethernet went balanced was actually for fault isolation
(star-topology vs. bus) and signal quality (IT people were horrible at
"sharking" and crimping coax.)
The reason Ethernet switched to a star topology was to take advantage of the
wires that were already installed in most buildings.
That came later.

The first twisted-pair stuff only allowed 30m of wire and was meant for
connecting workstations in an office to a hub, so that only the hub
had to attach to the "backbone" coax.
--
Poul-Henning Kamp | UNIX since Zilog Zeus 3.20
***@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.
_______________________________________________
time-nuts mailing list -- time-***@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Hal Murray
2016-03-26 21:22:01 UTC
Permalink
The one I have has both EFC+ and EFC- pins.
Is that a differential input so you don't have to worry about ground offset
shift when the oven current changes?
--
These are my opinions. I hate spam.



_______________________________________________
time-nuts mailing list -- time-***@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Poul-Henning Kamp
2016-03-26 23:00:01 UTC
Permalink
--------
Post by Hal Murray
The one I have has both EFC+ and EFC- pins.
Is that a differential input so you don't have to worry about ground offset
shift when the oven current changes?
I have no idea...

It's one I bought cheap and it's not particularly good, so I have not
spent a lot of time on it.
--
Poul-Henning Kamp | UNIX since Zilog Zeus 3.20
***@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.
_______________________________________________
time-nuts mailing list -- time-***@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Bob Camp
2016-03-26 23:19:18 UTC
Permalink
Hi

It’s probably bringing out both sides of the varicap to accomplish the same sort of thing.

Bob
Post by Hal Murray
The one I have has both EFC+ and EFC- pins.
Is that a differential input so you don't have to worry about ground offset
shift when the oven current changes?
--
These are my opinions. I hate spam.
_______________________________________________
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
_______________________________________________
time-nuts mailing list -- time-***@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Hal Murray
2016-11-21 23:16:04 UTC
Permalink
I think the installation manual for Trimbles timing products say you can use
either 75 or 50 Ohm cable...
I think they suggest using RG-6, the classic cable TV and/or satellite dish
cable. It's widely available at low cost.

The loss due to impedance mismatch is much less than the gain from the
reduced attenuation when compared to equivalent size 50 ohm cable.
--
These are my opinions. I hate spam.



_______________________________________________
time-nuts mailing list -- time-***@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Hal Murray
2018-06-24 21:05:52 UTC
Permalink
That's one of the reason I went with Bo's DDS (32bit0 instead of the eBay
modules (20-24 bit). Higher resolution mitigates the spur problem at least a
little bit.
I think it also moves the spurs closer in. But maybe if they are small enough
they get lost in the normal noise.

Does anybody have a handy formula for the spurs given the parameters for a DDS?

I've occasionally thought of writing code to generate the output of a DDS and
run it through a FFT. I haven't figured out how much memory that would need,
or rather how wide a DDS I could simulate with the memory I have. A 20 bit
accumulator repeats after a million cycles. At 8 bytes/sample that's 8
megabytes which I can do. (Round down if FFT needs another copy.)

We are only interested in the close-in area, so old brute-force calculations
maybe fast enough.
--
These are my opinions. I hate spam.



_______________________________________________
time-nuts mailing list -- time-***@lists.febo.com
To unsubscribe, go to https://lists.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
jimlux
2018-06-25 12:38:14 UTC
Permalink
Post by Hal Murray
That's one of the reason I went with Bo's DDS (32bit0 instead of the eBay
modules (20-24 bit). Higher resolution mitigates the spur problem at least a
little bit.
I think it also moves the spurs closer in. But maybe if they are small enough
they get lost in the normal noise.
Does anybody have a handy formula for the spurs given the parameters for a DDS?
There isn't one.. <grin>

The spur size and location is more determined by "how many loops through
the cosine table til you wind up back at zero" which gets down to the
phase quantization - or the cosine table size

Think of a cosine table that's 16 elements long (i.e. a 4 bit phase).

