Re: [time-nuts] WWVB / Xtendwave patents

[Joe Gwinn]

The US changed over to the rest-of-world patent system, where patent 
applications are public for a period before grant, precisely to get 
input from the entire technical community.


The granted patent (8270465) would be hard to overturn at this point.

As for the patent application (2012/0082008), one can file comments 
against a patent application pointing out prior art and suggesting 
that something being claimed is in fact obvious to those skilled in 
the art.  For prior art, one points out existing patents and 
published articles.  Simple assertions are not sufficient.  The 
Patent Office will then consider all prior art received when deciding 
to grant or to deny a patent.


Joe Gwinn

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

Re: [time-nuts] wwvb weak on east coast especially when the pre-amps under water.

[Joe Gwinn]

At 10:49 AM + 5/15/12, time-nuts-requ...@febo.com wrote:

Date: Tue, 15 May 2012 08:59:04 +0200
From: Attila Kinali 
To: Discussion of precise time and frequency measurement

Subject: Re: [time-nuts] wwvb weak on east coast especially when the
pre-amps under water.
Message-ID: <20120515085904.19669d7edd8b95454f176...@kinali.ch>
Content-Type: text/plain; charset=US-ASCII

On Mon, 14 May 2012 18:01:01 -0400
Joseph M Gwinn  wrote:


 Modern outdoor enclosures use a filter of some kind, but the underlying
 principle is the same.


I don't know what other types are around, but we use vents with
a gore-tex foil over them. Keeps water out but lets the case breath.
This prevents any pressure build up, which would then start to suck
water in from the seal.


Gore-tex blocks liquid water (and dust), but allows water vapor to 
pass, so one can still get pumping and the accumulation of condensed 
water.  Where are your enclosures used, and how exposed are they?



As for the other methods, so far the following have been discussed:

Box with long tube, where the tube volume exceeds the tidal volume of 
the enclosure, so outside air never manages to get to the enclosure. 
Typically, the tube is plugged with some cotton wool, to keep insects 
out.


Box with short tube and filter/dessicant.  The Gore-tex film is a 
filter, but not a dessicant.  A box with medium tube and filter plus 
dessicant at the box end can be very effective, the medium tube 
reducing the rate at which the dessicant is exhausted.


Box with short tube that opens on conditioned space.

Totally hermetic box.  Very effective, but very hard to do in 
practice, unless the box is small and strong.


Box where positive pressure is maintained using dry air, so all leaks 
are outward and the dew point is never reached inside the box.


Joe Gwinn

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

Re: [time-nuts] Neutrinos not so fast? (defective connector)

[Joe Gwinn]

At 3:24 PM + 2/24/12, time-nuts-requ...@febo.com wrote:

Date: Fri, 24 Feb 2012 14:42:36 +0100
From: b...@lysator.liu.se
To: "Discussion of precise time and frequency measurement"

Subject: Re: [time-nuts] Neutrinos not so fast? (defective connector)


Bj?rn,

[snip]

 >>

 High precision GPS receivers use various correlator schemes that try to
 minimize multipath. "Normal" GPS receivers are more vulnerable than
 geodetic quality receivers.

 http://webone.novatel.ca/assets/Documents/Papers/PAC.pdf


 You are of course correct, but timing receivers may not go to such lengths
 are are needed for geodetic receivers.  A lot of the magic of geodetic
 receivers is in the choke-ring antenna, which ignores signals arriving
 from too low an angle above the horizon.

 In the Neutrino case, the multipath is built into the cable between
 antenna
 and receiver, so the antenna cannot help.  But I will read the article,
 which looks interesting.  Wonder if it would solve such a triple-transit
 echo problem.

 Joe


I am not discussing antenna multipath attenuation. That is a separate
topic. Look at figure 12 in the Novatel paper - noting for the y-axis -
that 1 C/A chip length is about 300 meters.


Hmm.  Isn't 60 nanoseconds (18 meters at C) well within this envelope?



Here is an article from the Ashtech/JPS/JNS/Topcon-family.

 http://tinyurl.com/7w9wl4p

Overview results are seen in figure 5.


Thanks.  I'll read it.



Timing receivers used by time-labs to compare Cs are usually geodetic
receivers with the option to lock the internal clock to external
10MHz/1PPS signals coming from the Cs.

My point is that good receivers attenuate multipath fairly well.


If it's old, it may well have none of these improvements.  The timing 
receivers I have used usually say that they are within 100 
nanoseconds; I assume this to be the three sigma range.  I do know 
that in stationary receivers, the Allan Deviation of a given such 
receiver is quite good, implying that the error is a slowly drifting 
offset.  A cable multipath could well cause a bias error.




Is it known which GPS receiver type was used in the Neutrino experiment?


Another poster (Antonio I8IOV) answered this:

"It was a  Septentrio PolaRx2e. See the paper on setup and procedures:

"

I don't know anything about that make and model.   But someone will. 
I wonder when this unit was made.  And if it has a Rubidium local 
oscillator.



Actually, I bet there are multiple Time Nuts with more than one GPS 
timing receiver.  It is a simple experiment to feed a pair from a 
single antenna plus splitter, and put an impedance bump or two in one 
feed line, and compare the 1PPS outputs.  (I don't have the 
equipment.)  Or, split, delay one path, recombine, using a variable 
attenuatior in the delay path to adjust the overall multipath effect 
with which to challenge one of the two receivers receiver.


Joe

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

Re: [time-nuts] Thermal insulation choice?

[Joe Gwinn]

At 10:15 PM + 1/6/12, time-nuts-requ...@febo.com wrote:

Date: Fri, 06 Jan 2012 14:41:13 -0600
From: David 
To: Discussion of precise time and frequency measurement

Subject: Re: [time-nuts] Thermal insulation choice?
Message-ID: 
Content-Type: text/plain; charset=us-ascii

I completely forgot about those two.  [Balsa wood and foam-in=place 
urethane foam]


I have a good local hobby shop
with large pieces of balsa wood but I suspect it would be more
expensive than good quality expanded polystyrene bead sheet.  I will
have to check out foam board next time.  Both would be more difficult
to cut without a hot wire knife.


Balsa wood cuts very easily with a sharp knife, glues very well with 
carpenters' glue, and will not degrade over time.


Joe Gwinn

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

Re: [time-nuts] Thermal insulation choice?

[Joe Gwinn]

I wonder if Balsa wood would be suitable?  Like polystyrene foam, 
only stronger and easy to glue.


I think commercial OXCOs use polystyrene foam that's expanded to 
shape, which is fine in production, but making the mold for one use 
is not efficient.  However, there are lots of expand-in-place 
urethane foams available from building supply houses.


Joe Gwinn

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

Re: [time-nuts] crunching numbers from XOR phase detector

[Joe Gwinn]

At 10:16 AM + 1/4/12, time-nuts-requ...@febo.com wrote:

Date: Wed, 04 Jan 2012 00:58:24 -0800
From: Hal Murray 
To: Discussion of precise time and frequency measurement

Subject: Re: [time-nuts] crunching numbers from XOR phase detector
Message-ID:
<20120104085824.8426e800...@ip-64-139-1-69.sjc.megapath.net>
Content-Type: text/plain; charset=us-ascii


Neat.  Thanks for sharing.

With a XOR, you can't tell which input is higher frequency.  I think you can
fix that with a second XOR and a delay line.

I think 90 degrees of delay will provide the most information.  At 10 MHz,
that's 25 ns.  I think that's about 15 feet of good coax.


Beware of teflon dielectric cable, as the teflon knee is centered 
around room temperature, making delay cables of teflon quite 
sensitive to slight changes in temperature.


<http://www.micro-coax.com/pages/technicalinfo/applications/27.asp>

Joe Gwinn

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

Re: [time-nuts] Low noise power supplies for time nuts circuitry

[Joe Gwinn]

At 2:28 PM + 11/24/11, time-nuts-requ...@febo.com wrote:

Message: 3
Date: Thu, 24 Nov 2011 13:22:38 +0100
From: Attila Kinali 
To: Discussion of precise time and frequency measurement

Subject: Re: [time-nuts] Low noise power supplies for time nuts
circuitry
Message-ID: <2024132238.0c810b78.att...@kinali.ch>
Content-Type: text/plain; charset=US-ASCII

On Wed, 23 Nov 2011 04:50:49 -0800
Chuck Forsberg WA7KGX N2469R  wrote:


 How well would a pair of cascaded 3 terminal regulators do - say a 7815
 feeding a 7812.


78xx Regulators are quite noisy. You can use them to filter the
noise of a cheap DC/DC converter, but i wouldnt use them feed
high precision electronics. There are a lot better designs these days.


What I see done in low-noise circuits is a low-noise opamp used as a 
linear voltage regulator to clean up the output of a 78xx or the like 
regulator.


Joe Gwinn

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

Re: [time-nuts] Fast than light neutrino

[Joe Gwinn]

At 6:56 AM + 9/26/11, time-nuts-requ...@febo.com wrote: (really Javier S)

Date: Mon, 26 Sep 2011 08:56:12 +0200
From: Javier Serrano 
To: Discussion of precise time and frequency measurement

Subject: Re: [time-nuts] Fast than light neutrino
Message-ID:

Content-Type: text/plain; charset=ISO-8859-1

[snip]
 >



I did not express myself correctly. We know how to do accurate two-way sync
over a few km of fiber. See e.g.
http://dl.dropbox.com/u/13409775/pac2011/WEOAN1.pdf or
http://dl.dropbox.com/u/13409775/wrapper.pdf (our contribution to the CLIC
Conceptual Design Report).


Thanks for the reports.



What is new for us is going through more than
1000 km of fiber (only neutrinos have the luxury of going in a straight line
through the crust of the Earth, 732 km). I wonder who one calls for fibers
and also more technical things like optical amplifier technology with
typical ranges, etc. What I gather from the discussion so far is that 100 km
is within reach of available optical transceivers. I wonder how far one can
go with EDFAs.


EDFAs can easily achieve 10 dB or running closer to the edge 20 dB of 
gain, all with no electronics delay, but EDFAs are quite noisy, so 
there is a tradeoff to be made.


EDFAs are inherently bidirectional, although they usually contain an 
optical circulator to make them unilateral.  But it would not be hard 
to make EDFAs that amplified in one direction for one wavelength, and 
in the opposite direction at a different (but nearby) wavelength.


And one can also have a command wavelength, to allow for commanding 
of direction reversals and the like, so the fiber companies 
involvement is limited to hosting and installation of equipment.




One thing we could do is establish a fiber link between METAS
in Bern and CERN, and then look for a good metrology place in Rome (the
national one in Italy is in Torino I believe) and have a link between them
and Gran Sasso. Then we could use their UTC data sets to establish a paper
link between CERN and Gran Sasso which would be independent of the current
link.


Why do all that, versus just running an amplified fiber between CERN 
and Gran Sasso?  Two links are likely to be twice the trouble and 
error.



Joe Gwinn


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

Re: [time-nuts] Fast than light neutrino

[Joe Gwinn]

Date: Fri, 23 Sep 2011 14:51:26 +1000
From: Jim Palfreyman 
To: Discussion of precise time and frequency measurement

Subject: [time-nuts] Fast than light neutrino
Message-ID:

Content-Type: text/plain; charset=ISO-8859-1

For those of you who may be interested, here's the paper.

http://arxiv.org/ftp/arxiv/papers/1109/1109.4897.pdf


The path is through 730 kilometers of solid rock, so only neutrinos 
will do.  And neutrinos are the least understood of particles.


If I read the article correctly, the neutrinos appear to travel 25 
parts per million faster than C, which if true is still 
revolutionary.  But while the result is quite significant in 
statistical terms (6 sigma), 25 ppm is pretty small, and could easily 
be caused by some subtle systemic error.


One assumes *very* subtle, given that none of the ~100 coauthors 
could find it, and it won't have been for lack of trying.  Now the 
world physics community is on the case, so it may not take all that 
long.


Joe Gwinn

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

Re: [time-nuts] live 50 Hz measurements

[Joe Gwinn]

From digest vol 84 issue 7:

Date: Fri, 1 Jul 2011 13:06:31 -0700
From: Chris Albertson 
To: xfor...@citynet.net, Discussion of precise time and frequency
measurement 
Subject: Re: [time-nuts] live 50 Hz measurements
Message-ID: 
Content-Type: text/plain; charset=ISO-8859-1

On Fri, Jul 1, 2011 at 12:46 PM, Will Matney  wrote:
 >
  For accuracy sake, I'd still compare the line freq. to a 
standard, but that's just me.


If the computer used for this is also running "ntpd" the computer's
clock is disciplined.  How well depends on what ntpd is using as a
reference clock.


Actually, there *are* cases where ntpd is running without complaint 
but the local clock is not being disciplined.


The situation I most often run into is that ntpd has been started 
with insufficient privilege, and so its tweak-the-clock requests are 
being silently ignored by the operating system.


If there is no drift file, the usual symptom is that the clock offset 
grows without bound almost linearly.


If there is a valid drift file, the usual symptom is that the clock 
offset sways more or less sinusoidally, this being the residue after 
the average clock drift is removed.



Joe Gwinn.

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

Re: [time-nuts] Choke Ring Pictures

[Joe Gwinn]

Date: Sun, 14 Mar 2010 21:09:00 +0100
From: Magnus Danielson 
Subject: Re: [time-nuts] Choke Ring Pictures
To: Discussion of precise time and frequency measurement

Message-ID: <4b9d425c.2050...@rubidium.dyndns.org>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed

Bob Camp wrote:

 Hi

 Wedding cake pans normally come in 1" increments and are either 2" 
or 3" deep. Sets are 2" increments on the diameter:


 http://cooksdream.com/store/wedding-round.html


http://www.hubert.com/store/products.asp?CAWELAID=126235277&A=SB%2E58369%2E10738&Dn=0&An=966+966&Au=Presentation+Id&Ntt=10738&N=966+966&src=chanadv&Ntx=mode+matchall&D=10738&Ntk=SKU

 The height would be fairly easy to adjust. The diameter not so much so.

 Looks like 2 and 2.5" are typical dimensions for the depth. ~1" 
looks pretty typical for the width. A 2" deep / 2" diameter step 
set looks like it would do a pretty good job . It won't be accurate 
enough to be perfect. Without a 3D EM program it would be tough to 
figure out just what the errors would do to you.


With 2.5" depth and 14", 12", 10", 8" and 6" diameter pans you are not
completely in a different world from some antennas:

http://facility.unavco.org/project_support/permanent/equipment/antennas/ant_cals.html

Evaluating the performance may be a different thing.

One thing to care about is the leakage between the pans if you just 
stick them inside each other.


I'd be tempted to use EMI gaskets between the pans, except that 
contact will soon be lost as the aluminum grows a nice oxide layer. 
But it may not be necessary to make DC electrical contact, as the 
capacitance between the metal layers may suffice to act as a short at 
1.5 GHz.  A thin sheet of mylar (or a heavy anodization layer) 
between would help this along, by increasing the capacitance.


