non-directional navigation beacon location techique question
 

I'm trying to find the location of a low frequency mavigation beacon.
Two techniques come to mind.

1) Using the null of a loop antenna, find a vector to the beacon using
a compass. Note the location where the vector was found on a GPS. Do
this from multiple locations and the beacon is where the lines
intersect. Since there will be error in both the compass (2 degrees
using a Garmin GPS) and some error in finding the null, the result will
be more of a locus of points where the beacon could be located rather
than the beacon itself.

2) Again from various locations, find the direction of the unknown
beacon and a known beacon (about 40 miles apart). Use a turntable of
sorts to find the number of degrees between the known and unknown
beacon. [Note, this will not require a compass, just an angular
measurement) Given the angle difference between the two beacons, a
vector can be drawn from a well known (via gps) location. Again, where
these vectors intersect is the location of the unknown beacon.

Throw into the mix one more reference. It should be possible to find a
spot where both beacons null at the same time. The the unknown beacon
is along a line between your location and the known beacon. This would
be a line that should be quite accurate. This should be doable from two
locations, i.e one where the known beacon is the closer of the two, and
the opposite situation.

Comments, besides get a life?

wrote:

I'm trying to find the location of a low frequency mavigation beacon.
Two techniques come to mind.

1) Using the null of a loop antenna, find a vector to the beacon using
a compass. Note the location where the vector was found on a GPS. Do
this from multiple locations and the beacon is where the lines
intersect. Since there will be error in both the compass (2 degrees
using a Garmin GPS) and some error in finding the null, the result will
be more of a locus of points where the beacon could be located rather
than the beacon itself.

2) Again from various locations, find the direction of the unknown
beacon and a known beacon (about 40 miles apart). Use a turntable of
sorts to find the number of degrees between the known and unknown
beacon. [Note, this will not require a compass, just an angular
measurement) Given the angle difference between the two beacons, a
vector can be drawn from a well known (via gps) location. Again, where
these vectors intersect is the location of the unknown beacon.

Throw into the mix one more reference. It should be possible to find a
spot where both beacons null at the same time. The the unknown beacon
is along a line between your location and the known beacon. This would
be a line that should be quite accurate. This should be doable from two
locations, i.e one where the known beacon is the closer of the two, and
the opposite situation.

Comments, besides get a life?

The easiest way by far is to look up its location - no point in having a
navigation beacon if you don't publish the location.

Dave

This particular beacon is somewhere around Area 12 on the Nevada Test
Site. It is not published.

There are three unpublished NDB in the general area.
209Khz AEC located at Basecamp (near route 6 east of Warm Springs)
http://www.lazygranch.com/sound/beac...ecamp209_2.wav
278Khz XSD located at the north end of the Tonopah Test Range
www.lazygranch.com/sound/beacons/xsd278.wav
414Khz PYD somewhere near Area 12 on the Nevada Test Site
www.lazygranch.com/sound/beacons/pyd414.wav

Needless to say, these are not FCC licensed.

One more though comes to mind. Would it make sense to use a narrow band
filter (say a CW filter) on the beacon when determining the null since
the S meter will be used to determine the minimum signal strength.

One more though comes to mind. Would it make sense to use a narrow band
filter (say a CW filter) on the beacon when determining the null since
the S meter will be used to determine the minimum signal strength.

I have a Bendix 555 marine RDF, but the null is no where near as good
as using a loop with a Welbrook amp. Further, you really don't know the
bearing any better than a compass reading. That is, you can get the RDF
to point at the source (via a null), but that information needs to be
associated with a compass reading. Essentially, this is method one that
I listed.

On 9 Apr 2005 19:34:13 -0700, wrote:

I'm trying to find the location of a low frequency mavigation beacon.
Two techniques come to mind.

1) Using the null of a loop antenna, find a vector to the beacon using
a compass.

2) Again from various locations, find the direction of the unknown
beacon and a known beacon (about 40 miles apart). Use a turntable of
sorts to find the number of degrees between the known and unknown
beacon.

Throw into the mix one more reference. It should be possible to find a
spot where both beacons null at the same time.

Comments, besides get a life?