If the output frequency is fclk*8/16, your DDS puts out entry 0 and
entry 8, and repeats
If the output frequency is fclk*2/16, your DDS puts out entry
0,2,4,6,8,10,12,14,0,2,4,6

If the output frequency is fclk*3/16, your DDS puts out 0,
3,6,9,12,15,2,5,... and it takes a few cycles to repeat around

If the output frequency is fclk*2.5/16, your DDS puts out
0,2,5,7,10,12,15,1,4,6,9,11,14,0,3, etc.

these all have very different spur patterns.
Post by Hal Murray
I've occasionally thought of writing code to generate the output of a DDS and
run it through a FFT. I haven't figured out how much memory that would need,
or rather how wide a DDS I could simulate with the memory I have. A 20 bit
accumulator repeats after a million cycles. At 8 bytes/sample that's 8
megabytes which I can do. (Round down if FFT needs another copy.)
That is, in fact how people do it.
Post by Hal Murray
We are only interested in the close-in area, so old brute-force calculations
maybe fast enough.
_______________________________________________
time-nuts mailing list -- time-***@lists.febo.com
To unsubscribe, go to https://lists.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Dana Whitlow
2018-06-25 13:05:49 UTC
Permalink
A useful hardware demonstration of DDS phase truncation is to do step a DDS
through a long
series of frequencies at a relatively low stepping rate, and view the
output with an FFT-based
(real-time) spectrum analyzer set for spectrogram display. Watch this for
a while and it becomes
quite easy to reach a practical understanding of what's going on and how
the spurs behave.
Note that to avoid confusion from aliasing products, it is a good idea to
constrain the range of
frequencies swept to the first Nyquist zone (zero to < 1/2 the DDS clock
frequency).

There's a fairly simple regularity in the spur behavior which is not
particularly obvious from
looking at equations and tables (at least, not at my math skill level), but
which jumps out at
you in the spectrogram display (or even without the spectrogram).

Dana
Post by jimlux
Post by Hal Murray
That's one of the reason I went with Bo's DDS (32bit0 instead of the eBay
modules (20-24 bit). Higher resolution mitigates the spur problem at least a
little bit.
I think it also moves the spurs closer in. But maybe if they are small enough
they get lost in the normal noise.
Does anybody have a handy formula for the spurs given the parameters for a DDS?
There isn't one.. <grin>
The spur size and location is more determined by "how many loops through
the cosine table til you wind up back at zero" which gets down to the phase
quantization - or the cosine table size
Think of a cosine table that's 16 elements long (i.e. a 4 bit phase).
If the output frequency is fclk*8/16, your DDS puts out entry 0 and entry
8, and repeats
If the output frequency is fclk*2/16, your DDS puts out entry
0,2,4,6,8,10,12,14,0,2,4,6
If the output frequency is fclk*3/16, your DDS puts out 0,
3,6,9,12,15,2,5,... and it takes a few cycles to repeat around
If the output frequency is fclk*2.5/16, your DDS puts out
0,2,5,7,10,12,15,1,4,6,9,11,14,0,3, etc.
these all have very different spur patterns.
Post by Hal Murray
I've occasionally thought of writing code to generate the output of a DDS and
run it through a FFT. I haven't figured out how much memory that would need,
or rather how wide a DDS I could simulate with the memory I have. A 20 bit
accumulator repeats after a million cycles. At 8 bytes/sample that's 8
megabytes which I can do. (Round down if FFT needs another copy.)
That is, in fact how people do it.
Post by Hal Murray
We are only interested in the close-in area, so old brute-force calculations
maybe fast enough.
_______________________________________________
To unsubscribe, go to https://lists.febo.com/cgi-bin
/mailman/listinfo/time-nuts
and follow the instructions there.
_______________________________________________
time-nuts mailing list -- time-***@lists.febo.com
To unsubscribe, go to https://lists.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
jimlux
2018-06-25 14:22:08 UTC
Permalink
Post by Dana Whitlow
A useful hardware demonstration of DDS phase truncation is to do step a DDS
through a long
series of frequencies at a relatively low stepping rate, and view the
output with an FFT-based
(real-time) spectrum analyzer set for spectrogram display. Watch this for
a while and it becomes
quite easy to reach a practical understanding of what's going on and how
the spurs behave.
Note that to avoid confusion from aliasing products, it is a good idea to
constrain the range of
frequencies swept to the first Nyquist zone (zero to < 1/2 the DDS clock
frequency).
There's a fairly simple regularity in the spur behavior which is not
particularly obvious from
looking at equations and tables (at least, not at my math skill level), but
which jumps out at
you in the spectrogram display (or even without the spectrogram).
Here you go, as an example:

http://jim-lux.blogspot.com/2018/06/dds-spurs.html

16 bit DDS, 5 bit phase, 4096 samples for FFT spectra, etc.


_______________________________________________
time-nuts mailing list -- time-***@lists.febo.com
To unsubscribe, go to https://lists.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
c***@juno.com
2018-06-25 15:29:35 UTC
Permalink
Hi,

My purpose was to provide an inexpensive (<$30.00) and simple way to
replace a defective A1 without degrading the performance the original A1
gave.
The Chinese DDS allows this without having to solder all those tiny pins!
Both styles actually performed better than the original, and the same in
a "super" mod unit.
I was not claiming "theoretical" perfect performance and rejected more
complex schemes on purpose.
Also the current drawn by the A1 is part of the current budget for the
TED so an A1 replacements current draw should be between 100 and 150 mA!

Cheers,

Corby

_______________________________________________
time-nuts mailing list -- time-***@lists.febo.com
To unsubscribe, go to https://lists.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Attila Kinali
2018-06-25 16:42:30 UTC
Permalink
On Mon, 25 Jun 2018 08:29:35 -0700
Post by c***@juno.com
The Chinese DDS allows this without having to solder all those tiny pins!
Both styles actually performed better than the original, and the same in
a "super" mod unit.
BTW: what does limit the performance of the super-5065?
(short term, ie <100s)

Apparently, from the description of modifications I've seen,
the standard-5065 is mostly limited by detection noise, either
due to noise in the light, or the RF chain.

Could it be that the super-5065 and the A1 modifications have
shifted the limit to the detector/correlator and integrator?
Or does the Dick effect already limit you?

Attila Kinali
--
It is upon moral qualities that a society is ultimately founded. All
the prosperity and technological sophistication in the world is of no
use without that foundation.
-- Miss Matheson, The Diamond Age, Neil Stephenson
_______________________________________________
time-nuts mailing list -- time-***@lists.febo.com
To unsubscribe, go to https://lists.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
t***@timeok.it
2018-07-01 08:59:51 UTC
Permalink
Hi Corby,

I can not send you emails in your inbox. can you verify please?
Sorry to broadcast to all this service email.
Luciano


Da "time-nuts" time-nuts-***@lists.febo.com
A time-***@lists.febo.com
Cc
Data Mon, 25 Jun 2018 08:29:35 -0700
Oggetto [time-nuts] HP 5065A A1 replacement with DDS
Hi,

My purpose was to provide an inexpensive (<$30.00) and simple way to
replace a defective A1 without degrading the performance the original A1
gave.
The Chinese DDS allows this without having to solder all those tiny pins!
Both actually performed better than the original, and the same in
a "super" mod unit.
I was not claiming "theoretical" perfect performance and rejected more
complex schemes purpose.
Also the current drawn by the A1 is part of the current budget for the
TED so an A1 replacements current draw should be between 100 and 150 mA!

Cheers,

Corby

_______________________________________________
time-nuts mailing list -- time-***@lists.febo.com
To unsubscribe, go to https://lists.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
_______________________________________________
time-nuts mailing list -- time-***@lists.febo.com
To unsubscribe, go to https://lists.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Loading...