Joe Gwinn

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

Re: [time-nuts] Rack-mounting an LPRO? (Heatpipe cooler)

[Joe Gwinn]

At 2:02 AM + 2/28/10, time-nuts-requ...@febo.com wrote:

Date: Sat, 27 Feb 2010 20:02:05 -0600
From: Ed Palmer 
Subject: Re: [time-nuts] Rack-mounting an LPRO?
To: Discussion of precise time and frequency measurement

Message-ID: <4b89ce9d.4060...@sasktel.net>
Content-Type: text/plain; charset="iso-8859-1"; Format="flowed"

Sorry, the pictures got lost.  Let's try again.

Ed Palmer wrote:

 What else is going to be in the rack?  If your 1U enclosure is packed
 in tight between other devices there might be no cooling at all.  You
 might need a fan to move some air.

 I don't know if you can find something like this, but I scavenged heat
 sinks from an old Compaq DL760 server that might fit your situation. 
 Here's what the heat sink looks like.  The aluminum plate is about

 1/4" (6.4 mm) thick.  Notice the heat pipes.



 Here's what it looks like on the LPRO.  You'll have to drill holes in
 the plate to match the LPRO.



 It's not perfect, but it certainly does the job - particularly if you
 had a fan blowing through the fins.  The total height is about 1.75"
 (45 mm).

 But remember, the more you cool the LPRO, the more power it will draw
 to keep itself warm so you don't want to overdo the cooling.


 > Ed



[snip]
-- next part --
A non-text attachment was scrubbed...
Name: 1.jpg
Type: image/jpeg
Size: 37469 bytes
Desc: not available
URL: 
<http://www.febo.com/pipermail/time-nuts/attachments/20100227/9f21a5a8/attachment.jpg>

-- next part --
A non-text attachment was scrubbed...
Name: 2.jpg
Type: image/jpeg
Size: 33025 bytes
Desc: not available
URL: 
<http://www.febo.com/pipermail/time-nuts/attachments/20100227/9f21a5a8/attachment-0001.jpg>


The aluminum fin and copper pipe assembly integrated with the 
heatsink plate is most likely a heat pipe of some kind, as that's 
what Thermacore makes.  It's a model 2644, from the nameplate, but no 
joy at the Thermacore.com website.


Joe Gwinn

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

Re: [time-nuts] Variable Conductance Heat Pipes for temperature control

[Joe Gwinn]

At 7:31 PM + 1/23/10, time-nuts-requ...@febo.com wrote:

Date: Sat, 23 Jan 2010 14:31:40 -0500
From: Bob Camp 
Subject: Re: [time-nuts] Variable Conductance Heat Pipes for
temperature control
To: Discussion of precise time and frequency measurement

Message-ID: <08528fe8-4eec-4a8b-af68-fe045d227...@cq.nu>
Content-Type: text/plain; charset=us-ascii

Hi

After a quick read, I still don't see much on just how good the 
thermal control can get. Of course I could have missed that page ...


The performance of the Rb oscillator VCHP heatsink is in the book I 
left at work.  It was pretty good, especially considering that the 
heatsink system is wholly passive and electronics-free.


Joe



Bob


On Jan 23, 2010, at 1:33 PM, Joe Gwinn wrote:

 > Back in December 2009 I proposed (in "Re: [time-nuts] Cheap 
Rubidium (heatpipe cooling for)") use of heat pipes to stabilize the 
temperature of a Rubidium oscillator within an insulated box. 
Turns out I was anticipated by ~40 years.


 While standard heat pipes offer near isothermality , they don't 
offer temperature constancy as the heat flow and/or condenser 
temperature vary.  However, there are hints in the literature that 
near constant temperature can be achieved, complete with pictures 
of a Thermacore International product used to control the 
temperature of a Rb oscillator used in a cell phone tower or the 
like.

 >

[snip]

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

[time-nuts] Variable Conductance Heat Pipes for temperature control

[Joe Gwinn]

Back in December 2009 I proposed (in "Re: [time-nuts] Cheap Rubidium 
(heatpipe cooling for)") use of heat pipes to stabilize the 
temperature of a Rubidium oscillator within an insulated box.   Turns 
out I was anticipated by ~40 years.


While standard heat pipes offer near isothermality , they don't offer 
temperature constancy as the heat flow and/or condenser temperature 
vary.  However, there are hints in the literature that near constant 
temperature can be achieved, complete with pictures of a Thermacore 
International product used to control the temperature of a Rb 
oscillator used in a cell phone tower or the like.


General description from Thermacore: 
<http://www.thermacore.com/products/variable-conductance-heat-pipe.aspx> 



So, I did some digging.  It turns out that the technology was worked 
out circa 1970, under the rubric Variable Conductance Heat Pipes 
(VHCPs), which are described in all standard books on heat pipes, but 
usually only in passing, unclearly.  However, the older books tended 
to be better on the subject.


The basic principle of a VHCP is simple.  One builds a standard 
heatpipe, but in addition to the usual working fluid (that carries 
heat by evaporation and condensation) one puts some non-condensible 
gas inside as well.  A classic example would be a heatpipe made of 
copper and containing water plus nitrogen.


In operation, the flow of water vapor from evaporator to condenser 
sweeps the nitrogen into the condenser, which becomes partially 
blocked by the resulting nitrogen-gas plug.  If the heat flow 
increases, the pressure inside the pipe increases, compressing the 
nitrogen plug, unblocking more of the condenser area, thus reducing 
the temperature change at the evaporator.


Anyway, the modern textbook on heatpipes is "Heat Pipes - Theory, 
Design, and Application" fifth edition, David Reay and Peter Kew, 
Butterworth-Heinemann (Elsevier) 2006.  (The Thermacore product for 
Rb oscillators mentioned above is shown in Figure 7.2 on page 277.)


The references in Reay and Kew lead one to "Heat pipe theory and 
practice - a sourcebook", S.W. Chi, Hemisphere Publishing 
(McGraw-Hill), 1976.  (This has an entire chapter on "controlled heat 
pipes", including a nice description of how to design a VHCP, all 
based on work done at TRW for the space program.


The key reference in Chi is to the report written by TRW, "Theory and 
design of variable conductance heat pipes", B.D. Marcus, NASA 
Contractor Report CR-2018, April 1972.  A pdf scan of this 252-page 
report is available for free download from NASA'a Technical Report 
Server: 
<http://ntrs.nasa.gov/search.jsp?R=190521&id=9&as=false&or=false&qs=Ns%3DPublicationYear%257c0%26N%3D4294915469>


Copper-water-nitrogen VCHPs seem simple enough for fabrication in a 
home shop, although long-term retention of nitrogen (or any such gas) 
will require hard brazed or welded hermetic seams and seals (versus 
soft solder and ball valves).


Joe Gwinn

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

Re: [time-nuts] Cheap Rubidium (heatpipe cooling for)

[Joe Gwinn]

At 12:00 PM + 12/27/09, time-nuts-requ...@febo.com wrote:


Date: Sat, 26 Dec 2009 17:04:46 -0700
From: Robert Darlington 
Subject: Re: [time-nuts] Cheap Rubidium (heatpipe cooling for)
To: Discussion of precise time and frequency measurement


My comments are in-line, below

On Fri, Dec 25, 2009 at 4:38 PM, Joe Gwinn  wrote:


 At 12:45 AM + 12/25/09, time-nuts-requ...@febo.com wrote:



 Date: Thu, 24 Dec 2009 17:14:38 -0700
 From: Robert Darlington 
 Subject: Re: [time-nuts] Cheap Rubidium (heatpipe cooling for)
 To: Discussion of precise time and frequency measurement



 On Thu, Dec 24, 2009 at 1:32 PM, Bob Camp  wrote:

   Hi


  A heat pipe might work if the fluid had a sufficiently low boiling
 point.




 The working fluid in a heat pipe will boil at every temperature above its
 melting point.



 Well, I've been thinking about this, and I used the term "heat pipe" too
 loosely.  Both the one- and two-pipe systems mentioned here have no wicks,
 and so technically are two-phase thermosyphons, which depend on gravity to
 circulate vapor and condensate.  A true heat pipe has a wick, and will work

 > in zero gravity.


 One gets significant heat transfer by phase change so long as the vapor
 pressure in the heat input end is high enough to generate enough vapor to
 carry the thermal power flow, and this makes the pipe isothermal.  However
 the temperature (although constant along the pipe) varies with the thermal
 power flow (in thermal watts) being carried.

 What I'm looking for is related but different:  A device where the heat
 transfer capacity varies sharply with temperature, so that there is a range
 of heat transfer rates over which the input-end temperature will be
 substantially constant.  This is why I envision the fluid boiling (versus
 evaporating), which is actually out of the operating regime of a true heat
 pipe.


 >

 I tend to use water because it's cheap, but have made them

 >> with 3M "engineered fluids", Fluorinert, and denatured alcohol.
 >

 Fluorinert.  I think that's what the expensive commercial CPU-cooling
 heatpipes use.


$1000 a gallon!  Or $5 a drum when you buy it at a salvage auction.


That explains why low-end heatpipes use alcohol or acetone.

Actually, one ought to be able to use the freon intended for 
automobile air conditioners, for a whole lot less money, even new.




 >>  I've found

 that ordinary solder works just fine.  A trick to make these things easy
 to build is to use a ball valve at the top (I'm assuming there is a top and
 we're going with gravity return because it's simple).  I've got a few that
 are still under vacuum for several years now in this configuration.  My
 giant heat pipe of doom is a 10 foot stick of 1/2" copper with a ball valve
 at one end and an end cap at the other.  There is perhaps 100ml water in
 there total, and no air.  You can either boil the liquid until it builds up
 a nice head of steam, or go the easy way and pull a vacuum with a pump and

 >> just close the valve.
 >

 I wouldn't have thought that an ordinary ball valve would be tight enough,
 allowing the water to escape and the air enter, slowly, although I suppose

 > one can replace the water if it comes to that.
 >
Mine have been running for a few years with no sign of needing to be pumped
down again.  They just work.

 > But I think people want to build this exactly once, so I followed

 refrigeration practice.  A properly made hermetically sealed refrigeration
 system keeps its working fluid essentially forever.  I suppose one can use a
 refrigeration fill valve, say from an automobile air conditioning system,
 but these all leak to some degree.

 Is the ball valve anything special?

 >
Nope, just whatever was on the shelf at the local hardware store.
Stainless ball with brass valve body.  Teflon bearing surface.


Ahh.  A quarter-turn ball valve, used as a cutoff.  The term "ball 
valve" isn't quite precise in plumbing parlance.


These are very good, but still they are not hermetic, and will over 
decades (if not a few years) lose their working fluid.  I bet that 
while water will be contained, freon will diffuse right through the 
teflon seal of the ball valve.  So, there's the tradeoff.




 >>  These things are incredible.  If you pack snow around

 the end of this thing, the other end that is ten feet away gets cold almost
 immediately.  They want to stay isothermal and the heat transfer is at the
 speed of sound through the working fluid.  Delays are introduced because
 you're dealing with a thermal mass of copper pipe that needs to change

 >> temperature along with the working fluid so it's not quite instant, but

 still about 10,000 times faster heat transfer than copper by itself.  They
 are certainly handy for getting heat out of confined spaces.

 >>



 &

Re: [time-nuts] Cheap Rubidium (heatpipe cooling for)

[Joe Gwinn]

At 12:45 AM + 12/25/09, time-nuts-requ...@febo.com wrote:


Date: Thu, 24 Dec 2009 17:14:38 -0700
From: Robert Darlington 
Subject: Re: [time-nuts] Cheap Rubidium (heatpipe cooling for)
To: Discussion of precise time and frequency measurement


On Thu, Dec 24, 2009 at 1:32 PM, Bob Camp  wrote:


 Hi

 A heat pipe might work if the fluid had a sufficiently low boiling point.



The working fluid in a heat pipe will boil at every temperature above its
melting point.


Well, I've been thinking about this, and I used the term "heat pipe" 
too loosely.  Both the one- and two-pipe systems mentioned here have 
no wicks, and so technically are two-phase thermosyphons, which 
depend on gravity to circulate vapor and condensate.  A true heat 
pipe has a wick, and will work in zero gravity.


One gets significant heat transfer by phase change so long as the 
vapor pressure in the heat input end is high enough to generate 
enough vapor to carry the thermal power flow, and this makes the pipe 
isothermal.  However the temperature (although constant along the 
pipe) varies with the thermal power flow (in thermal watts) being 
carried.


What I'm looking for is related but different:  A device where the 
heat transfer capacity varies sharply with temperature, so that there 
is a range of heat transfer rates over which the input-end 
temperature will be substantially constant.  This is why I envision 
the fluid boiling (versus evaporating), which is actually out of the 
operating regime of a true heat pipe.




I tend to use water because it's cheap, but have made them
with 3M "engineered fluids", Fluorinert, and denatured alcohol.


Fluorinert.  I think that's what the expensive commercial CPU-cooling 
heatpipes use.




I've found
that ordinary solder works just fine.  A trick to make these things easy to
build is to use a ball valve at the top (I'm assuming there is a top and
we're going with gravity return because it's simple).  I've got a few that
are still under vacuum for several years now in this configuration.  My
giant heat pipe of doom is a 10 foot stick of 1/2" copper with a ball valve
at one end and an end cap at the other.  There is perhaps 100ml water in
there total, and no air.  You can either boil the liquid until it builds up
a nice head of steam, or go the easy way and pull a vacuum with a pump and
just close the valve.


I wouldn't have thought that an ordinary ball valve would be tight 
enough, allowing the water to escape and the air enter, slowly, 
although I suppose one can replace the water if it comes to that.


But I think people want to build this exactly once, so I followed 
refrigeration practice.  A properly made hermetically sealed 
refrigeration system keeps its working fluid essentially forever.  I 
suppose one can use a refrigeration fill valve, say from an 
automobile air conditioning system, but these all leak to some degree.


Is the ball valve anything special?



These things are incredible.  If you pack snow around
the end of this thing, the other end that is ten feet away gets cold almost
immediately.  They want to stay isothermal and the heat transfer is at the
speed of sound through the working fluid.  Delays are introduced because
you're dealing with a thermal mass of copper pipe that needs to change
temperature along with the working fluid so it's not quite instant, but
still about 10,000 times faster heat transfer than copper by itself.  They
are certainly handy for getting heat out of confined spaces.


What is the purpose of the heatpipe of doom?  Education?




-Bob




 The rubidium isn't terribly tolerant of high temperatures, and I'm going to
 pick up some heat rise as I put it inside some baffles / shields. You need
 to find something that fits a fairly narrow window.

 I suspect that a recirculating water loop is a more practical approach to
 carry away the heat. It's got a pump to move the water, but the rest of it

 > is fairly simple.


