Bob Camp wrote:
Hi
There's always the ever popular 5370 (a or b) off auction on the cheap ...
The box we had did indeed have a pretty good time stamper. It would take in
"anything or everything" and ultimately tell you what was happening with it. I
don't recall the GPS data being good to the ~20 ps level.
It isn't (in the short term) even with carrier phase data.
They just made the timestamp resolution so high (perhaps it became an
exercise in itself to see how good they could make it) that its
contribution to the measurement noise is insignificant compared to the
GPS receiver timing noise.
Simply doing<1 ns level time transfer would be a major upgrade over what most of us have
available. If you could do a number of "shots" over an hour or so that certainly
would do me a *lot* of good. That of course assumes I'm borrowing a very good standard on
the other end.
The final comparison doesnt even have to be done in real time so the
other end could publish measurements on a website for example.
The better GPS timing receivers allow the timing data for each SV
relative to the (sawtooth corrected) PPS output to be recorded so that
timing relative to those SVs seen at both ends can be computed.
Bob
Bruce
On Mar 15, 2010, at 9:40 PM, Bruce Griffiths wrote:
According to the available documentation for one variant of the NIST box, it
had (apart from the GPS receiver) a multichannel time stamping counter to time
stamp PPS pulses from external signal sources (generated using a programmable
divider so that any source with a frequency of up to around 100MHz or so that
was nominally a harmonic of 1Hz could be used).
The timestamp resolution of was about 25ps.
Bruce
Bob Camp wrote:
Hi
One could always use the ever popular super cooled sapphire resonator
oscillator and an equally handy optical ion standard.
--------------------
Anything that involves comparing same to same is vulnerable to all sorts of
common mode effects. Ensembles of cesiums all made on the same line can have
issues. Even having something like (yikes!) WWVB to throw in the mix will help
rule out part of the errors that creep in.
One thing we haven't talked about (but could) are time transfer setups via
common view GPS. Back when we had a NIST box it did them under robotic control.
Judging from the data plots it did a very good job.
There really wasn't a lot in the silly box. Their system did indeed use a (at
that time) fancy antenna and a pretty good receiver. It's not clear to me that
the 1980's technology in the box could not be duplicated today for a lot less
than they put into it. Anything you can toss into the mix is a good thing. With
good enough time transfer you could use the H-Maser in somebody else's basement
.....
Bob
On Mar 15, 2010, at 8:55 PM, Bruce Griffiths wrote:
Unless you have a CSO or a hydrogen maser absolute measures of ADEV and phase
noise arent feasible for the range of Tau of interest.
Even an indirect method such as measuring the location of the apparent minimum
in ADEV between the GPS SV constellation observables and the OCXO when it is
undisciplined depend heavily on the ADEV characteristics of the OCXO being used.
Bruce
WarrenS wrote:
yeah,
So many variables, ALL the more reasion to just see what the overall effect is
on the more common type of GPSDO receviers at a few sites.
So did you have a better plan?
ws
****************
Bruce Griffiths Added:
WarrenS wrote:
Bruce wrote:
"Which antenna performance metric do you have in mind?"
Could do GPSDO hold over performance, but that would not be much of
a test of the antenna.
How about the antenna's effect on the ADEV Osc noise and Phase noise.
What else does the Time Nut care about?
Since the better timing receivers use carrier phase ...
I don't remember you ever finding ANY Time Nut that is now using one.
So may be simpler for now to just stick to the more common type of
GPSDO in use.
The Motorola M12+T and iLotus M12M use carrier phase smoothing of the
code phase observables.
Its highly likely that a number of the better performance GPS timing
receivers also use carrier phase smoothing.
Thus whether one is aware of it or not the antenna carrier phase
properties are likely to be of some importance.
In the absence of complete information on how your particular GPS timing
receiver uses carrier phase and code phase observables, the best you can
do is compare the performance of a range of antennas using a given
timing receiver.
Such results will only apply to a particular site and receiver.
Specifying the pertinent characteristics (eg isolated on a flat plain,
surrounded by a set of hills, mountains which obscure the sky below 10
degrees, surrounded by trees that obscure everything below 40 degeees
elevation, etc) of your antenna location and the particular GPS receiver
used will be helpful to others in selecting an antenna that suits their
budget, receiver, antenna location constraints, etc.
