Now here's a new idea

http://www.nofs.navy.mil/projects/npoi/

Vacuum pipes suggest a laser interferometer, where I wanted to use wires.

"John Schutkeker" wrote in message
. ..
(Greg Hennessy) wrote in news:d17vv2$in5$1
@tantalus.no-ip.org:

If someone can repeat the question (I only see replies, not the head
article) I can try to answer the question. I am an employee of USNO.

I wanted to build an optical imaging interferometer that worked over a
phone line.

Why didn't you?

In article ,
John Schutkeker wrote:
If someone can repeat the question (I only see replies, not the head
article) I can try to answer the question. I am an employee of USNO.

I wanted to build an optical imaging interferometer that worked over a
phone line.

How were you planning on downconverting the terahertz frequency
optical frequencies to a bandwidth a phone line could handle?

John Schutkeker wrote:
(Greg Hennessy) wrote in news:d17vv2$in5$1
@tantalus.no-ip.org:

If someone can repeat the question (I only see replies, not the head
article) I can try to answer the question. I am an employee of USNO.

I wanted to build an optical imaging interferometer that worked over a
phone line.

Why? What makes you think you would *ever* see any fringes?

BTW A single interferometer pair cannot form an image. So your claimed
objective is impossible or an oxymoron. You need a several different
baselines and a range of PA angles to stand any chance of forming even a
primitive image from interferometer data. Look up aperture synthesis
radio telescopes for more information at much easier wavelengths.

State of the art research optical interferometers like COAST use
relatively modest path lengths and devious path compensation measures in
climate controlled bunkers. I am unsure what the longest optical
interferometry baseline to see useful fringes is to date but I would be
surprised if it was more than 500m (the longest working optical baseline
I know of is 330m at CHARA).

ADS abstracts has some of the technical details online. Try the JPL
OLBIN coordinating site for an explanation of what you are up against:

http://olbin.jpl.nasa.gov/intro/

Regards,
Martin Brown

"R.Lewis" wrote in
:

I wanted to build an optical imaging interferometer that worked over
a phone line.

Why didn't you?

I'm already a two or three years behind schedule on my main project, and
since I don't have my PhD yet, I'd never be able to get anybody to take me
seriously in a grant application. I'm still in the process of thinking it
through, to see if it makes sense to try, and I'll give it a priority bump
if I can get a paper out of my proto-thesis.

Besides, I'm just a humble mathematician, and I really don't know as much
about hardware as I'd like.

(Greg Hennessy) wrote in news:d1d401$7ns$1
@tantalus.no-ip.org:

In article ,
John Schutkeker wrote:
If someone can repeat the question (I only see replies, not the head
article) I can try to answer the question. I am an employee of USNO.

I wanted to build an optical imaging interferometer that worked over a
phone line.

How were you planning on downconverting the terahertz frequency
optical frequencies to a bandwidth a phone line could handle?

I'm not a heterodyne savant, but somebody told me that lasers might do it,
although I'm still open to suggestions. Perhaps you could have a staged
system, where you step it down once or twice by several dB using different
methods. HeNe's would be great, because they're also dirt cheap.

I'm willing to sacrifice bandwidth for twisted pair (or shielded twisted
pair) because the heart of the idea is to gain distance (which translates
to aperture size) cheaply.

Martin Brown wrote in news:d1e4mt$tfj
:

Why? What makes you think you would *ever* see any fringes?

Twisted pair is radially symmetrical. If that's too demanding, I can
relax the constraints to shielded-twisted pair or coax. However, I'm
following this idea on someone else's word that fiber optic can do it.
Was that guy telling me the truth?

BTW A single interferometer pair cannot form an image. So your claimed
objective is impossible or an oxymoron. You need a several different
baselines and a range of PA angles to stand any chance of forming a
primitive image from interferometer data.

If each baseline is cheap enough, you can afford to install several
detectors.

State of the art research optical interferometers like COAST use
relatively modest path lengths, devious path compensation measures and
climate controlled bunkers.

I'm gambling that by the time I finish this, an enabling technology for
phase preserving delay lines will emerge. Hopefully somebody smart is
working on that, but if they're not, then I have yet another hard
problem to work on.

Climate controlled signal pipes don't seem like too much of a reach.

I am unsure what the longest optical interferometry
baseline to see useful fringes is to date but I would be
surprised if it was more than 500m (the longest working optical
baseline I know of is 330m at CHARA).

Let me know if you can find any hard numbers about this. For proof of
concept, we just need to match the state of the art.

ADS abstracts has some of the technical details online. Try the JPL
OLBIN coordinating site for an explanation of what you are up against:

I guess my homework this weekend is to read up on ADS, OLBIN, COAST and
aperture synthesis.

John Schutkeker wrote in
:

http://www.nofs.navy.mil/projects/npoi/

Vacuum pipes suggest a laser interferometer, where I wanted to use wires.

These guys just have beam conduits.

In article 2,
John Schutkeker wrote:
I wanted to build an optical imaging interferometer that worked over a
phone line.

How were you planning on downconverting the terahertz frequency
optical frequencies to a bandwidth a phone line could handle?

I'm not a heterodyne savant, but somebody told me that lasers might do it,
although I'm still open to suggestions.

You apparently missed my point, which is no one knows how to
downconvert an optical frequency to something lower.

Radio interferometers such as the VLA have detectors that make use of
the wave nature of the radiation, so they mix with with a local
oscillator, convert to voltages, and then combine the signals in a
correlator.

Optical interferometers have vaccuum delay lines, mix the light beams
physically, and then have APD or CCD detectors to measure the fringes.

The reason I asked you how you were planning on downconverting the
signal was a hint that the technology of doing what you asked for
doesn't exist.

I'm willing to sacrifice bandwidth for twisted pair (or shielded twisted
pair) because the heart of the idea is to gain distance (which translates
to aperture size) cheaply.

What is the bandwidth you think you can get out of twisted pair? If
you express that as a fraction bandwidth of an optical signal (a few
hundred terahertz) what sort of fractional bandwidths are you seeing?
If too low you'd need the Keck telescope to measure Sirius.

In article 2,
John Schutkeker wrote:
I'm gambling that by the time I finish this, an enabling technology for
phase preserving delay lines will emerge.

I hope you aren't putting much money on that bet.