 Bob


 On Dec 24, 2009, at 2:49 PM, Joe Gwinn wrote:

 > A dodge occurs to me - a homebrew heat pipe: <
 http://en.wikipedia.org/wiki/Heat_pipe>.
 >
 > Make the cold plate of copper, to which is soldered a meandering piece of
 copper tubing, which tubing is also soldered to a copper radiator plate that
 is above the coldplate, forming a closed loop with a fill tube attached by a
 T.  Braze all tubing connections, as for freon refrigeration systems.  (Soft
 solder is too porous to work for the joints, but is OK for attaching tubes

 > to plates.)
[snip]  [This is really a kind of "thermosyphon", as discussed above.]
 > >

 > Anyway, a heat pipe system will stabilize the coldplate temperature
 fairly accurately despite variations in thermal load, has no moving or
 electrical parts, and may be sufficient by itself.  If not sufficient, it
 can be used as the outer stage in a two-stag

Re: [time-nuts] Cheap Rubidium (heatpipe cooling for)

[Joe Gwinn]

A dodge occurs to me - a homebrew heat pipe: 
<http://en.wikipedia.org/wiki/Heat_pipe>.


Make the cold plate of copper, to which is soldered a meandering 
piece of copper tubing, which tubing is also soldered to a copper 
radiator plate that is above the coldplate, forming a closed loop 
with a fill tube attached by a T.  Braze all tubing connections, as 
for freon refrigeration systems.  (Soft solder is too porous to work 
for the joints, but is OK for attaching tubes to plates.)


Insulate the two tubes running between coldplate and radiator plate 
from one another.


Put enough working fluid into the system to fill the tubing that is 
soldered to the coldplate, but no more.  Warm the system up so the 
vapor drives all the air out, pinch the fill tube off and fold it 
back, and braze the end shut.   (It's not critical to get absolutely 
all the air out.)


Making the radiator plate be above the coldplate (the boiler) 
implements what amounts to an oldtime two-pipe water vapor heating 
plant.  Vapor goes up one pipe, condensed fluid returns via the 
other.  I lived in a house with such a system.  The difference 
between a vapor plant and a steam plant is pressure:  the vapor plant 
runs below atmospheric pressure, while the steam plant runs at or 
slightly above.


Make sure that things are arranged so the returning fluid does not 
pool anywhere but in the coldplate, or the heat pipe will bang like 
an old steam heating system.


There is a brazing filler metal intended for copper-to-copper joints 
that is widely used for freon systems: 
<http://www.uniweld.com/catalog/alloys/silver_brazing_alloys/phos_copper.htm>. 
The zero silver phos stuff is adequate, cheap and widely available. 
While copper-to-copper needs no flux, copper-to-brass does, so also 
get the flux.  Plumbing supply houses and welding equipment stores 
are likely sources.  You will also need a torch or pair of torches 
able to raise the tubing joints to an orange heat in a reasonable 
length of time.


Depending on the chosen working fluid, the cold plate temperature 
will not rise above the boiling point of the fluid unless the system 
is too small (in radiator heat removal capacity) to easily handle the 
10 or 20 thermal watts that are passing through.


What fluid to use?  Anything common and thermally stable that does 
not attack copper.  Alcohol (methyl or ethyl) and water are common 
choices, as are the various freons.  I bet acetone would also work. 
Anyway, one controls the coldplate temperature by a combination of 
choice of working fluid and internal pressure.



I have seen commercially made heat pipes for cooling Intel CPUs 
advertised, but I don't know that these units can be adapted.


Anyway, a heat pipe system will stabilize the coldplate temperature 
fairly accurately despite variations in thermal load, has no moving 
or electrical parts, and may be sufficient by itself.  If not 
sufficient, it can be used as the outer stage in a two-stage ovening 
scheme.



Joe Gwinn

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

Re: [time-nuts] Cheap Rubidium

[Joe Gwinn]

At 10:06 PM + 12/23/09, time-nuts-requ...@febo.com wrote:

Date: Thu, 24 Dec 2009 09:46:13 +1300
From: Bruce Griffiths 
Subject: Re: [time-nuts] Cheap Rubidium
To: Discussion of precise time and frequency measurement


Joe Gwinn wrote:

 Date: Thu, 24 Dec 2009 08:57:42 +1300
 From: Bruce Griffiths 
 Subject: Re: [time-nuts] Cheap Rubidium
 To: Discussion of precise time and frequency measurement

 >> 




[snip]


 >> Distributed heating using wire wound or printed heaters perhaps, but to

 reduce the associated magnetic field bifilar winding should be
 considered.


 Non-inductive power resistors, which are commercially available, have
 very low magnetic fields.

 The low-inductance resistors have Ayrton-Perry windings, which are
 bifilar.


No, Ayrton-Perry windings arent bifilar.

Classically a flattened helical winding was made on a insulating card.
An identical winding was then wound in the opposite direction on top of
the first winding and the 2 were connected in parallel.
The idea being that the small magnetic field produced by one flattened
helix is cancelled by that of the other flattened helix.


True enough - while there are two conductors, they are not close and parallel.

Anyway, the point is that non-inductive components by definition have 
low magnetic fields, and that non-inductive power resistors are 
common.


To eliminate the field from the loop of resistors, one can have a 
linear string of A-P resistors in series, with a pair of return wires 
in parallel, with the return wires on either side of the resistor 
string, thus reducing the effective loop area.


Joe


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

Re: [time-nuts] Cheap Rubidium

[Joe Gwinn]

Date: Thu, 24 Dec 2009 08:57:42 +1300
From: Bruce Griffiths 
Subject: Re: [time-nuts] Cheap Rubidium
To: Discussion of precise time and frequency measurement


Bob Camp wrote:

 Hi

 So if I want to set up 4 uncorrelated systems, that would require 20 tons of
 water split into 4 tubs. Each tub would be roughly 3' x 4' x 15'. Of course
 > if they are all in the same basement, I still have a correlation 
problem. My

 guess is that no matter what I do, any system that controls all the systems
 the same way will run into correlation.

 Oils, silicon fluids, and the like mostly hold less heat than water so the
 tubs would get bigger. Maybe a few tons of mercury...
  

Try about 145 tons of mercury per rubidium source as the specific heat
of mercury is about 1/29 that of water.
The redeeeming feature is that it will only occupy about 2.14x the volume.
The specific of some oils may be as large as 1/2 that of water however
the density is around 10-20% lower.


 Active heat control and a rational heat sink is sounding like a better
 approach...

  

Distributed heating using wire wound or printed heaters perhaps, but to
reduce the associated magnetic field bifilar winding should be considered.


Non-inductive power resistors, which are commercially available, have 
very low magnetic fields.


The low-inductance resistors have Ayrton-Perry windings, which are bifilar.

<http://www.token.com.tw/resistor-pd/power-resistor-ah.htm>



The major limitation is that the 25W or so dissipated by the rubidium
source has to be transferred to ambient without raising the rubidium
temperature too much.
This limits the maximum thermal resistance between the baseplate and
ambient that can be safely used.


I would be tempted to regulate temperature by actively controlling 
the speed of the fan (or pump) driving air (or oil) through the heat 
sink, as has been suggested.


Joe Gwinn

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

Re: [time-nuts] 60Hz mains clocking in computers

[Joe Gwinn]

At 4:53 PM + 12/13/09, time-nuts-requ...@febo.com wrote:

Date: Sun, 13 Dec 2009 17:42:12 +0100
From: Magnus Danielson 
Subject: Re: [time-nuts] 60Hz mains clocking in computers
To: Discussion of precise time and frequency measurement

Message-ID: <4b251964.5040...@rubidium.dyndns.org>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed

Joe Gwinn wrote:

 At 1:44 AM + 12/13/09, time-nuts-requ...@febo.com wrote:

 Date: Sat, 12 Dec 2009 17:29:17 -0800
 From: Colby Gutierrez-Kraybill 
 Subject: [time-nuts] 60Hz mains clocking in computers
 To: time-nuts@febo.com
 Message-ID: <3058527a-cc99-4174-be75-21dd92334...@astro.berkeley.edu>
 Content-Type: text/plain; charset=US-ASCII; format=flowed; delsp=yes


 I'm trying to get to the bottom of whether or not any computing
 equipment made around the advent of UNIX systems (or any time-slicing
 system) used the mains cycles of 60Hz as phase lock for the internal
 system clock.  My guess is that perhaps they did not as the computing
 logic is DC based, but, I have memories of using an 68000 based UNIX
 system that I thought had its internal clock based off of the 60Hz
 mains...  Not sure the vendor anymore.

 >

 In the 1980s and 1990s, before networks capable of carrying NTP time to
 the millions became common, the computer local clock was very often
 derived from the local AC power mains, and the frequency was steered to
 match atomic time once per day.  The POSIX standards reflect this common
 approach by the tolerance on CLOCK_REALTIME, 20 milliseconds, this being
 one cycle of 50 Hz power.


Which does not perfectly match the 60 Hz being used in some countries or
for that matter traditional division for PC clocks (derived from
14,31818 MHz, over 4,77 MHz and what the 8253 divider allows).


 The CPU logic clock was not generally phase-locked to the AC power
 lines, instead being generated by a cheap crystal having a very large
 tempco.  The exception to this was that video generators were (and still
 are) often locked to the AC line so that hum bars would not drift across
 the screen.


I have never seen this in any of the devices I've seen. It is certainly
not what we do in the TV world either. Examples would be good.


It was certainly true in black and white TVs, and I'm pretty sure 
that the early Macintosh computers did the same.


NTSC color TV runs almost at 60 Hz, so the hum bars drift slowly 
enough to not be visually disturbing.


The Television Handbook will have the relevant standards for TV. 
Because NTSC was intended for North America, behavior with 50 Hz 
power was not a concern.  European standards are different, but they 
must have dealt with hum bars somehow.


When all the vacuum tubes (except the picture tube) disappeared from 
TVs, hum bars became far less of a problem, and subsequent TV 
standards don't worry about the effect.



Joe Gwinn


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

Re: [time-nuts] 60Hz mains clocking in computers

[Joe Gwinn]

At 1:44 AM + 12/13/09, time-nuts-requ...@febo.com wrote:

Date: Sat, 12 Dec 2009 17:29:17 -0800
From: Colby Gutierrez-Kraybill 
Subject: [time-nuts] 60Hz mains clocking in computers
To: time-nuts@febo.com
Message-ID: <3058527a-cc99-4174-be75-21dd92334...@astro.berkeley.edu>
Content-Type: text/plain; charset=US-ASCII; format=flowed; delsp=yes


I'm trying to get to the bottom of whether or not any computing 
equipment made around the advent of UNIX systems (or any time-slicing 
system) used the mains cycles of 60Hz as phase lock for the internal 
system clock.  My guess is that perhaps they did not as the computing 
logic is DC based, but, I have memories of using an 68000 based UNIX 
system that I thought had its internal clock based off of the 60Hz

mains...  Not sure the vendor anymore.


In the 1980s and 1990s, before networks capable of carrying NTP time 
to the millions became common, the computer local clock was very 
often derived from the local AC power mains, and the frequency was 
steered to match atomic time once per day.  The POSIX standards 
reflect this common approach by the tolerance on CLOCK_REALTIME, 20 
milliseconds, this being one cycle of 50 Hz power.


The CPU logic clock was not generally phase-locked to the AC power 
lines, instead being generated by a cheap crystal having a very large 
tempco.  The exception to this was that video generators were (and 
still are) often locked to the AC line so that hum bars would not 
drift across the screen.


Joe Gwinn

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

Re: [time-nuts] Measuring phase shift between 1 Hz DMTD signals by I+Q processing

[Joe Gwinn]

Bruce,

At 1:00 AM + 7/26/09, time-nuts-requ...@febo.com wrote:

Message: 3
Date: Sun, 26 Jul 2009 12:29:24 +1200
From: Bruce Griffiths 
Subject: Re: [time-nuts] Measuring phase shift between 1 Hz DMTD
signals by I+Q processing
To: Discussion of precise time and frequency measurement



If one uses a mixer output frequency of several kHz then one can avoid
the flicker noise region if one uses a high pass filter between the ADCs
and the mixer preamps.

Does such a system have a performance advantage over direct RF sampling?
Perhaps it does if and only if the phase noise floor of the lower
bandwidth ADCs that are used is lower than the noise floor of the ADCs
that would be required to sample the RF signals directly?
The noise floor of state of the art ADCs suitable for direct RF sampling
is around -150dBFS/Hz.
The noise floor of  "typical" high resolution ADC(AD7762, AD7641)
capable of sampling at around 1MSPS or so appear to be similar.


It strikes me that one can do a double conversion here, with one 
conversion done in analog hardware, the other in a DSP.


The analog mixer would go from megahertz to say 100 KHz or 30 KHz, 
and this kilohertz signal would be digitized, yielding a stream of 
I+Q samples.


The resulting stream of digital I+Q samples would then be numerically 
mixed down to 1 Hz, and the relative phase between the two 1 Hz 
signals would be measured.


This two-step approach should neatly sidestep the flicker-noise issue.

As for the ADC yielding I+Q samples, given the great oversampling 
possible with current ADCs, one can use a single ADC and 
mathematically generate the I+Q streams.  There are many methods, 
invented because high=performance ADCs are very expensive, and 
because it is difficult to find sufficiently well matched pairs of 
such ADCs.



Joe Gwinn

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

Re: [time-nuts] Measuring phase shift between 1 Hz DMTD signals by I+Q processing

[Joe Gwinn]

Magnus,

At 1:00 AM + 7/26/09, time-nuts-requ...@febo.com wrote:

Message: 4
Date: Sun, 26 Jul 2009 03:00:28 +0200
From: Magnus Danielson 
Subject: Re: [time-nuts] Measuring phase shift between 1 Hz DMTD
signals by I+Q processing
To: Discussion of precise time and frequency measurement


Joe Gwinn wrote:

 Magnus,

 At 4:01 PM + 7/25/09, time-nuts-requ...@febo.com wrote:


 Message: 5
 Date: Sat, 25 Jul 2009 16:38:23 +0200
 From: Magnus Danielson 
 Subject: Re: [time-nuts] Measuring phase shift between 1 Hz DMTD
 signals by I+Q processing
 To: Discussion of precise time and frequency measurement
 

 Joe Gwinn wrote:

  It occurs to me that there is a possible alternative to the ZCD-chain
  approach typical in DMTDs, if one is willing to provide two mixers and
  two ADCs per channel, with a 90 degree phase offset between LO signals
  provided to the mixers of a channel.  The output of the four ADCs will
  be a pair of I+Q signals, one pair per DMTD channel.

  The key observation is that if one has two signals, one being a time
  delayed replica of the other, if one multiplies one signal by the

 >>  > complex [conjugate] of the other signal, the result is Exp[j(phase

  > difference)].  This is true whatever the waveform of the signal, so long

 >>>  as the only difference in signals is a delay.  The mathematical
 >>>  argument function of this exponential is the desired phase.


