Bob Camp wrote:

Putting it on a constantly rotating platform would work too. I actually
made a comment about moving the antenna as an alternative to the use of
multiple antennas. You can use the velocity vector in correlation with
the local rotation indication to transform the local reference frame to
that of the direction to north. Just swinging it around is pretty
useless unless you bring that in correlation to that local reference
frame. Also, you would like a fairly stable swingrate so that
correlation becomes easier. Also, having a higher rate receiver, such as
10 or 20 Hz would also be nice. The PPS is a suitable reference pulse to
correlate angle-measurements with that of the position given.

Cheers,
Magnus

Not so. I'm very familiar with laying in accurate North lines for gyro
testing. To get anything close to accurate (1 arc second or better) takes
many hours of stellar observation with a Wild T-3 class instrument.

-John

===============

When a star tracker is used as a stellar compass in effect takes
simultaneous fixes on several stars and the better ones are capable of
an rms error of a few arc seconds, largely limited by atmospheric
instability.
These are usually used for determining space vehicle attitude, in which
case the instability due to seeing is much smaller than when immersed in
the atmosphere.
They have been used as relatively inexpensive position encoders for
pointing a telescope to within a few arc seconds.
Pattern recognition techniques are used to identify stars in a
relatively wide field (a few degrees) slightly defocused image (improves
centroiding accuracy).

Star tracker for Clementine mission:
http://www.osti.gov/bridge/servlets/purl/86977-ULkJcR/webviewable/86977.pdf

A more accurate version:
http://www.newworldt.com/download/DTU/microASC%20Summary.pdf

telescope pointing application:
http://www.ifa.hawaii.edu/faculty/pickles/AJP/spie3351.07.pdf
http://www.ifa.hawaii.edu/users/pickles/AJP/AMOS2003_v4.pdf

Measuring the position (altitude + azimuth) of one star at a time using
a theodolite is not the most efficient method of determining the
direction of the meridian.
Its far better to measure the position of several stars at once as in
the stellar compass, however a longer focal length camera lens than
usually used in a star tracker is desrable for increased accuracy.

Bruce

J. Forster wrote:

I once went through the exercise of finding North from a station at
home.
A gps fix would give me +- - 1metre northing and  + - 1 metre
easting, another
fix 1000m away would be the same, a lousy base for an accurate azimuth.
Using an ephemeris for constants, a theodolite fix on a circumpolar star
taken at 6 hour intervals starts to get a value with the limits being
the
usual limits of observing stars but less the refraction errors.
I recorded the data with respect to a Referred Object, ( a street
light 4 km away)
Several times I repeated this exercise ( on other nights) and I was
disssapointed by the precision.
Now my theodolite is a Wild T1, and the method does not lend itself
to getting the maximum
precision from the instrument, but I was deeply impressed by the
difficulty of getting
an accurate measure of true North.
cheers, Neville Michie

On 22/11/2009, at 1:13 PM, J. Forster wrote:

Am Saturday 21 November 2009 20:32:11 schrieb J. Forster:

Just a wild guess: The GPS receiver also knows its location, and magnetic
deviation is known to some degree in its variation over earth's surface. So,
why not introduce a map of magnetic deviation that basically tells you: at
location Lat; Long, your magnetic north points at 357 degrees instead of 360.
Bingo, you're done. That would be especially useful in the pole region, where
deviation becomes large, as the accuracy of magnetic sensors probably is in
the order of a couple degrees.

Sounds reasonable, doesn't it?

HTH,
Florian

Hi Florian,

That is basicly the way it is done. The "source" is a magnetic model - a
"formula" with lots of coefficients, you input your position and the
output is an approximation of your magnetic declination at that position.
It can be implemented by precomputing a lockup table (map) with the
required bin size, or store coefficients and do the calculation in the
receiver.

This is actually an exact analogy with the computation of Mean Sea Level
height. The GPS receiver can only measure ellipsoid height, but many
receivers will output a MSL heigt, by computing or search in a precomputed
"geoid separation"-map.

--

Björn

Stanley Reynolds wrote:

As expected:

Antenna

Type Active dual-loop eLoran H-field antenna with GPS patch antenna

dual-loop H-field Loran antenna is expected. H-field antennas has a zero
in the plane of the loop. To compensate another antenna is mounted with
90 degrees difference. Now, this allows the use of the signal from both
antennas to measure the heading towards each of the LORAN stations,
which together with the position can be converted to heading towards North.

If the GPS receiver had two antennas with similar directional
characteristics, it would too be able to do similar tricks.

Cheers,
Magnus

Yes, found this patent :

http://www.freepatentsonline.com/y2008/0186232.html

"A method and radio navigation system compass apparatus for determining true north or azimuth or orientation of a vehicle or the like by the use of integrated Loran and satellite radio navigation receivers employing crossed-loop H-field antennas for the Loran reception, or the use of at least three Loran type transmitter, or two Loran type transmitter and a synchronized clock for determining both position and azimuth. "

synergy, the whole is greater than the sum of the individual parts

Both GPS and loran depend on the phase/frequency/time accuracy of their source, but I wonder if other signals not as accurate but maybe short-term stable could be used to improve combined stability such as TV and radio transmitters. With such a large source of possible transmitters comparison of them to GPS and loran to pick a few good ones to add both hold-over from gps and or loran loss. For example a CDMA cell site that is dependent on GPS would slowly deteriorate if GPS was lost, but a large number of CDMA cell sites would continue to work if they could be synced to another source.

Besides the hold over capability I would be interested in ways to combine as many independent sources to improve accuracy like the use of more accurate frequency sources to improve GPS accuracy:
http://plan.geomatics.ucalgary.ca/papers/ENC06_Paper74_WatsonetAl_v3Web.pdf

Or maybe a time-nut antidote would be better ;-)

Stanley

----- Original Message ----
From: Magnus Danielson magnus@rubidium.dyndns.org
To: Discussion of precise time and frequency measurement time-nuts@febo.com
Sent: Sun, November 22, 2009 11:18:58 AM
Subject: Re: [time-nuts] OT - GPS and North

Stanley Reynolds wrote:

As expected:

Antenna

Type Active dual-loop eLoran H-field antenna with GPS patch antenna

dual-loop H-field Loran antenna is expected. H-field antennas has a zero in the plane of the loop. To compensate another antenna is mounted with 90 degrees difference. Now, this allows the use of the signal from both antennas to measure the heading towards each of the LORAN stations, which together with the position can be converted to heading towards North.

If the GPS receiver had two antennas with similar directional characteristics, it would too be able to do similar tricks.

Cheers,
Magnus

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.

Stanley Reynolds wrote:

Traditional Directional Finding techniquest combined with positional
information from the Loran-C is a good mix. Question is if the invention
is that big that a good patent would hold.

You can use either the timing or direction or both from many different
sources in combination with that of a number of different satellite
systems in a combined pseudo-range and orientation input for a position
and heading analysis. You can weigth in timing from several VLF sources
in combination with the pseudo-ranges of the sats. The problem with VLF
sources is the shift in delay due to shift in atmosphere, so their
contribution would be limited unless fairly nearby.

Maybe. :)

Cheers,
Magnus