Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.naic.edu/~phil/hardware/12meter/danasFiberTest.txt Дата изменения: Fri Feb 4 15:55:28 2011 Дата индексирования: Fri Feb 11 16:13:30 2011 Кодировка: Поисковые слова: equinox

EXPLANATION OF DANA'S FIBER TESTS ON THE 12M TELESCOPE...


MOTIVATION:

For the VLBI observations using the 12m telescope it is
desirable to send a clock signal to the feed cone which
has a constant delay relative to the signal launched
from our hydrogen maser. But we know full well that this
is not directly possible- the delay in the optical fiber
carrying this signal will change both with temperature
and to some degree with fiber bending as the telescope
tracks objects or moves from one object to another.

The next best thing is to measure the one-way delay change
and report that measurement result along with data from
the telescope at frequent intervals so the delay changes
can be compensated out. But all we can really do is to
measure round trip delay changes and hope to calculate
one-way changes from this. This is pretty simple if we
can find a matched pair, otherwise it gets harder.

So we would like to identify at least one pair of single
mode fibers in our bundle, whose delay changes due to
antenna tracking etc are pretty well matched. If we
can find such a pair, then we're in business just that
simply. But this may not happen, in which case life
gets tougher. It may then prove necesary to measure
individual fiber delays and use a more complicated
calculation to figure the uplink delay changes when
the two fibers in final use are not matched.

This is the purpose of the temporary system that I'm
presently checking out in the RX Lab.


SYSTEM DESCRIPTION:

This system uplinks a signal on one fiber, then turns it
around and splits it into four separate signals which are
donwlinked on any desired test group of four fibers. The
idea is that if the returned signals on any two fibers
show adequately matched delay variations, we've found
what we need. Otherwise I'll need to do a somewhat more
complicated set of measurements to establish independent
values for pairs of uplink and downlink fibers.


HW DESCRIPTION:

The main elements of the system are two big aluminum
plates full of RF components and with temperature
stabilizing devices. The massive plates are used to
provide lots of thermal inertia to make it easier to
avoid rapid temperature changes, which could affect
the measurement electronics and masquerade as delay
changes.

The slightly smaller plate is at the 12m antenna, as
close as is practical to the feeds. The larger plate
will be down in the operations building, near where
the fiber bundle is brought in.


12m SITE NEEDS:

The "hilltop" plate needs to be in or near the feed
cone, above the joints about which tracking motions
occur so that the fiber bundle to this plate is
exposed to the full normal gamut of bending, twisting,
temperature, etc.

This plate will be accompanied by several small power
supplies (or one larger one if preferred), plus two
small electronic units for the temperature regulation
system. The plate and all this other stuff needs to
be well-protected from the weather.

The plate will need easy access because a full search
of the fibers in the bundle will require changing
and/or swapping of fiber connections at least several
times (possibly many times).

The antenna has a reinforced flat plate at the "bottom"
of the movable part, currently with a large coverable
hatch and a smaller oblong hole for cables. I'm
thinking that mounting my stuff on this plate might
suit my needs best, assuming that it's permissible for
us to drill mounting holes in this plate.

The antenna part of the system will need 120VAC
single-phase power, in the neighborhood of 100-200
watts, from a UPS capable of keeping the system alive
in the face of the usual interruptions that occur at
the AO.

It will also have five optical fibers connected at any
given time, although the entire bundle of single-mode
fibers must be available at the plate (terminated in
FC/APC connectors).

Additionally I should really have RS232 access to the
temperature controller for monitoring and tweaking
the controller parameters. It would be sufficient if
this were just brought down into the pedestal, but a
remote link to the operations bldg would be really
nice. The software for doing this expects to be
driving an RS232 port on its host.


OPERATIONS BUILDING NEEDS:

Here I just need space, AC power, access to 10 MHz
derived from our hydrogen maser, and access to the
single-mode fiber bundle (with FC/APC terminations).

The space should be large enough to comfortably
accommodate the big "downstairs" plate plus a small
CW synthesizer and associated power supply, a small
data acquisition device, two small electronics
packages for the downstairs temperature controller,
and a laptop PC. It would be best if the space were
air conditioned, but I could probably live without if
needed.

AC power requirements will also be modest, probably
under 200 watts. Here especially it is essential
that the power come from a reliable UPS.


THE FUTURE:

Assuming that all this works out, the "hilltop" plate
will be replaced with a similar but perhaps somewhat
smaller plate. This new plate will contain just two
fiber link components (instead of the current five)
for the delay monitoring task, plus possibly the two
fiber link transmitters for bringing RF down from the
feeds.

The new hilltop plate might also include the VLBI phase
CAL pulse generator, which we will be purchasing from
an outside supplier.