  In practice, one will sample far faster than 1 Hz, say 1 MHz, and will
  heavily average the resulting stream of products.

  Now I have not gone through the math to estimate performance
 compared to
  the traditional ZCD approach, but the complex multiply and average
  approach should be quite robust against noise, and is easily
 implemented in a DSP or FPGA.


 The time-difference between the two sampling points could be minimized
 in such an approach as the phase could be shifted arbitrarily in the
 post-processing such that the effective phase difference between the two
 chains reduces to near zero and hence the correlation between the
 channels for the transfer oscillator would be better in phase and cancel
 the transfer oscillator out better.


 It would be nice, but I need to think about this.  I'm not sure that you
 don't have to use a real physical delay out in the analog hardware.


If you play the vector/phasor game you know that while you shift
frequency, you don't shift phase, so whatever phase-change you want to
apply to the carrier level (10 MHz) you can apply to the mixed down case.


I haven't had time to play the math gain, but I suspect that you are right.



Recall, the problem with not full cancelation of the transfer oscillator
is due to the time-difference of the beat notes and that the ZCD
detectors by design adapts to what it percieves to be 0 degree and that
the the time occurence of this is different between the two beat note
channels.

Now, if we phase shift at the beat note frequency we can make that
channels beat-note time occurance in the ZCD to be come arbitrarilly
shifted and thus close the time-difference considerably until they occur
too close in which case we get unwanted degradation due to cross-talk.

Phase-shifting at the carrier frequency is just to achieve the same
thing, infact that is doing it one step away from where it is expected
to occur. If we do this in stable digital domain, there is less
stability issues.

It should be noted that such shifting needs to be continously monitored
and adjusted as the carrier frequencies drift appart. Any adjustments
needs to be compensated or otherwise phase-steps will be introduced into
the datastream. A carefull correction actually compensate for both gain
and phase shifts.


I would set up a tracking loop to keep the two 1 Hz signals 
coincident, and the loop implementation would report how much shift 
war required to achieve this coincidence.




 >> The postprocessing would then slowly tune the I/Q phase and keep a phase

 adjustment track such that post-correlation could turn it back for
 proper phase-trace.


 But, unlike ZCD-triggered counters, there is no disadvantage or
 difficulty if the phase difference is adjusted exactly to zero, where
 the two 1 Hz sinewaves coincide.


Depends on your post-processing. If you attempt to emulate ZCD in
firmware, then you get that result. If you rather do the
phase-subtraction processing no phase-shifting is needed as it is done
sample-for-sample and the issue is gone and you have a clear
phase-difference record that builds.


Yes.



 >> An alternative approach is to use the Costas tracking loop as Bruce

 suggested.


 A Costas loop is far more complex, but they do work well.  Given near
 constant phase delay, don't know if a Costas loop is worth the trouble.

 The Costas loop will not by itself solve the problem of
 transfer-oscillator noise.


Costas loops isn't that expensive these days, rathe

Re: [time-nuts] Measuring phase shift between 1 Hz DMTD signals by I+Q processing

[Joe Gwinn]

Magnus,

At 4:01 PM + 7/25/09, time-nuts-requ...@febo.com wrote:


Message: 5
Date: Sat, 25 Jul 2009 16:38:23 +0200
From: Magnus Danielson 
Subject: Re: [time-nuts] Measuring phase shift between 1 Hz DMTD
signals by I+Q processing
To: Discussion of precise time and frequency measurement


Joe Gwinn wrote:

 It occurs to me that there is a possible alternative to the ZCD-chain
 approach typical in DMTDs, if one is willing to provide two mixers and
 two ADCs per channel, with a 90 degree phase offset between LO signals
 provided to the mixers of a channel.  The output of the four ADCs will
 be a pair of I+Q signals, one pair per DMTD channel.

 The key observation is that if one has two signals, one being a time
 delayed replica of the other, if one multiplies one signal by the

 > complex [conjugate] of the other signal, the result is Exp[j(phase
 > difference)].  This is true whatever the waveform of the signal, so long

 as the only difference in signals is a delay.  The mathematical argument
 function of this exponential is the desired phase.

 In practice, one will sample far faster than 1 Hz, say 1 MHz, and will
 heavily average the resulting stream of products.

 Now I have not gone through the math to estimate performance compared to
 the traditional ZCD approach, but the complex multiply and average
 approach should be quite robust against noise, and is easily implemented
 in a DSP or FPGA.


The time-difference between the two sampling points could be minimized
in such an approach as the phase could be shifted arbitrarily in the
post-processing such that the effective phase difference between the two
chains reduces to near zero and hence the correlation between the
channels for the transfer oscillator would be better in phase and cancel
the transfer oscillator out better.


It would be nice, but I need to think about this.  I'm not sure that 
you don't have to use a real physical delay out in the analog 
hardware.




The postprocessing would then slowly tune the I/Q phase and keep a phase
adjustment track such that post-correlation could turn it back for
proper phase-trace.


But, unlike ZCD-triggered counters, there is no disadvantage or 
difficulty if the phase difference is adjusted exactly to zero, where 
the two 1 Hz sinewaves coincide.




An alternative approach is to use the Costas tracking loop as Bruce
suggested.


A Costas loop is far more complex, but they do work well.  Given near 
constant phase delay, don't know if a Costas loop is worth the 
trouble.


The Costas loop will not by itself solve the problem of 
transfer-oscillator noise.




Regardless this first stage of digital processing can be done in a FPGA
frontend and bring the resulting signal bandwidth into very reasonable
rates, just as for a GPS receiver.


Yes.  Is 0.01 Hz slow enough?

Joe Gwinn

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

[time-nuts] Measuring phase shift between 1 Hz DMTD signals by I+Q processing

[Joe Gwinn]

It occurs to me that there is a possible alternative to the ZCD-chain 
approach typical in DMTDs, if one is willing to provide two mixers 
and two ADCs per channel, with a 90 degree phase offset between LO 
signals provided to the mixers of a channel.  The output of the four 
ADCs will be a pair of I+Q signals, one pair per DMTD channel.


The key observation is that if one has two signals, one being a time 
delayed replica of the other, if one multiplies one signal by the 
complex complement of the other signal, the result is Exp[j(phase 
difference)].  This is true whatever the waveform of the signal, so 
long as the only difference in signals is a delay.  The mathematical 
argument function of this exponential is the desired phase.


In practice, one will sample far faster than 1 Hz, say 1 MHz, and 
will heavily average the resulting stream of products.


Now I have not gone through the math to estimate performance compared 
to the traditional ZCD approach, but the complex multiply and average 
approach should be quite robust against noise, and is easily 
implemented in a DSP or FPGA.


Joe Gwinn

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

Re: [time-nuts] Broken Ovenaire OSC 85-50 (Cutting ferrite cores)

[Joe Gwinn]

At 1:13 PM + 7/3/09, time-nuts-requ...@febo.com wrote:

[snip]
-Original Message-
From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On
Behalf Of Bruce Griffiths
Sent: Thursday, May 28, 2009 7:13 PM
To: Discussion of precise time and frequency measurement
Subject: Re: [time-nuts] Broken Ovenaire OSC 85-50


Joe [Trantham]

Whilst its easy enough to construct a miniature RF current probe using a
small ferrite core which is slipped over the wire in which the current is to
be measured, constructing one using a split ferrite toroid is more difficult
as such toroids are difficult to come by. In principle one can construct
ones own by cutting and grinding a pair of toroids. Unless one has access to
diamond lapping equipment and diamond saws this is probably impractical.


Actually, at least in the US it's easy to get such blades from 
lapidary shops for less than US $10.  Here's one:


<http://www.dadsrockshop.com/blades.html>

The traditional approach is to hold the oddly shaped object to be cut 
in pitch.  By lapidary standards, ferrites are pretty soft, and so 
should cut quite quickly.


Joe Gwinn

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

[time-nuts] PCI IRIG receiver card for AIX?

[Joe Gwinn]

Does anybody know of any PCI cards that will receive IRIG-B time 
signals and come with an I/O driver for AIX (IBM's flavor of UNIX)? 
Industrial-grade commercial products are preferred.


Thanks,

Joe Gwinn

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

Re: [time-nuts] Standards sought for immunity of shielded cable links to power-frequency ground loops

[Joe Gwinn]

Magnus,

At 6:02 PM + 1/10/09, time-nuts-requ...@febo.com wrote:
>
>Message: 4
>Date: Sat, 10 Jan 2009 19:02:09 +0100
>From: Magnus Danielson 
>Subject: Re: [time-nuts] Standards sought for immunity of shielded
>   cable links to power-frequency ground loops
>To: Discussion of precise time and frequency measurement
>   
>Joe,
>
   For digital signals (1PPS, various triggers), it's RS422 over 100 ohm
   twinax (fancy shielded twisted pair).

   The 10 MHz sinewave is sent over a pair of 50 ohm coax links, with the
   signals 180 degrees out of phase.  This is acheived with a pair of hybrid
>>>   > transformers which convert from one-cable to two-cable and then back to
   one-cable, where all cables are 50 ohm coax.
>  >>
>>>  OUCH! The trouble with that arrangement is that the coax cables MUST be
>>>  twisted or else H-fields will induce differential mode current. It will
>>>  induce current into both directions which through the 180 degree will
>>>  not cancel but add up. The 0/180 degree arrangement will save you from
>>>  common mode problems. You would prefer a twisted cable over a twisted
>>>  cable pair, as the later allows for installation procedure errors to
>  >> have huge impact and the twisting properties will not be as good either
>>>  and thus compromising the quality. A single ended coax is not as
>>>  sensitive to H fields to induce diffrential currents, but can have some
>>>  other problems.
>>
>>  You are right about the twisting.  The cables are close and parallel,
>>  and ground offsets are the big problem, versus magnetic fields.
>
>I just want you to end up having that trouble instead. I think you
>should consider a shielded twisted pair instead. Use the transformer to
>go between 50 Ohm and 100-110 Ohm while also getting the common mode
>isolation. A double-transformer approach can be used in which the
>launch/receive-transformer has a center tap on the "inside" which is
>wired to local ground (needs to be very low impedance). This improves
>capacitive isolation for common mode currents. The inner transformers do
>impedance matching. This is really an alternative to getting isolation
>transformers, it might even be cheaper. Dual-shielded isolation
>transformers is better thought, as capacitive coupling as spread out
>over the coil is always terminated to each side own shield which reduces
>common-mode to diffrential mode conversion.

The engineer wanted to use catalog components, which means 
connectorized hybrid transformers, probably from Minicircuits or the 
like.   He did use real twinax elsewhere, and the hum pickup issue 
has occurred to him.

The connectors are Type N, and the cable will be some kind of robust 
double-shielded flexible type.  He may already be twisting the two 
cables, which are about 30 meters long.


>  > My worry was that the ground currents might be enough to saturate the
>>  tiny ferrite cores in the hybrid transformers.  The engineer's
>>  reaction to this was on the following day to say that if this turns
>>  out to be a problem, he will add DC blocks.  This would have to be
>>  the kind that blocks both center and shield paths.
>
>I have a bit hard [time] to realize how the common mode ground current would
>saturate the hybrid transformers unless the current is so high that the
>asymmetry in the transformers helps. Some form of DC blocker or LF
>current limiting may be wise thought.

The 60 Hz limit in MIL-STD-461 CS109 is one amp (120 dB over one 
microamp).  The EMI guy said that this limit was arrived at for 
submarines in the 1970s, and the currents were primarily due to 
charging currents from capacitor-input power supplies, and the like.

What saves us with the hybrids is that while the cores are small, the 
windings might have five turns, so it will take a very substantial 
current to have any effect, and the winding will blow out first.


>  > The problem is that the radar and the ship are not yet built, so we
>  > cannot yet make tests.
>
>So much better. You have a chance to get things right before it is too
>late and too expensive.

Yes and no.  The drawings are done long before, and change is 
painful.  But necessary.


>I am sure we can send a sub to sink it late if needed.

I'm not sure that solves the problem, but it certainly eliminates the problem.


>    energy straight through and allow for a higher resistive path for the
>   low frequency energy.
   The ground grid impedance between any two points is well less than one
   ohm, so 100 ohms will pretty much abolish all ground loops.  I've used 10
   ohms in like labs, with success.  I'll grant that this would 
not work with
   long wires outside.
>>>  Should be sufficient then. But remember that capacitive coupling helps
>>>  you in the RF area and impulse protection.
>>
>>  True.
>
>The reason I keep mentioning it is since that it is easy to focus and
>make a design "optimum" for one case and forgetting about other aspects

Re: [time-nuts] Standards sought for immunity of shielded cable links to power-frequency ground loops

[Joe Gwinn]

Magnus,

At 10:31 AM + 1/10/09, time-nuts-requ...@febo.com wrote:
>
>Message: 5
>Date: Sat, 10 Jan 2009 11:06:39 +0100
>From: Magnus Danielson 
>Subject: Re: [time-nuts] Standards sought for immunity of shielded
>   cable   links to power-frequency ground loops
>To: Discussion of precise time and frequency measurement
>   
>
>Joseph,
>
>>  time-nuts-boun...@febo.com wrote on 01/07/2009 10:47:46 PM:
>>
>>>  Joseph,
>>>
>>  Could be a differential TX and RX.  I recall that they send a RS422
  signal.
>  Depending on the speed, RS422 works fine with transformers.
  Yes.  It would be 10 MHz or 20 MHz, depending on coding.  Or 5 MHz, so
>>  the
  transitions are at 10 MHz.  I don't recall, or never knew.
>>>  RS422 does not imply any encoding as such so it would be 10 MHz but
>>>  naturally there is twice that many transitions, but it is the frequency
>>>  of the signal you are interested in for this case.
>>
>>  I know that RS422 is not the encoding.  I cheated, and talked to the
>>  relevant engineer.
>
>That is to cheat! :)
>
>>  For digital signals (1PPS, various triggers), it's RS422 over 100 ohm
>>  twinax (fancy shielded twisted pair).
>>
>>  The 10 MHz sinewave is sent over a pair of 50 ohm coax links, with the
>>  signals 180 degrees out of phase.  This is acheived with a pair of hybrid
>  > transformers which convert from one-cable to two-cable and then back to
>>  one-cable, where all cables are 50 ohm coax.
>
>OUCH! The trouble with that arrangement is that the coax cables MUST be
>twisted or else H-fields will induce differential mode current. It will
>induce current into both directions which through the 180 degree will
>not cancel but add up. The 0/180 degree arrangement will save you from
>common mode problems. You would prefer a twisted cable over a twisted
>cable pair, as the later allows for installation procedure errors to
>have huge impact and the twisting properties will not be as good either
>and thus compromising the quality. A single ended coax is not as
>sensitive to H fields to induce diffrential currents, but can have some
>other problems.