Sounds like an interesting project. Why not use all three techniques
together? After all, you are trying to locate something as accurately
as you can - all the data you collect is relevant to solving the
problem.

The poistion polygon that you will determine used to be known as a
'cocked hat' - the smaller it is, the better your determination. There
used to be an expression "I knocked him into a cocked hat', meaning
that you had negated someone's argument. Good luck.

1) Using the null of a loop antenna, find a vector to the beacon using
a compass. Note the location where the vector was found on a GPS. Do
this from multiple locations and the beacon is where the lines
intersect. Since there will be error in both the compass (2 degrees
using a Garmin GPS) and some error in finding the null, the result will
be more of a locus of points where the beacon could be located rather
than the beacon itself.


I have posted a law-of-cosines calculation example several times where
coordinates of an intersection point are determined from two GPS locations
with a bearing from each location...to the intersection point. I can find
one of those in my files and post it later if that is needed.

But it could been done graphically with 'Scratch-Plot' because one of the
newer features of 'Scratch-Plot' is a plot by azimuth and distance. Just
set-up a scale of UTM coordinates, plot each GPS location in UTM
coordinates, plot a line from each location with an azimuth direction and a
dummy-distance so that the lines intersect, mouse-click a point at the
intersection of the lines, and check the coordinates of the mouse-click
point immediately in the help file or later in the plot text file.

Here is a user link to 'Scratch-Plot' :

http://www.kbhscape.com/plot.htm


Also, the difference between a geodetic direction and a UTM grid direction
can be determined with 'Geodetic/UTM-Grid Utility' by entering the latitude
and longitudes and comparing geodetic directions to grid directions.

And here is a user link to 'Geodetic/UTM-Grid Utility' :

http://www.kbhscape.com/gps.htm

I have posted a law-of-cosines calculation example several times where
coordinates of an intersection point are determined from two GPS locations
with a bearing from each location...to the intersection point. I can find
one of those in my files and post it later if that is needed.


Oh, intersection of two directions is fundamentally a law-of-sines
calculation and that is a much easier calculation than a law-of-cosines
calculation. A law-of-cosines calculation is use for the intersection of two
distances...

But it could been done graphically with 'Scratch-Plot' because one of the
newer features of 'Scratch-Plot' is a plot by azimuth and distance. Just
set-up a scale of UTM coordinates, plot each GPS location in UTM
coordinates, plot a line from each location with an azimuth direction and
a dummy-distance so that the lines intersect, mouse-click a point at the
intersection of the lines, and check the coordinates of the mouse-click
point immediately in the help file or later in the plot text file.


Compass directions with declination should also be adjusted with UTM
convergence before they are used with UTM coordinates...

"KBH" wrote in message
.. .
....
I have posted a law-of-cosines calculation example several times where
coordinates of an intersection point are determined from two GPS locations
with a bearing from each location...to the intersection point. I can find
one of those in my files and post it later if that is needed.

The free software FAA software compsys21 will do this and more
http://www.naco.faa.gov/index.asp?xm...online/compsys

"KBH" wrote in message
. ..
The free software FAA software compsys21 will do this and more

That's interesting but for something the size of half-a-state I would more
likely work with UTM coordinates and then use plane survey formulas...

Why use an inaccurate approximation when better software already exits???

In RDF navigation the more "fixes" one has the smaller the
ccccircle of confussion. For a variety of reasons, measurement
erros, propagation anolomies etc, the bearing is seldom perfect.
These imperfections create a difussed "circle" that the calculated
powiton lies witihin. Moer sighting/bearings gives smaller errors,
but using home built equipment one shouldn't expect world class
accuracy.

As a basic check pick a non NDB,make your measurements,
being carefuyll to note the GPS reading for each, and see how
close you come to hitting the transmitters location.

I su pect that yoru antenna will be far from accuarte with some
imbalance. that will be hard to quantify exactly. As a fun project,
that will teach you a lot about navigation go for it, but as a serious
attempt I suspect you will be dispointed.