Bruce
ws
*************
Bruce said:
Which antenna performance metric do you have in mind?
There are several, some of which are considered in the paper:
http://www.novatel.com/Documents/Papers/effectofantenna.pdf
Since the topography surrounding the antenna, its height and location on
the Earth all affect measured performance any comparative measurements
should use the same receiver and antenna location.
Some estimates for the effect of multipath on code phase receivers can
be found:
http://bbcr.uwaterloo.ca/~wzhuang/papers/iee95_gps.pdf
http://bbcr.uwaterloo.ca/%7Ewzhuang/papers/iee95_gps.pdf
Since the better timing receivers use carrier phase smoothing of the
code phase timing, both the carrier phase and code phase performance of
the antenna are important.
A phased array antenna like the one in the following papers may provide
better performance than alternative antennae:
http://www.navsys.com/Papers/0001002.pdf
http://www.congrex.nl/07c12/papers/day1_s1_paper05_Konovaltsev.pdf
Some measurements with geodetic antennae:
http://www.fig.net/pub/fig2008/papers/ts05g/ts05g_03_eventzur_shaked_2816.pdf
Comparison of code phase and carrier phase time transfer:
http://tycho.usno.navy.mil/ptti/ptti2004/paper41.pdf
Bruce
******************
WarrenS wrote:
Brian wrote:
"There were also comments about surveying and timing antennas."
Those may of been from me, unsuccessfully trying to make a point of
the difference between what is 'Best' and what is 'GOOD enough'.
"about every national timing laboratory uses choke ring antennas.
... for timing stability reasons."
Then again they also have multiple CS and Just their Antenna budget is
likely more than the annual income of most time nuts.
Can you do a test to show IF there is ANY improvement for the AVERAGE
time nut when compared to a well setup (Tbolt) GPSDO using a TacoSalad
antenna?
Would be interesting to see a plot of cost vs. performance for the
various antenna types,
Scaled to show the performance improvement that the average Time nut
would see.
The TacoSalad antenna, originally cost me a total of $7.95, And took
under 30 seconds to build.
That cost should be discounted because those parts had been considered
just throw away junk up until now.
ws
**********************
----- Original Message ----- From: "Brian Kirby"
<kilodelta4foxmike at gmail.com>
To: "Discussion of precise time and frequency measurement"
<time-nuts at febo.com>
Sent: Monday, March 15, 2010 2:09 PM
Subject: Re: [time-nuts] Choke Ring Pictures
Dr. Clark passed on a tip that I used. Put the funnel in a microwave
oven and run it and see if the funnel warms up. If it warms up, you
do not use it. I do not know what type of plastic the funnel was
made out of; it was white, semi-transparent.
There were also comments about surveying and timing antennas. If you
investigate about every national timing laboratory uses choke ring
antennas. Some enclose the antenna unit and they temperature control
it. They do this for timing stability reasons.
The commercial timing antenna is bullet shaped and is operated
without a ground plane. They are patch antennas. When there is not
ground plane, the antenna picks up best from the overhead and less
towards the horizon. These antennas usually have a lot more gain
(30-50 db vs most normal antennas in the 15-25 db range).
Also in surveying, we cut off the horizon at 15 degrees in software.
A free Army Corp of Engineering manual on GPS Surveying is at
http://140.194.76.129/publications/eng-manuals/em1110-1-1003/toc.htm
The main difference in surveying and timing is in surveying they use
the carrier phase method, were in timing most use a solution derived
from the processing of the coarse acquisition code, in were the
receiver is in a fixed over-determined position . Some timing labs
are using carrier phase method, when they need more resolution.
Brian - KD4FM
****************
warrens wrote:
...
Preliminary results for the Taco Dish GPS antenna as an indoor
antenna are looking good.
Certainly worth considering if your GPS antenna is stuck indoors,
'Out of the rain in the living room'.
I find it best to rise it up near the ceiling such as on an upper
shelf with nothing above it.
It would be hard to tell the difference between the GPSDO
performance obtained from this or the Best outdoor antenna if using
a Tbolt set to the standard default settings.
Picture attached
ws
**************
_______________________________________________
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 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 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 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 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 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.