You are right about the twisting.  The cables are close and parallel, 
and ground offsets are the big problem, versus magnetic fields.

My worry was that the ground currents might be enough to saturate the 
tiny ferrite cores in the hybrid transformers.  The engineer's 
reaction to this was on the following day to say that if this turns 
out to be a problem, he will add DC blocks.  This would have to be 
the kind that blocks both center and shield paths.

The problem is that the radar and the ship are not yet built, so we 
cannot yet make tests.


>   But you should never let the screen float in the far end, you should
>  terminate it with a 10M resistor and a sparkgap in parallel to the
>  local ground.
>
>  The resistor takes care of static electricity and the sparkgap will
>  do lightnings.
  I've done such things, but with a 100 ohm resistor (and a safety
>>  ground to
  ensure that the voltage doesn't get too large.  But this was
>>>  a lab lashup.
>>>
>>>  The trouble with 100 ohm is that still can be a little low in relation
>  >> to ground loop impedances, it still allow some fair current to roll down
>  >> the cable. A capacitor in parallel would cut most of the transient
>>>  energy straight through and allow for a higher resistive path for the
>>>  low frequency energy.
>>
>>  The ground grid impedance between any two points is well less than one
>>  ohm, so 100 ohms will pretty much abolish all ground loops.  I've used 10
>>  ohms in like labs, with success.  I'll grant that this would not work with
>>  long wires outside.
>
>Should be sufficient then. But remember that capacitive coupling helps
>you in the RF area and impulse protection.

True.


>  > By the way, I also finally talked to one of our most experienced EMI/EMC
>>  engineers.  He suggested using MIL-STD-461 test CS109, even though CS109
>  > was developed for enclosures.  It turns out he was involved in developing
>>  CS109 when he worked for the US Navy.
>
>Need to look it up. Never had to do any of the MIL-STD-461 stuff.

It's available for free on the web. 


Joe

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

Re: [time-nuts] Standards sought for immunity of shielded cable links to power-frequency ground loops

[Joe Gwinn]

At 4:59 AM + 1/7/09, time-nuts-requ...@febo.com wrote:
>
>Message: 6
>Date: Tue, 06 Jan 2009 21:54:41 -0600
>From: Brian Kirby 
>Subject: Re: [time-nuts] Standards sought for immunity of shielded
>   cable links to power-frequency ground loops
>To: Discussion of precise time and frequency measurement
>   
>Message-ID: <49642781.2020...@bellsouth.net>
>Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>
>During my experiences involving audio/phone, video and data
>transmission, we were taught to ground the shield at one end only so we
>would not cause a ground loop.

Yes, it's impossible to do this in a system of any size.  In my 
experience, the RF cables connect the arms of the star-grounding 
system, causing loops.  So, the receivers had to be immune.  The 
problem is to quantify and specify the required degree of immunity.


>I ran into problems everywhere I went with this and as much as folks
>disdain transformers, they are your friend in this type of problem.

DC blocks (usually a series capacitor) also work at RF.  But we would 
have a lot of them.  And we would still need some kind of spec to 
require, to know when we are done.


>Don White Consultants/Interference Control Technology published a whole
>series on EMI, Grounding, and EMC for the military.  They are located in
>Gainesville, VA.

But do they publish formal and official requirements documents? 
That's what I need, versus training.

Thanks,

Joe

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

Re: [time-nuts] Leap seconds and POSIX

[Joe Gwinn]

Magnus,

At 12:00 PM + 1/2/09, time-nuts-requ...@febo.com wrote:
>Message: 1
>Date: Fri, 02 Jan 2009 09:45:23 +0100
>From: Magnus Danielson 
>Subject: Re: [time-nuts] Leap seconds and POSIX
>To: Discussion of precise time and frequency measurement
>   
>
>>>  : In practice, in platforms that have access to GPS, NTP is used to
>>>  : servo the local computer clock into alignment with UTC (or GPS System
>>>  : Time (UTC without the accumulated leaps) in systems that abhor time
>>>  : steps), and there is a transient error just after a leap second while
>>>  : NTP recovers.
>>>
>>>  When the INS bit is set in the NTP packets, NTP tells the kernel about
>>>  it, which replays the last second of the day to keep in step.  I'm not
>>>  sure this is a transient error or not, since ntp_gettime can be used
>>>  to determine that this is the leap second for applications that care.
>>>  However, it does introduce a glitch in the data produced by system
>>>  interfaces that don't have leap second indicators...
>>
>>  Platforms vary because NTP is at the mercy of the kernel developers.
>>   From the standpoint of the average user, there is a transient error.
>>  Not that many average users will notice, so long as nothing crashes
>>  or hangs.
>
>For many of the places where POSIX is being used these days, this kind
>of reasoning is not helpful to say the least.

But is it correct?  Remember, most people understand little and care 
less about time, except insomuch as it affects something they do care 
about.

There has been a flurry of reports of leap-second induced failures on 
"comp.protocols.time.ntp".  This is precisely why some systems must 
avoid UTC.


>For instance, it is being used in many systems where high availability
>as well as propper logging is concerned. In addition to that, UTC may
>very well be the time-scale of choice, thought TAI could be used if
>properly identified as such.

In the financial and legal worlds, UTC is the standard choice, and 
many GPS receivers are purchased for legally-tracable timestamping.


>If POSIX standard time were TAI and NTP was steering it as TAI in all
>the boxes, then things would work. We would need to convert into UTC and
>then into local time-zone as part of presentation.

I recall Dave Mills discussing this as a possibility, but being 
unable to achieve it, for reasons I don't recall.


>If POSIX standard time were UTC and NTP was steering it as UTC in all
>the boxes, then things would work. We would need to convert into local
>time-zone as part of presentation.
>
>Oh, time-zone is a presentation issue and not a "box" issue.

There is also a move afoot to cease to have leap seconds , instead 
correcting UTC manually every ten or more years.   If this were done, 
then UTC would become uniform and smooth, and the TAI possibility 
would be eclipsed. The first possible ITU vote on this would be at 
their 2010 meeting, the ITU isn't likely to approve something that 
large the first time, and the DoD proposes that the change not be 
effective before 2018 at the earliest.   Ref: "Discontinuance of Leap 
Second Adjustments", John G. Grimes, Assistant [US] Secretary of 
Defense, Networks and Information Integration, 3 September 2008, 3 
pages.


Joe

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

Re: [time-nuts] Leap seconds and POSIX

[Joe Gwinn]

Magnus,

At 8:39 AM + 1/2/09, time-nuts-requ...@febo.com wrote:
>
>Message: 9
>Date: Fri, 02 Jan 2009 09:35:59 +0100
>From: Magnus Danielson 
>Subject: Re: [time-nuts] Leap seconds and POSIX
>To: Discussion of precise time and frequency measurement
>   
>Message-ID: <495dd1ef.7030...@rubidium.dyndns.org>
>Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>
>Joe Gwinn skrev:
>>  Having worked in the POSIX committee for many years, I can shed some
>>  light on how POSIX handles leap seconds:
>>
>>  In short, POSIX adamantly ignores leap seconds.  All days in POSIX
>>  have the same length, 86,400 seconds.
>>
>>  This omission is not by accident, instead having been violently
>>  debated at length, and voted upon.
>>
>>  The rationale is that one cannot assume that all POSIX systems have
>>  access to leap second information, or even the correct time, and yet
>>  must work in a reasonable manner.  In particular, file modification
>>  timestamps must allow one to determine causal order (to within one
>>  second in the old days) by comparison of timestamps.  (Yes, people do
>>  realize that timestamps are not the perfect way to establish causal
>>  order, but are nonetheless widely used in non-critical applications.
>>  Critical applications instead use some kind of atomic sequence-number
>>  scheme.)
>>
>>  So, at least in theory, POSIX time is a form of TAI, having a
>>  constant offset from TAI.
>>
>>  In practice, in platforms that have access to GPS, NTP is used to
>>  servo the local computer clock into alignment with UTC (or GPS System
>>  Time (UTC without the accumulated leaps) in systems that abhor time
>>  steps), and there is a transient error just after a leap second while
>>  NTP recovers.
>
>The problem with the POSIX time is that it can't be UTC but fails to
>identify itself as either TAI, UT1 or even UT2. If it clearly identified
>itself as being one of those, or similar (say GPS time) then things
>would be much improved. However, it is being interprented as being UTC
>which it fails to handle. I think the UT1 and UT2 alternatives would be
>best suited.
>
>I don't mind that the default time in POSIX remains there, but I don't
>think it helps that there is no way to get proper UTC time by a
>standardized interface.

Actually, this isn't quite true, as POSIX provides a standard (POSIX) 
Re: [time-nuts] Leap seconds and POSIX way to define non-standard (to 
POSIX) clocks.  So, if a vendor felt the need, they could define and 
provide a real TAI clock for instance.

That said, I don't know of anybody having done this, but I have not 
looked either.


>Attempting to say "It's UTC, except on leap seconds" is not a workable
>solution if you are locked up and a leap-second do occur. It only works
>if you are not locked up and free runs on whatever you have. However,
>more and more systems actually needs to be locked up and they may also
>need to have propper UTC. We see needs to have logs in UTC and
>coordinated over many countries and time-zones.
>
>This is not only a matter of how we deal with leap-seconds, but how we
>deal with time. We need to be able to actually deal with propper UTC
>regardless, and we need to know it is UTC "traceable" (in a wider sense
>most of the time, but occasionally in propper sense) or not.

There are something like 20 named timescales at the international 
level, most being paper clocks to be sure, but each has a purpose and 
a set of properties, and serves some need.  By implication, there is 
no single One True Clock.

A better way to approach this is to see that POSIX time is yet 
another timescale, with a purpose and a set of properties.


>I find the current situation unsatisfactory.

It is what it is.  Computer platform vendors traditionally cared 
little about time, and about realtime.  The rise of interest in audio 
and video media caused the vendors to become interested in realtime 
issues and performance, and thus indirectly about time.  But only to 
the degree needed to support handling of such media.


Joe

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

Re: [time-nuts] Leap seconds and POSIX

[Joe Gwinn]

Steve,

At 8:39 AM + 1/2/09, time-nuts-requ...@febo.com wrote:
>
>Message: 6
>Date: Fri, 2 Jan 2009 14:00:56 +1300
>From: "Steve Rooke" 
>Subject: Re: [time-nuts] Leap seconds and POSIX
>To: "Discussion of precise time and frequency measurement"
>   
>
>2009/1/2 Joe Gwinn :
>
>>  Platforms vary because NTP is at the mercy of the kernel developers.
>>   From the standpoint of the average user, there is a transient error.
>>  Not that many average users will notice, so long as nothing crashes
>>  or hangs.
>>
>>  In systems where the transient error and possibility of a crash or
>>  hang cannot be tolerated, one common dodge is to lie to NTP by
>>  configuring the GPS receiver and NTP timeserver to emit GPS System
>>  Time, and live with the fact that the local computer clocks are ~14+
>>  seconds off of UTC.  Purpose-built user displays are programmed to
>>  compute and use the correct time.
>
>Examples please.

Examples of what?

The systems of my direct experience are radars and the like.  Such 
systems always include trackers, and having a time step or a re-lived 
second can really destabilize things.   To avoid all such problems, 
one common approach is to create a uniform and smooth timescale for 
use by the radar software.

This was done long before GPS was invented, or NTP for that matter. 
Now days, the same smooth and uniform timescale is often implemented 
using a GPS receiver set to emit GPS System Time and NTP to 
distribute this time to the rest of the radar system.

Joe

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

Re: [time-nuts] Leap seconds and POSIX

[Joe Gwinn]

Warner,

At 6:35 PM + 1/1/09, time-nuts-requ...@febo.com wrote:
>
>Message: 6
>Date: Thu, 01 Jan 2009 11:28:03 -0700 (MST)
>From: "M. Warner Losh" 
>Subject: Re: [time-nuts] Leap seconds and POSIX
>To: time-nuts@febo.com, joegw...@comcast.net
>Message-ID: <20090101.112803.179959520@bsdimp.com>
>Content-Type: Text/Plain; charset=us-ascii
>
>In message: 
> Joe Gwinn  writes:
>: Having worked in the POSIX committee for many years, I can shed some
>: light on how POSIX handles leap seconds:
>:
>: In short, POSIX adamantly ignores leap seconds.  All days in POSIX
>: have the same length, 86,400 seconds.
>:
>: This omission is not by accident, instead having been violently
>: debated at length, and voted upon.
>:
>: The rationale is that one cannot assume that all POSIX systems have
>: access to leap second information, or even the correct time, and yet
>: must work in a reasonable manner.  In particular, file modification
>: timestamps must allow one to determine causal order (to within one
>: second in the old days) by comparison of timestamps.  (Yes, people do
>: realize that timestamps are not the perfect way to establish causal
>: order, but are nonetheless widely used in non-critical applications.
>: Critical applications instead use some kind of atomic sequence-number
>: scheme.)
>
>If POSIX had allowed for the system time to be TAI, and have the
>offset applied for the display of times, then there would be no
>ambiguity.  However, this is not allowed because one must be able to
>do math on time_t such that time_t % 86400 is midnight.

Defining a formal equivalence of some kind to TAI was proposed, but 
was not accepted, largely because they had were not linked at the 
beginning, and there would be many details that were not quite right.


>: So, at least in theory, POSIX time is a form of TAI, having a
>: constant offset from TAI.
>
>Except that the offset isn't constant :(.

True, as discussed next.


>: In practice, in platforms that have access to GPS, NTP is used to
>: servo the local computer clock into alignment with UTC (or GPS System
>: Time (UTC without the accumulated leaps) in systems that abhor time
>: steps), and there is a transient error just after a leap second while
>: NTP recovers.
>
>When the INS bit is set in the NTP packets, NTP tells the kernel about
>it, which replays the last second of the day to keep in step.  I'm not
>sure this is a transient error or not, since ntp_gettime can be used
>to determine that this is the leap second for applications that care.
>However, it does introduce a glitch in the data produced by system
>interfaces that don't have leap second indicators...

Platforms vary because NTP is at the mercy of the kernel developers. 
 From the standpoint of the average user, there is a transient error. 
Not that many average users will notice, so long as nothing crashes 
or hangs.

In systems where the transient error and possibility of a crash or 
hang cannot be tolerated, one common dodge is to lie to NTP by 
configuring the GPS receiver and NTP timeserver to emit GPS System 
Time, and live with the fact that the local computer clocks are ~14+ 
seconds off of UTC.  Purpose-built user displays are programmed to 
compute and use the correct time.

Joe

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

[time-nuts] Leap seconds and POSIX

[Joe Gwinn]

Having worked in the POSIX committee for many years, I can shed some 
light on how POSIX handles leap seconds:

In short, POSIX adamantly ignores leap seconds.  All days in POSIX 
have the same length, 86,400 seconds.

This omission is not by accident, instead having been violently 
debated at length, and voted upon.