Even at NDB low fRF frequencies, there are too many things
that will cause transmission path errors. I live about 15 miles from
the Lexington Blue Grass Fild NDB and even with a barrowed
mil grade RDF setup it's apparent bearing changed by several
degrees the month I had the unit. The NDB in Frankfort and Mt.
Steerling "wondered" all over the place. with sudden randon sshifts
up up to +/- 5 degrees. The pilot who loaned me the unit told me that
the closer to the earths surface you got with a LF RDF the more error
you
pikced up. At 5000' the error was very small. This was 25 years ago
and I suepct not much has changed.

This was an aircraft RDF and at that time still was within the
calibration cycle.

Good luck and have fun.

Terry

The free software FAA software compsys21 will do this and more

That's interesting but for something the size of half-a-state I would
more likely work with UTM coordinates and then use plane survey
formulas...

Why use an inaccurate approximation when better software already exits???

You give yourself away with a comment like that...

Projections to rectangular grids are not intended to be approximations but
are intended to be rectangular grids.

Latitudes and longitudes labeled on an atlas are first converted to
rectangular coordinates, plotted as rectangular coordinates, and then
labeled as latitude and longitude. In other words any point (within range)
has both latitude / longitude location and UTM grid location. Simply there
is a conversion between the two.

Okay, UTM and geodetic have different directional orientations. But in
project layout any point is relative to two or more other points. In project
layout there is no such thing as one point relative to only one other point
because that would be a magical creation. The point is that UTM directional
orientation is used with UTM points and that geodetic directional
orientation is used with geodetic points. (For example, the consumer GPS
user can do this by getting their GPS location, getting a GPS location of a
skyscraper or transmission tower that can be seen in the distance, and then
laying out an angle to a required point using their home point location and
the line of sight to the tower. Note the three points and that is project
layout.)

Also, a bearing in UTM is one direction to the point. And that is likely
what is required on a project. A geodetic bearing is simply a beginning
direction to the point and directional corrections are required. That may be
okay for navigation but would very strange in project layout...So a layout
in UTM is a rumb line while a geodetic layout is a great circle. Note that a
point laid out on a rumb line in a UTM grid could after layout be converted
to latitude and longitude. In other words simply meet requirements and
output in any required format.

Of course for higher accuracy, projections to state plane coordinates are
used instead of UTM coordinates. And state plane coordinates are used
extensively in project layout. There are not a bunch of engineers walking
around saying "...why use an inaccurate approximation..." as the project is
simply defined with a rectangular coordinate system.

Finally, one benefit is using rectangular coordinates are that simple
formulas can be used with rectangular coordinates. Someone with an
inexpensive scientific calculator can make on-site calculations that for
instance a construction crew might be waiting on. And of course a $90 HP48
will hold and quickly run all of the plane survey formulas and will fit in
shirt pocket. Furthermore, plane survey formulas can be developed for PC
programs in short periods of time and without access to highly developed
formal sources of information or expertise.

So if the project is defined with rectangular coordinates why have a
software that does not apply to the task ?

"KBH" wrote in message
...
You give yourself away with a comment like that...

With a PhD in mathematics specializing in an area of geometry, I think
you
misjudge me.


The terminology "...in an area of geometry..." is another give-away.

Geometric topology with graduate coursework including differential
geometry--in that you learn the substance of approximating a curved surfaces
with euclidean pieces ("maps"). And yes, I have actually studied the
differences in distances and angles from approximating the earth's surface
as rectangular on the small scale--not in a course but by doing the
calculations.

But that's not the point.

Agree.

Projects are not defined with great-circles because the direction to a
great-circle is just a beginning direction. Navigation on the other hand
seeks the shortest route and thus correction of the course is accepted. So
projection to rectangular coordinates systems is used for projects while
geodetic systems are used for navigation.

As I understand the original problem. it was to triangulate to locate one of
the NV beacons and the software I mention contains menu selection that will
do the needed calculation.

I would argue that rectangular coodrinates are used simply because we like
to think in straight lines, they are easy to construct, and on a small
scale they approximate a geodesic (great circle) on a sphere.

I might add that I have done long range triangulation of VLF stations (the
stations were 1000's of km away). This would have been impossible using
rectangular coordinates. Additonally, originally the FCC HF DF network had
to use a special device to triangulate--it would have been much easier if
they could have just done rectangular geometry. But again, the distances of
the project proposed are much smaller and that level of accurate modeling is
certainly not needed as other errors will be more significant.

