On Jun 19, 11:20*am, Robert Clark wrote:
On Jun 16, 6:57 pm, Robert Clark wrote:


From: (Robert Clark)
Date: 23 May 2001 11:15:06 -0700
Subject: Will amateur radio astronomers be the first to directly
detect extrasolar planets?
Newsgroups: rec.radio.amateur.space, rec.radio.amateur.antenna,
sci.astro, sci.astro.seti, sci.space.policyhttp://groups.google.com/group/sci.astro.seti/browse_frm/thread/c0018...

*The long wavelengths should make the requirements for accurate
distance information and timing synchrony between the separate
detectors easy to manage even for amateur systems.

Not easy, not for the precision required.
You need not only precise time (straightforward), but also precise
location (not so straightforward)

You're interested in roughly 20MHz, as I recall. Wavelength of 15
meters. In time, about 45 nanoseconds.

Let's start with a real relaxed requirement, comparable to the mirror
flatness for a telescope of lambda/14. That means a time knowledge
of about 3 ns and a position knowledge of 1 meter, in absolute terms.

Typical GPS receivers that have a 1pps output are good to about 20-30
nanoseconds. Using that to discipline a quartz oscillator, you can do
a bit better, but it's non trivial to get to the 1-2 ns range.
Remember, you're also planning on integrating over time, so you have
to hold that tolerance for a long time.

It would be difficult to determine your position to an absolute
accuracy of 1 meter, much less the phase center of the antenna (which
will change as a function of the angle of incidence, quite
substantially, unless you're putting those dipoles up 100s of feet in
the air.


Using this method
might make the detection achievable even if the power or transmitting
antenna size requirements are not practical for a low cost, low weight
lander *on the Moon for an individual detector on Earth.
*The recent achievement of real-time very long baseline interferometry
should make it possible to integrate these separate detector signals
in real-time as well:

snip

You need to go beyond looking at press releases from radio
astronomers.


*However, instead of using the satellite TV dishes, we could use
individual dipole antennas attached to each house. You would need to
communicate high data rates for the signals detected so you would need
broadband internet access for this.
*These dipole antennas as per the Radio JOVE project are just simple
vertical wires so could be attached to the house when the installer is
connecting the wiring for the broadband. Possibly you could use the
same external wiring as for the broadband but that might cause
interference with the internet signals.

Radio Jove uses a pair of horizontal dipoles connected together to
create a single narrower lobe pointing up.



*As shown on the Radio JOVE page the receivers for these dipole
antennas are quite simple so would contribute minimally to the cost of
installation.

Who's paying, and how minimal? I don't think so.



You do need accurate positional determination and timing
synchrony for each receiving system to do the very long baseline
interferometry, but at these decametric wavelengths this would be easy
to do with GPS receivers carried by the installers.

No they can't. You need position accuracy of sub-1 meter accuracy,
and that isn't achievable by simple handheld devices, like your Garmin
E-trex, etc. A surveyor using a survey GPS system can get there,
although absolute position (relative to, say, the center of the earth,
or some standard datum) to 1 meter would be very challenging.

There's also the not so little problem of tidal bulge. Your position
changes in absolute (relative to a stellar reference) terms several
tens of cm. On top of that, tectonic plate movement is on the order of
several cm/year, which is in the same general ballpark as your
accuracy requirement.

To do the kind of large area combining you're contemplating requires
geodetic quality surveying or some form of in-situ calibration using
known sources (which the folks doing LOFAR and SKA have thought
about). When DSN does accurate interferometric measurements of deep
space probes (a process called Delta DOR) they use a "common view"
quasar as a timing reference, because the Hydrogen maser normally used
for VLBI kinds of things isn't good enough.



Over time you
could keep the systems in synchrony by timing stamps accessed over the
internet.

NTP over the internet is only good to tens of milliseconds. You need
nanosecond precision.

*New broadband subscribers would automatically get the dipole
antennas. At the rate of increase of broadband subscribers, it would
only take 3 months to reach 10 million separate dipoles. If each
installer when setting up a new system, also retrofitted an another
existing broadband system, then you could reach the full coverage of
all the broadband subscribers dipoles in 6 years.


Let's see, leaving aside the surveying and time synchronization
problems, in economic terms this is a non-starter. Say it costs $100
for each "station"... that's a billion dollars for your 10 million
stations. And $100 is a very, very low cost estimate, because
installer time isn't free (probably about $25/hr with all benefits,
insurance, equipment, added in).
BTW, if you really want to do something like this, think in terms of
an addon to a cell site. They already have to have nanosecond
precision timing and surveys in order to do E-911 position
trilateration.



*The number of world-wide broadband subscribers will be 500 million by
2010. At current growth rates it would be 900 million within the 6
years it took to equip each broadband subscriber system with one of
the antenna dipoles. This is nearly two orders of magnitude better
sensitivity than a 10 million dipole system. You could detect out to
100 light-years, opening up many more stars to the possibility of
detection.

I think the 100 billion dollars could be better spent in other ways,
if looking for planets is your goal. Check out Terrestrial Planet
Finder (TPF) for one approach.



* * * Bob Clark