The rationale is that one cannot assume that all POSIX systems have 
access to leap second information, or even the correct time, and yet 
must work in a reasonable manner.  In particular, file modification 
timestamps must allow one to determine causal order (to within one 
second in the old days) by comparison of timestamps.  (Yes, people do 
realize that timestamps are not the perfect way to establish causal 
order, but are nonetheless widely used in non-critical applications. 
Critical applications instead use some kind of atomic sequence-number 
scheme.)

So, at least in theory, POSIX time is a form of TAI, having a 
constant offset from TAI.

In practice, in platforms that have access to GPS, NTP is used to 
servo the local computer clock into alignment with UTC (or GPS System 
Time (UTC without the accumulated leaps) in systems that abhor time 
steps), and there is a transient error just after a leap second while 
NTP recovers.

Joe

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

Re: [time-nuts] Sub Pico Second Phase logger

[Joe Gwinn]

Pete,

At 4:58 AM + 12/27/08, time-nuts-requ...@febo.com wrote:
>
>Message: 2
>Date: Fri, 26 Dec 2008 19:10:47 -0700
>From: "Pete" 
>Subject: Re: [time-nuts] Sub Pico Second Phase logger
>To: "Discussion of precise time and frequency measurement"
>   
>
>Joe,
>
>I have been following this thread from the start & learned a lot from it.
>
>Thanks to you and Bruce for the insights.
>
>I thought I would take a look at a simple, passive approach, just to
>see what happens. I also limited my attention to stable 5 or 10 MHz
>sources & observation windows of <100s. I used a Mini-circuits
>SYPD-1 PD with a PSCQ-2-10.5 90 deg hybrid plus an ATM
>P1506 phase shifter (delta=0.32ns). There are lots of other items
>which will work just as well, this stuff happens to be on hand.
>
>The SYPD-1 also needs a diplexer to terminate the high order
>products & low pass filter the desired output; this is a cobble
>together item from the parts box. I used a PC logging program
>with my HP 3478A to take DC readings from the diplexer output.

What do you mean by "diplexer"?


>Thus armed, I first checked the DVM logging with no RF in to see
>if it was stable enough to yield a useable noise floor. This first result
>was good enough. Reading 720 points over 100 seconds; the data
>was limited to -1, 0 or +1 uV.
>
>Then I applied 10MHz @ to the 90 deg hybrid & connected one
>hybrid output to the SYPD-1 LO & the other hybrid output to the
>ATM P1506. The ATM output then drives the SYPD-1 RF port.
>I measured both SYPD-1 inputs to be +6.4dBm. Now I was able
>to set the ATM phase to zero the SYPD-1 output & verify it was
>stable over the 100s time I was interested in. The 3478A readings
>ranged from -1 to +2uV during this time with STDDEV of 0.6uV
>-again, good enough for me.
>
>Finally, I used a HP 105B to get a quiet 5MHz source & a 1MHz
>ext. reference for my HP3336C sythesizer. All this to give me 2
>"nearly" identical. phase locked sources to measure the calibration
>factor of the SYPD-1/Diplexer output. The 3336C was set to
>5,000,000.005Hz & +6.4dBm driving the SYPD-1 LO port. The
>105B 5MHz output was attenuated to +6.4dBm driving the RF
>port. The resulting 5mHz data was logged & analyzed at the zero
>crossings (max gain). The result is a phase sensitivity of 1.22uRad/uV.
>The + & - zero crossings were identical and the 1.22uRad/uV factor
>is linear (+/- 2 %) up to +/- 20 % of FS. FS = 1.023V
>
>If my math is about right, @10MHz, 1uRad = 16fs; so a noise floor
>of +/- 2uV (+/-2.5uRad) equals +/-40fs. I expect that, if 2 sources
>are near zero phase, then it should only require <10s to estimate
>the actual phase error within 1ps, or less. Anyone agree?
>
>I await & hope for constructive comments and/or corrections.

I've done similar things, and can also get random errors in the 
femtoseconds, but only over short periods of time.  The key is 
temperature sensitivity.   I would be suspicious of the 90 degree 
hybrid in particular, although the phase detector also has an 
opinion. The splitter phase unbalance is also quite sensitive to 
drive frequency, so the frequency stable sources are essential.

Joe

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

Re: [time-nuts] Sub Pico Second Phase logger

[Joe Gwinn]

Bruce,

At 10:16 PM + 12/23/08, time-nuts-requ...@febo.com wrote:
>
>Message: 1
>Date: Wed, 24 Dec 2008 08:47:21 +1300
>From: Bruce Griffiths 
>Subject: Re: [time-nuts] Sub Pico Second Phase logger
>To: Discussion of precise time and frequency measurement
>   
>
>[snip]
>  >
>    Using a DDS avoids the requirement for a pair of low phase noise VCOs.
>  
>  >>>  If we can control the spurs, many DDS chips are very good.
>  >>>
>  >> [BG] A DDS, unlike a conventional digital frequency divider, 
>doesn't suffer
>  >> from aliasing of phase noise into the output passband.
>  >
>>  [JG] How true is this, in practice?  A DDS is at the mercy of phase noise
>>  in its reference clock, by much the same mechanism as for a simple
>>  divider chain.  And the variable-factor dividers (the M and N above)
>>  work in a manner similar to a DDS, but with far coarser increments
>>  and limits.  Both DDS and M/N PLL chips use a PLL to clean up the
>>  resulting ref signal.  Many DDS chips incorporate a M/N PLL to
>  > multiply the ref frequency.
>  >  
>For NIST's measurements of this effect see:
>http://tf.nist.gov/timefreq/general/pdf/1380.pdf

Thanks for the reference.  Very interesting paper, although speckled 
with typos and gaps.   For instance, equation 5 is not grammatical, 
and appears to have been mangled.  There is a later and longer IEEE 
paper by the same authors that I'll get in January.

One would have expected this paper to have appeared 15 or 20 years 
earlier.  I guess the effect wasn't elucidated until many people 
tried to lock microwave sources to 10 MHz.

It also shows how to avoid the aliasing effect: Put an antialiasing 
filter and comparitor between stages.  This is what Ascarrunz does in 
his US patent 6,278,330, specifically Figure 4.  (This patent is what 
ref 3 of paper 1380 became.)


>[snip]
>  >> The simplest way of achieving the required performance is preferable.
>>>
>>
>>  Yes, but aren't we Time Nuts?
>>
>>
>Usually with finite budgets.

Few dollars, many hours.


Joe

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

Re: [time-nuts] New topics (was Re: He is a Time-Nut Troublemaker....)

[Joe Gwinn]

At 5:05 PM + 12/23/08, time-nuts-requ...@febo.com wrote:
>Date: Tue, 23 Dec 2008 11:40:51 -0500
>From: John Ackermann N8UR 
>Subject: [time-nuts] New topics (was Re: He is a Time-Nut
>   Troublemaker)
>To: Discussion of precise time and frequency measurement
>   
>
>Magnus Danielson wrote:
>
>>  My intent is to get some stuff done in the lab during the vacation.
>>  (Desperatly trying to get some more on-topic discussions going).
>
>Here are two questions that have been running around my head:
>
>[snip]
>2.  Several measurement techniques require a given phase relationship
>(e.g., quadrature) between DUT and reference.  For HF frequencies (ie, 5
>or 10 MHz) is there a *practical* phase shifter design covering 180+
>degrees that doesn't involve switching various lengths of coax in and
>out of the line?

Why the aversion to coax delay lines?  They are simple and reliable. 
The reason will help people to tell what alternatives would be best.

Joe

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

Re: [time-nuts] Sub Pico Second Phase logger

[Joe Gwinn]

At 5:05 PM + 12/23/08, time-nuts-requ...@febo.com wrote:
>Date: Tue, 23 Dec 2008 16:10:29 +
>From: Mark Sims 
>Subject: [time-nuts] Sub Pico Second Phase logger
>To: 
>Message-ID: 
>Content-Type: text/plain; charset="Windows-1252"
>
>
>Since when has prior art or obviousness kept the US Patent Orifice 
>from issuing a patent?  ;-)
>
>-
>
>[JG]  I just read 5,267,182.  It appears to anticipate Sotiriadis in every
>aspect, so it may be a *long* time before Sotiriadis's patent is
>issued.

The USPTO is in over their heads, both on expertise and on sheer 
volume of applications.

Given that the uncommon word "Diophantine" appears in both, there is 
a possibility that the Examiner will find 5,267,182.

Joe

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

Re: [time-nuts] Sub Pico Second Phase logger

[Joe Gwinn]

Bruce,

At 11:43 PM + 12/22/08, time-nuts-requ...@febo.com wrote:
>
>Date: Tue, 23 Dec 2008 12:01:40 +1300
>From: Bruce Griffiths 
>Subject: Re: [time-nuts] Sub Pico Second Phase logger
>To: Discussion of precise time and frequency measurement
>   
>
>[snip]
>  >> [BG]  Control of the close in spur levels produced by the NCO 
>may also be an
>>>  issue.
>>>
>>
>>  [JG]  If the division ratio is an integer, the spurs are reduced.
>>
>>  Analog Devices has a web calculator that's useful for fast
>>  exploration.  The calculator operation is explained and the reference
>>  provided in their big tutorial on DDS principles.
>>
>>
>>  
>It's a bit misleading in that it indicates that the performance is
>impaired when there is no phase truncation spur.

I just look at the spur chart.


>This may well be true for frequencies close to that frequency that has
>no phase truncation spur, however in this application one wouldn't
>choose an output frequency that has significant phase truncation spurs.

The value of the calculator is to look around for such sweet spots.

One can also generate the traditional spur chart, the one with the 
nest of diagonal lines.


>  >>>  I found and read the basic articles, which can be downloaded from
  
>>>   > Prof Sotiriadis' website:
>>>
   


  
>>>  [BG]  There is also US patent 5267182.
>>>
>>
>  > I did find 5,267,182 while searching for Sotiriadis' patent, but
>>  forgot to mention it.  Don't know what the difference is.  In the US,
>>  patent applications are made public after 18 months, or immediately
>>  if the person is also applying for a European patent.  The US Patent
>  > Office is years behind, so patent issuance may be a while.

I just read 5,267,182.  It appears to anticipate Sotiriadis in every 
aspect, so it may be a *long* time before Sotiriadis's patent is 
issued.


>  >>  > Items J10, J13, and J15 seem particularly relevant.
>>>   >
>>>
   All one needs is the M/N chip, although one can certainly use DDS chips.



  
>>>  Using a DDS avoids the requirement for a pair of low phase noise VCOs.
>>>
>>
>>  If we can control the spurs, many DDS chips are very good.
>>
>>  
>A DDS, unlike a conventional digital frequency divider, doesn't suffer
>from aliasing of phase noise into the output passband.

How true is this, in practice?  A DDS is at the mercy of phase noise 
in its reference clock, by much the same mechanism as for a simple 
divider chain.  And the variable-factor dividers (the M and N above) 
work in a manner similar to a DDS, but with far coarser increments 
and limits.  Both DDS and M/N PLL chips use a PLL to clean up the 
resulting ref signal.  Many DDS chips incorporate a M/N PLL to 
multiply the ref frequency.


>  >>  >>  >> A conventional mixer would then be used to either add or 
>subtract the
>>>
>   
>
>>>two DDS output frequencies.
>>>If the ratio of the 2 DDS clock source frequencies is appropriately
>>>chosen the spacing between the resultant mixer output 
>>>frequencies can be
>>>much finer than the spacing between the truncation spur 
>>>free outputs of
>>>either DDS chip.
>>>The DDS and mixer outputs should be filtered to remove harmonics and
>>>other unwanted frequencies.
>>>  
>>>   
>>>
>>If the DDS chips are well chosen, we will get sin and cos outputs,
>>and so can implement a dual-mixer phasing scheme to yield only the
>  >   sum frequency or only the difference frequency, greatly reducing the
>>amount of filtering needed.  The better balanced the channels are the
>> 
>>  
>> better the cancellation of the unwanted term.  This is basically the
>> phasing method of single-sideband signal generation.
>   
>
   This would be a reason to use DDS chips instead of M/N PLL chips,
   unless there are M/N PLL chips that provide quadrature outputs.
>  >>>  SiLabs Si5338 may suffice, as it allows one to control the relative
   phase of its outputs.


 
  
>>>  If one were to divide the output frequency of the diophantine
>>>  synthesizer by 4 using a 2 bit Johnson counter then quadrature phase
>>>  outputs are available.
>>>  However the filters used to extract the fundamental from the divider
>>>  outputs would need to be matched.
>>>  If the diophantine frequency synthesiser output frequency doesn't vary
>>>  too much one can always use a quadrature hybrid.
>>>
>>
>>  This seems like a lot of work.  Hmm.  Now that I think of it, the
>>  SiLabs chips emit square waves (logic signals), not sine waves.  Back
>>  to DDS chips, it seems.
>>
>>  Joe
>>
>>
>The simplest way of achieving the required performance is preferable.

Yes, but aren't we Time Nuts?

Joe

___

Re: [time-nuts] Sub Pico Second Phase logger

[Joe Gwinn]

Bruce,

At 9:08 PM + 12/22/08, time-nuts-requ...@febo.com wrote:
>
>Date: Tue, 23 Dec 2008 10:08:35 +1300
>From: Bruce Griffiths 
>Subject: Re: [time-nuts] Sub Pico Second Phase logger
>To: Discussion of precise time and frequency measurement
>   
>
>Joe
>Joe Gwinn wrote:
>>  At 2:01 AM + 12/22/08, time-nuts-requ...@febo.com wrote:
>>  
>>>  Message: 4
>>>  Date: Mon, 22 Dec 2008 11:19:06 +1300
>>>  From: Bruce Griffiths 
>>>  Subject: Re: [time-nuts] Sub Pico Second Phase logger
>>>  To: Discussion of precise time and frequency measurement
>>> 
>>>
>>>  Joe
>>>
>>>  Joe Gwinn wrote:
>>>   > Bruce,
>>>
>>>   >>
>>>
>  >> [snip]
>
>  >
>  >>>>  [BG]  The DDS based equivalent (of the dual PLL Diophantine 
>synthesiser) would
>>>>>   use a pair of DDS chips each replacing a conventional PLL in the
>>>>>   Diophantine frequency synthesiser, the output frequency of each having
>  >>>>  zero phase truncation spurs.
>>>>>
>>>>>   Both DDS clock sources should be spur free and have a frequency ratio
>>>>>   that is a selected fixed rational fraction.
>>>>>   
>>>>>
>>>>   [JG]  A M/N PLL chip can arrange this.  I recall that Silicon Labs makes
>  >>>  such a chip, which requires a parameter load on power-up, so a
>  >>> computer or FPGA is needed.
>  >>
>>
>  > 
><https://www.silabs.com/products/clocksoscillators/clocks/Pages/default.aspx>
>>
>>  Also of interest:
>>
>> 
>><https://www.silabs.com/products/clocksoscillators/clocks/Pages/Any-RateJitterAttenuatingClockMultipliers.aspx>
>>
>>
>>  
>The close in phase noise of these devices appears to be very high.
>This may preclude their use in a system that just uses 2 of them directly.