Per the original plan, I like his differencing the angles from a known
source as that is better than just trusting the null to be where he things
it is in the antenna pattern. Ideally use a radio where one can turn off
the AGC--it will make things a lot easier. Be sure to make several (lots)
indpendent measurements at each location and average them. If you take
enough (ideally at least 30), you can then construct error bound for the
bearings using simple statistics. And the thing usually forgotten, practice
the technique on known locations (beacons) first to obtain realistic
expectations..

As I understand the original problem. it was to triangulate to locate one
of the NV beacons and the software I mention contains menu selection that
will do the needed calculation.


I don't dislike using the geodetic software to determine the intersection
point but as I said before I personally would work with UTM coordinates and
a simple law-of-sines calculation. The law-of-sines calculation could be
demonstrated in a small paragraph while the ellipsoidal intersection
calculation would be a page of formulas.

The final result of working with UTM is UTM and can always be converted to
latitude and longitude.

Also, with rectangular coordinates the problem could be worked out on graph
paper...or as I said before worked out graphically.

But suppose the compass readings were perfect and the distances involved
were 40 miles...and say that the becon is the size of a 16d nail and that
determination of its location is critcial within two diameters of the becon.
Is the compass pointing a rumb line or is the compass indicating an initial
direction ? In other words with the geodetic software are we seeking the
intersection of two great-circles or the intersection of two rumb lines ?


I would argue that rectangular coodrinates are used simply because we like
to think in straight lines, they are easy to construct, and on a small
scale they approximate a geodesic (great circle) on a sphere.

We do more than think in straight lines. Survey instruments run straight
lines or determine straight baselines. A control survey is points connected
with lines even though points on some curve near the control survey can be
set from the control survey...

The historical requirement of the division of Federal lands calls for the
construction of an East great-circle as points offset from a straight
baseline...In other words the straight baseline is the given and the
great-circle is the construction.

The routes intersect on the map.That is, you have to enter a waypoint
where the lines cross.

Indicentally, I think UTM is much better system, but it is sort of like
Betamax and VHS.

It's probably loaded to increase the electrical length. Here is a
photo:
http://www.lazygranch.com/images/basecamp/ndb1.jpg
I cranked up the contrast, but you can't see the wire.

I've run into state plane as well. I believe they work in feet, which
is one of the differences between state plane and UTM. However, my
knowledge of either is pretty limited.

http://www.lazygranch.com/images/fau...rojfault-3.jpg
I found this marker and was baffled about the coordinate system until
someone on this group pointed out it was in state plane coordinates.

I know of a Jim Gosnell who lives in Florida.You any kin to him? He was
in Vietnam.Which reminds me,I got an email from him yesterday and I need
to reply.About updating an email list.
cuhulin

Hi Miso

Well asking here was the easiest DFing ever for you.
Here are locations as found

On 9 Apr 2005 20:31:52 -0700, in rec.radio.shortwave you wrote:

This particular beacon is somewhere around Area 12 on the Nevada Test
Site. It is not published.

There are three unpublished NDB in the general area.
209Khz AEC located at Basecamp (near route 6 east of Warm Springs)
http://www.lazygranch.com/sound/beac...ecamp209_2.wav
'Base Camp' Tonopah
AWOS-3; Official Location: Warm Springs Actual as listed.
Lat 38.3125
Lon -116.292

278Khz XSD located at the north end of the Tonopah Test Range
www.lazygranch.com/sound/beacons/xsd278.wav
Tonopah Range
Lat 37.8542
Lon -116.792


414Khz PYD somewhere near Area 12 on the Nevada Test Site
www.lazygranch.com/sound/beacons/pyd414.wav

'Area 19' Groom Lake
Lat 37.1875
Lon -115.958

Regarding basecamp, I'd have to say this information is wrong. The NDB
is on the other side of Tybo road. The NDB is at least a mile to the
north east. I don't have a waypoint, but I've been to the general area.

Regarding the Tonopah Test Range, the NDB is at least 2.5 miles to the
north based on the map I acquired.
http://www.lazygranch.com/images/ttr/xsdsectional.gif

Regarding the Area 19 location given , it is actually closer to area
10.

If you have a link to your source, I'd like to look at it anyway.