The SiLabs "Precision" units claim very low jitter (and thus low 
phase noise).  The above "jitter-attenuating" units are in this 
group.  I see that the spec numbers are still TBD though.

There may be other makers; this is a very competitive area.  I got 
the source from a RF engineer at work.

By the way, when we first used one of the simple PLL chips, we had 
failures.  It turned out that one had to bake the chips (to drive 
moisture off) longer than the original instructions said, and some of 
the chips cracked during soldering.  The new instructions eliminated 
the problem.

I suppose that all single-IC solutions will have worse phase noise 
than a well-engineered discrete unit, simply because it's hard to 
prevent undesired leakage of signals between parts of the the IC. 
However, the ICs may still be more than adequate, and are a whole lot 
easier than rolling our own discrete units.


>However in the high resolution version where the output frequencies of
>the 2 M/N synthesisers are each divided by a large factor before being
>added and subtracted from the reference (10MHz ??) frequency the
>resultant close in phase noise will be much lower.

And can be combined with use of the precision units.


>Control of the close in spur levels produced by the NCO may also be an
>issue.

If the division ratio is an integer, the spurs are reduced.

Analog Devices has a web calculator that's useful for fast 
exploration.  The calculator operation is explained and the reference 
provided in their big tutorial on DDS principles.


>  >> M and N only have to be relatively prime (ie the GCD of M and N is 1).
>>>  The ratio of M/N should also be close to 1.
>>>  If the spacing of the phase truncation spur free output frequencies is
>>>  about 10kHz (for either DDS) and M, N ~ 1000 the resultant mixer output
>>>  frequencies would have a spacing of about 10Hz which may be adequate for
>>>  this application.
>>>
>>
>>  I found and read the basic articles, which can be downloaded from
>  > Prof Sotiriadis' website:
>>
>>  <http://www.ece.jhu.edu/~pps/WEB/Publications/pub.html>
>>
>>
>There is also US patent 5267182.

I did find 5,267,182 while searching for Sotiriadis' patent, but 
forgot to mention it.  Don't know what the difference is.  In the US, 
patent applications are made public after 18 months, or immediately 
if the person is also applying for a European patent.  The US Patent 
Office is years behind, so patent issuance may be a while.


>  > Items J10, J13, and J15 seem particularly relevant.
>  >
>>  All one needs is the M/N chip, although one can certainly use DDS chips.
>>
>>
>>
>Using 

Re: [time-nuts] Sub Pico Second Phase logger

[Joe Gwinn]

At 2:01 AM + 12/22/08, time-nuts-requ...@febo.com wrote:
>
>Message: 4
>Date: Mon, 22 Dec 2008 11:19:06 +1300
>From: Bruce Griffiths 
>Subject: Re: [time-nuts] Sub Pico Second Phase logger
>To: Discussion of precise time and frequency measurement
>   
>
>Joe
>
>Joe Gwinn wrote:
>  > Bruce,
>>
>  >>
>
>[snip]
>>>
>>>
>>>> 
>>>>  
>>>>>   Some claim to be able to sync to an SPDIF input but the resultant jitter
>>>>>   may be large.
>>>>>   
>>>>>
>>>>   Why large jitter?  Bad implementation?
>>>>
>>>>
>>>> 
>>>>  
>>>  I'm just suspicious, although I did see some data somewhere that seemed
>>>  to confirm my suspicions.
>>>  The S/PDIF signal has to be a valid SPDIF signal not just a square or
>>>  sine wave clock.
>>>  Output sample rates (for the AP192) are then identical to that of the
>>>  the S/PDIF source which is limited to
>>>  192, 176.4, 96,88.2 48, 44.1 32 KSPS.
>>>
>>
>>  I did a little looking.  I bet that the sync quite well, but this
>>  signal is pretty complex.  One assumes that there is a box that takes
>  > in a 10 MHz ref and does the rest, because the broadcast industry
>>  does use atomic clocks.
>>
>>
>>  
>Ulrich has built a circuit that takes a sampling frequency input derived
>from a 10MHz GPSDO output and produces an S/PDIF output for this
>application.
>Its certainly worth trying since all the specs for the sound card aren't
>readily available.

Ulrich posted some details, and I got the datasheet for study.

But I bet someone already makes the necessary box.


>  >>>>  Conventional Diophantine synthesis uses number theory 
>together with 2 or
>>>>>   3 conventional synthesiser loops to achieve very fine resolution whilst
>>>>>
>>>   >> maintaining a high PLL phase detector input frequency.
>>>   >
>>>
>>>>   In a sense, the concatenated DDS approach is a divide-and-mix chain.
>>>>  
>>>   > Perhaps there is a parallel here.
>>>   > 
>>>  The DDS based equivalent (of the dual PLL Diophantine synthesiser) would
>>>  use a pair of DDS chips each replacing a conventional PLL in the
>>>  Diophantine frequency synthesiser, the output frequency of each having
>>>  zero phase truncation spurs.
>>>
>>>  Both DDS clock sources should be spur free and have a frequency ratio
>>>  that is a selected fixed rational fraction.
>>>
>>
>>  A M/N PLL chip can arrange this.  I recall that Silicon Labs makes
>>  such a chip, which requires a parameter load on power-up, so a
>  > computer or FPGA is needed.

<https://www.silabs.com/products/clocksoscillators/clocks/Pages/default.aspx>

Also of interest:

<https://www.silabs.com/products/clocksoscillators/clocks/Pages/Any-RateJitterAttenuatingClockMultipliers.aspx>


>M and N only have to be relatively prime (ie the GCD of M and N is 1).
>The ratio of M/N should also be close to 1.
>If the spacing of the phase truncation spur free output frequencies is
>about 10kHz (for either DDS) and M, N ~ 1000 the resultant mixer output
>frequencies would have a spacing of about 10Hz which may be adequate for
>this application.

I found and read the basic articles, which can be downloaded from 
Prof Sotiriadis' website:

<http://www.ece.jhu.edu/~pps/WEB/Publications/pub.html>

Items J10, J13, and J15 seem particularly relevant.

All one needs is the M/N chip, although one can certainly use DDS chips.


>  >> A conventional mixer would then be used to either add or subtract the
>>>  two DDS output frequencies.
>>>  If the ratio of the 2 DDS clock source frequencies is appropriately
>>>  chosen the spacing between the resultant mixer output frequencies can be
>>>  much finer than the spacing between the truncation spur free outputs of
>>>  either DDS chip.
>>>  The DDS and mixer outputs should be filtered to remove harmonics and
>>>  other unwanted frequencies.
>>>
>>
>>  If the DDS chips are well chosen, we will get sin and cos outputs,
>>  and so can implement a dual-mixer phasing scheme to yield only the
>>  sum frequency or only the difference frequency, greatly reducing the
>>  amount of filtering needed.  The better balanced the channels are the
>  > better the cancellation of the unwanted term.  This is basically the
>  > phasing method of single-sideband signal generation.

This would be a reason to use DDS chips instead of M/N PLL chips, 
unless there are M/N PLL chips that provide quadrature outputs. 
SiLabs Si5338 may suffice, as it allows one to control the relative 
phase of its outputs.


Joe

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

Re: [time-nuts] Sub Pico Second Phase logger

[Joe Gwinn]

Ulrich,

At 7:57 AM + 12/22/08, time-nuts-requ...@febo.com wrote:
>
>Message: 7
>Date: Mon, 22 Dec 2008 08:56:55 +0100
>From: "Ulrich Bangert" 
>Subject: Re: [time-nuts] Sub Pico Second Phase logger
>To: "'Discussion of precise time and frequency measurement'"
>   
>
>Joe and Bruce,
>
>>  Ulrich has built a circuit that takes a sampling frequency
>>  input derived from a 10MHz GPSDO output and produces an
>>  S/PDIF output for this application. Its certainly worth
>>  trying since all the specs for the sound card aren't readily
>>  available.
>
>I use a cheap and easy to use CRYSTAL CS8402 S/PDIF transmitter to
>generate the S/PDIF signal for the sound card. Data input just tight to
>ground. The balanced output is done with a Minicircuits T1-1T
>transformer. The chip needs an 12.844 MHz clock for 96 kHz sample rate
>and 6.144 MHz for 48 kHz sample rate. However, these two frequencies are
>generated by an old HP3330 locked to 10 MHz in my case. If the circuit
>had to be designed new I would have used a ANALOG DEVICES dds circuit
>with a built in clock multiplier.

Thanks for the pointer.  I'll look at the datasheet of the CS8406, 
which replaces the CS8402.



Joe

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

Re: [time-nuts] Sub Pico Second Phase logger

[Joe Gwinn]

Bruce,

At 9:50 PM + 12/19/08, time-nuts-requ...@febo.com wrote:
>
>Message: 8
>Date: Sat, 20 Dec 2008 10:50:46 +1300
>From: Bruce Griffiths 
>Subject: Re: [time-nuts] Sub Pico Second Phase logger
>To: Discussion of precise time and frequency measurement
>   
>
>Joe
>
>Joe Gwinn wrote:
>>  Bruce,
>>
>>  At 3:54 AM + 12/19/08, time-nuts-requ...@febo.com wrote:
>>  
>>>  Message: 6
>>>  Date: Fri, 19 Dec 2008 16:54:12 +1300
>>>  From: Bruce Griffiths 
>>>  Subject: Re: [time-nuts] Sub Pico Second Phase logger
>>>  To: Discussion of precise time and frequency measurement
>  >>   
>>>
>[snip]
>  >> I'll breadboard an oscillator over the holidays and take some samples.
>>>
>>
>>  A rainy-day activity for sure.   In the Boston area, the issue is
>>  snow - a noreaster is just now starting, with a foot of snow expected.
>>
>>  Joe
>>
>>  
>AFAIK it has never snowed here.

So it instead rains heavily?


>It does snow on the central NI Plateau but not usually in December.

Well, I've just dug out of 1.5 feet of snow already, and it continues 
to snow.  The threat is that later we will get rain, and then it will 
all freeze into solid ice.  This in the Boston area.


>I'll start with a diode limited (soft clamps - resistance in series with
>diodes) oscillator and move to a lower distortion one if necessary.

What distortion level do we need?  Given that the parts cost nothing, 
I'd be tempted into some overkill.

Joe

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

Re: [time-nuts] Sub Pico Second Phase logger

[Joe Gwinn]

Bruce,

At 9:50 PM + 12/19/08, time-nuts-requ...@febo.com wrote:
>
>Message: 7
>Date: Sat, 20 Dec 2008 10:38:26 +1300
>From: Bruce Griffiths 
>Subject: Re: [time-nuts] Sub Pico Second Phase logger
>To: Discussion of precise time and frequency measurement
>   
>
>Joe
>
>Joe Gwinn wrote:
>>  Bruce,
>>
>>  At 3:54 AM + 12/19/08, time-nuts-requ...@febo.com wrote:
>>  
>>>  Message: 5
>>>  Date: Fri, 19 Dec 2008 16:51:55 +1300
>>>  From: Bruce Griffiths 
>>>  Subject: Re: [time-nuts] Sub Pico Second Phase logger
>>>  To: Discussion of precise time and frequency measurement
>>> 
>>>
>>>  Joe
>>>
>>>  Joe Gwinn wrote:
>>>
>>>>   At 11:48 PM + 12/18/08, time-nuts-requ...@febo.com wrote:
>  >>>
>[snip]
>
>>  
>>>  Some claim to be able to sync to an SPDIF input but the resultant jitter
>>>  may be large.
>>>
>>
>>  Why large jitter?  Bad implementation?
>>
>>
>>  
>I'm just suspicious, although I did see some data somewhere that seemed
>to confirm my suspicions.
>The S/PDIF signal has to be a valid SPDIF signal not just a square or
>sine wave clock.
>Output sample rates (for the AP192) are then identical to that of the
>the S/PDIF source which is limited to
>192, 176.4, 96,88.2 48, 44.1 32 KSPS.

I did a little looking.  I bet that the sync quite well, but this 
signal is pretty complex.  One assumes that there is a box that takes 
in a 10 MHz ref and does the rest, because the broadcast industry 
does use atomic clocks.


>  >
>  >>  >>  >> A few divide and mix stages will be required to achieve 
>a spur free resolution of 10Hz.
>  >>>>>>
>>>>>>That is a traditional approach.  But are there alternate 
>>>>>>approaches that
>  >>>>>   have now become practical?
>  >>>>>
>  >>>>  Diophantine frequency synthesis?
>  >>>>
>>>>From the sound of the name I think so, at least in the last DDS
>>>>   stage, as done by that patent.
>>>>
>  >>>  But I was fishing.
>  >>>  
>>>  Conventional Diophantine synthesis uses number theory together with 2 or
>>>  3 conventional synthesiser loops to achieve very fine resolution whilst
>  >> maintaining a high PLL phase detector input frequency.
>  >
>>  In a sense, the concatenated DDS approach is a divide-and-mix chain.
>  > Perhaps there is a parallel here.
>  >  
>The DDS based equivalent (of the dual PLL Diophantine synthesiser) would
>use a pair of DDS chips each replacing a conventional PLL in the
>Diophantine frequency synthesiser, the output frequency of each having
>zero phase truncation spurs.
>
>Both DDS clock sources should be spur free and have a frequency ratio
>that is a selected fixed rational fraction.

A M/N PLL chip can arrange this.  I recall that Silicon Labs makes 
such a chip, which requires a parameter load on power-up, so a 
computer or FPGA is needed.


>A conventional mixer would then be used to either add or subtract the
>two DDS output frequencies.
>If the ratio of the 2 DDS clock source frequencies is appropriately
>chosen the spacing between the resultant mixer output frequencies can be
>much finer than the spacing between the truncation spur free outputs of
>either DDS chip.
>The DDS and mixer outputs should be filtered to remove harmonics and
>other unwanted frequencies.

If the DDS chips are well chosen, we will get sin and cos outputs, 
and so can implement a dual-mixer phasing scheme to yield only the 
sum frequency or only the difference frequency, greatly reducing the 
amount of filtering needed.  The better balanced the channels are the 
better the cancellation of the unwanted term.  This is basically the 
phasing method of single-sideband signal generation.


>One drawback is that selecting the output frequencies of the 2 DDS chips
>required to produce the desired output frequency is somewhat complex.
>Since one almost certainly needs a computer of some sort to set the DDS
>frequencies this shouldn't be a significant issue.

Yes.  Basically, every digital chip in this scheme requires a 
computer of some kind.  What I've seen used is a small CPLD to do the 
initial parameter loads, and then to run the show.

Joe

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

Re: [time-nuts] Sub Pico Second Phase logger

[Joe Gwinn]

Bruce,

At 3:54 AM + 12/19/08, time-nuts-requ...@febo.com wrote:
>
>Message: 6
>Date: Fri, 19 Dec 2008 16:54:12 +1300
>From: Bruce Griffiths 
>Subject: Re: [time-nuts] Sub Pico Second Phase logger
>To: Discussion of precise time and frequency measurement
>   
>
>Joe
>Joe Gwinn wrote:
>>  This is from home.  I'll not be at work until next year.
>>
>>  At 11:48 PM + 12/18/08, time-nuts-requ...@febo.com wrote:
>>  
>>>  Message: 4
>>>  Date: Fri, 19 Dec 2008 12:17:33 +1300
>>>  From: Bruce Griffiths 
>>>  Subject: Re: [time-nuts] Sub Pico Second Phase logger
>>>  To: Discussion of precise time and frequency measurement
>  >>   
>  >>
>>>  Joe
>>>
>>>
>>>   >>>
>  >>
>[snip]
>  >>  >> Sources:
>>>
>>>>>   1) Wien bridge or equivalent (eg state variable oscillator with soft
>>>>>   clamping) relatively low distortion oscillator.
>>>>>
>>>>>   2) Buffered low pass filtered output of binary divider driven by a
>>>>>   crystal oscillator
>>>>>   
>>>>>
>>>   > RC oscillator sounds far simpler and more flexible.
>>>
>>>  A Wien bridge using a lamp is perhaps the simplest.
>>>  I'll create a circuit schematics for this using an OPA2134 (dual lowish
>>>  noise JFET opamp).
>>>  One opamp for the oscillator one to drive the sound card (attenuation of
>>>  the oscillator output will be required for some sound cards and it is
>>>  desirable to have a low output impedance driver).
>>>
>>
>>  Jim Williams of Linear Technology had a very good low-distortion AGC
>>  controlled Wein Bridge oscillator.  If I recall, he used a photo-FET
>>  or the like as the servoed resistor in the bridge.  There may be an
>>  application note on the LT website, but I saw it in a chapter of a
>>  book on analog circuitry, the chapter author being Jim W.
>>
>>
>>  
>I have the application note.

LT App Note 43 is what JW mentions.  Figure 47 is the key.  (He also 
mentions App Note 5.)

These cites are from the refs of "Max Wien, Mr. Hewlett, and a Rainy 
Sunday Afternoon", Jim Williams, Chapter 7 in "Analog Circuit Design 
-- Art, Science, and Personalities, edited by Jim Williams, 
Butterworth-Heinemann 1991.  The relevant section of App Note 43 
seems to be shared with this chapter.

It wasn't a photo-FET he used, it was a Vactec model VTL5C10 
optically (LED) driven CdS photocell.

Most of it is here (what's missing is some history and the refs): 
<http://books.google.com/books?id=Il4xxTTyhbEC&pg=PA43&lpg=PA43&dq=Max+Wien,+Mr.+Hewlett,+and+a+Rainy+Sunday+Afternoon&source=bl&ots=jnHNItJrOF&sig=ZL_CB_yIdT0ZbsD7Ieu_CjVYJHk&hl=en&sa=X&oi=book_result&resnum=1&ct=result#PPA53,M1>.


>  >> For an AP192 the directly (without sample rate interpolation) available
>>>  output sample rates are:
>>>
>>>  192, 96, 64, 48, 32, 8 KSPS.
>>>
>>
>>  OK.  I would start with 8 ksps.  We will end up decimating below that
>>  anyway, except for 1 KHz test signals.
>>
>>
>>  Joe
>>
>>
>I'll breadboard an oscillator over the holidays and take some samples.

A rainy-day activity for sure.   In the Boston area, the issue is 
snow - a noreaster is just now starting, with a foot of snow expected.

Joe

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

Re: [time-nuts] Sub Pico Second Phase logger

[Joe Gwinn]

Bruce,

At 3:54 AM + 12/19/08, time-nuts-requ...@febo.com wrote:
>
>Message: 5
>Date: Fri, 19 Dec 2008 16:51:55 +1300
>From: Bruce Griffiths 
>Subject: Re: [time-nuts] Sub Pico Second Phase logger
>To: Discussion of precise time and frequency measurement
>   
>
>Joe
>
>Joe Gwinn wrote:
>>  At 11:48 PM + 12/18/08, time-nuts-requ...@febo.com wrote:
>>  
>>>  Message: 5
>>>  Date: Fri, 19 Dec 2008 12:48:27 +1300
>>>  From: Bruce Griffiths 
>>>  Subject: Re: [time-nuts] Sub Pico Second Phase logger
>>>  To: Discussion of precise time and frequency measurement
>  >>   
>  >>
>  >> [snip]
>  >  
>>>  Using a second sound card to generate the test signal may overcome this
>>>  problem at increased cost, and for some it may not even be an option.
>  >>
>>
>>  It may not work with PCI soundcards, as the card clock may be
>>  synchronized to the PCI bus clock.  Firewire/USB cards will have
>>  their own independent clocks.
>>
>>  
>Most recent design PCI sound cards have their own independent crystal
>oscillator.

It won't be a great crystal, but a crystal nonetheless.


>Some claim to be able to sync to an SPDIF input but the resultant jitter
>may be large.

Why large jitter?  Bad implementation?


>  >  
>>>   >> 10Hz resolution whilst avoiding phase truncation spurs is impractical
>>>
>>>>>   with a DDS chip by itself.
>>>>>   Depending on the DDS and its clock frequency, the frequency spacing of
>>>>>   phase truncation spur free outputs may be as large as several kHz.
>>>>>   
>>>>>
>  >>>  Is this true of concatenated DDS chips?  I recall a patent to the
>  >>> contrary.
>  >>  >
>  >> Which patent?
>  >
>>  Hmm.  It's at work.  I'll look it up in January.  As I recall, the
>>  second DDS made a small integer conversion, and so had low spurs,
>>  while the first DDS was set to whatever was needed.
>>
>>
>Do you mean US5598440?

Yes, that's it.


>  >> If the zero crossings are time stamped and do not occur simultaneously
>>>  in each channel then the phase noise of the offset oscillator will
>  >> affect the measurement.
>  >
>>  I'm not following.  Please expand.  The zero crossings are never
>>  aligned unless there is no phase delay.
>>
>>
>>  
>Yes, thats the point, the offset generator phase noise contribution isnt
>the same for both zero crossings.
>Greenhall et al correct for this to some extent, but at least for short
>tau, one is then no longer measuring ADEV, MDEV etc.
>There is some advantage in having a higher beat frequency as the offset
>generator phase noise has less time to accumulate.

I recall reading an article on this by Greenhall, probably _the_ 
article.  My reaction was that I would be fortunate to have a setup 
where I could even detect such a problem.


>  >>  >> A few divide and mix stages will be required to achieve a spur free
>>>
>>>>>   resolution of 10Hz.
>>>>>   
>>>>>
>>>>   That is a traditional approach.  But are there alternate approaches that
>>>>   have now become practical?
>>>>
>>>>
>>>>  
>>>  Diophantine frequency synthesis?
>>>
>>
>>   From the sound of the name I think so, at least in the last DDS
>>  stage, as done by that patent.
>>
>>  But I was fishing.
>>
>>  
>Conventional Diophantine synthesis uses number theory together with 2 or
>3 conventional synthesiser loops to achieve very fine resolution whilst
>maintaining a high PLL phase detector input frequency.

In a sense, the concatenated DDS approach is a divide-and-mix chain. 
Perhaps there is a parallel here.


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

Re: [time-nuts] Sub Pico Second Phase logger

[Joe Gwinn]

At 11:48 PM + 12/18/08, time-nuts-requ...@febo.com wrote:
>
>Message: 5
>Date: Fri, 19 Dec 2008 12:48:27 +1300
>From: Bruce Griffiths 
>Subject: Re: [time-nuts] Sub Pico Second Phase logger
>To: Discussion of precise time and frequency measurement
>   
>Message-ID: <494ae14b.2050...@xtra.co.nz>
>Content-Type: text/plain; charset=ISO-8859-1
>
>[snip]
>
>  >>>  
>>>  This just adds another layer of complexity for little immediate gain.
>>>  Making the algorithms robust against small drifts in beat frequency is
>>>  more useful in the general case (when 2 different test sources are being
>>>  compared) than just assuming that the the beat frequency is very stable
>>>  and fixed.
>>>
>>
>>  Yes, but I'm not sure we are solving the same problem.
>>
>>  I suppose the sound card could drive a simple PLL signal cleanup circuit.
>>
>>
>>  
>One potential problem with using a sound card for a test source is that
>the output DAC may share the same clock as the ADC ensuring that the
>output signal is locked to the ADC sampling rate.

True.  I would expect everything on a given card to use the same clock.


>In an actual dual mixer system the beat frequency and the ADC clock won't
>necessarily be synchronised (it's difficult to lock the sampling clock of
>most sound cards to an external reference).

Maybe impossible.  While stuff used by the Broadcast folk probably 
does accept external sync, it may as well be made of purest beaten 
Gold.


>If one isn't careful the algorithms developed may work well when the ADC
>sample clock and the test frequencies are locked, but have problems when
>they are not.

True.


>Using a second sound card to generate the test signal may overcome this
>problem at increased cost, and for some it may not even be an option.

It may not work with PCI soundcards, as the card clock may be 
synchronized to the PCI bus clock.  Firewire/USB cards will have 
their own independent clocks.


>[snip]
>  >>>  
>>>  The immediate task is actually to evaluate sound cards for their
>>>  suitability, preferably without the added cost and complexity of a DDS
>>>  LO and mixer.
>>>
>>
>>  Suitability for what?  That is the point of enumerating tasks.
>>
>>  
>Suitability for use in a dual mixer system.

Ah.


>  >> 10Hz resolution whilst avoiding phase truncation spurs is impractical
>>>  with a DDS chip by itself.
>>>  Depending on the DDS and its clock frequency, the frequency spacing of
>>>  phase truncation spur free outputs may be as large as several kHz.
>>>
>>
>>  Is this true of concatenated DDS chips?  I recall a patent to the
>  > contrary.
>  >  
>Which patent?

Hmm.  It's at work.  I'll look it up in January.  As I recall, the 
second DDS made a small integer conversion, and so had low spurs, 
while the first DDS was set to whatever was needed.


>If the zero crossings are time stamped and do not occur simultaneously
>in each channel then the phase noise of the offset oscillator will
>affect the measurement.

I'm not following.  Please expand.  The zero crossings are never 
aligned unless there is no phase delay.


>  >> A few divide and mix stages will be required to achieve a spur free
>>>  resolution of 10Hz.
>>>
>>
>>  That is a traditional approach.  But are there alternate approaches that
>>  have now become practical?
>>
>>
>Diophantine frequency synthesis?

 From the sound of the name I think so, at least in the last DDS 
stage, as done by that patent.

But I was fishing.


>  >  
>>>  To broaden participation we need to broaden the scope of the project to
>>>  include dual mixer system with statistically independent test sources as
>>>  well as the more specialised case where the 2 input frequencies differ
>>>  only in phase.
>>>
>>>  1) Evaluate sound cards for suitablility.
>>>  Initially use simple less stable sources and follow up with more stable
>>>  test sources for the more promising cards.
>>>  Need to measure crosstalk, temporal instability of interchannel phase
>>>  shift, system noise etc.
>>>
>>
>>  Generally agree, but there is that undefined elastic term "suitability"
>  > again.
>>
>>  
>Replace suitability with:
>Measure the characteristics of a sound card that affect the performance
>when used in a dual mixer system used to measure the relative phases of
>a pair of RF signals.
>Where the pair of RF signals may either:
>
>1) originate from 2 statistically independent sources (OCXOs, GPSDO's etc)
>
>OR
>
>2) originate from the same source and just differ in phase.

That covers it for sure.


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

Re: [time-nuts] Sub Pico Second Phase logger

[Joe Gwinn]

This is from home.  I'll not be at work until next year.

At 11:48 PM + 12/18/08, time-nuts-requ...@febo.com wrote:
>Message: 4
>Date: Fri, 19 Dec 2008 12:17:33 +1300
>From: Bruce Griffiths 
>Subject: Re: [time-nuts] Sub Pico Second Phase logger
>To: Discussion of precise time and frequency measurement
>   
>Message-ID: <494ada0d.3060...@xtra.co.nz>
>Content-Type: text/plain; charset=ISO-8859-1
>
>Joe
>
>
>  >>>  
>>>  [BG] Proposed test setup:
>>>  (preliminary to be refined)
>>>
>>>  Drive 2 sound card inputs in parallel with the same source.
>>>
>>>  Source amplitude:
>>>  Max sound card input -3dB
>>>
>>
>>  What kind of dB?
>>
>>  
>Peak input signal voltage = 70% of sound card maximum peak input voltage.
>Just to leave some margin for gain tolerances.

OK.


>  >> Sources:
>>>
>>>  1) Wien bridge or equivalent (eg state variable oscillator with soft
>>>  clamping) relatively low distortion oscillator.
>>>
>>>  2) Buffered low pass filtered output of binary divider driven by a
>>>  crystal oscillator
>>>
>>
>  > RC oscillator sounds far simpler and more flexible.
>
>A Wien bridge using a lamp is perhaps the simplest.
>I'll create a circuit schematics for this using an OPA2134 (dual lowish
>noise JFET opamp).
>One opamp for the oscillator one to drive the sound card (attenuation of
>the oscillator output will be required for some sound cards and it is
>desirable to have a low output impedance driver).

Jim Williams of Linear Technology had a very good low-distortion AGC 
controlled Wein Bridge oscillator.  If I recall, he used a photo-FET 
or the like as the servoed resistor in the bridge.  There may be an 
application note on the LT website, but I saw it in a chapter of a 
book on analog circuitry, the chapter author being Jim W.


>  >> Test frequencies:
>>>
>>>  100Hz
>>>
>>>  1kHz
>>>
>>
>>  Why no 10 Hz?  (Well, 20 Hz.)
>>
>>  
>No particular reason other than some complications if a lamp stabilised
>oscillator is used.
>A diode soft (series R) clamped RC oscillator is more flexible in this
>regard.
>I'll also produce a circuit schematic for one of these oscillators.

Jim Williams' circuit would handle 10 Hz if I recall.


>  >> Sound card sample rate:
>>>
>>>  ~24KSPS
>>>
>>
>>  I assume that this is the lowest rate supported, and certainly is overkill
>>  for 1 KHz.
>>
>>  
>It varies with the sound card.
>I just suggested that for a starting point in the discussion.
>
>For an AP192 the directly (without sample rate interpolation) available
>output sample rates are:
>
>192, 96, 64, 48, 32, 8 KSPS.

OK.  I would start with 8 ksps.  We will end up decimating below that 
anyway, except for 1 KHz test signals.


Joe

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