DEVELOPMENT OF BROADBAND PHASE DISCONTINUITY SENSING
ALGORITHM

1. Gary Chanan, Principal Investigator
Agusti Pinto, Co-Investigator


2. Purpose of test: development project

3. Test number: not applicable

4. Test duration: 10 hours (full night)

5. Brief description of test:

Broadband Phase Discontinuity Sensing (BBPDS) is a variation of the
current (narrowband) PDS algorithm which has a much larger capture
range - as much as two orders of magnitude greater than the current
400 nm range of PDS. [BBPDS bears a similar relation to narrowband
PDS as broadband PCS does to narrowband PCS.] Although the narrow-
band PDS algorithm capture range is large enough for most maintenance
phasing work, it is not adequate for post-segment exchange phasing.
Additionally, if the normal phase errors do exceed the PDS capture
range (as happened recently), the current system is rendered completely
helpless.

A. To begin this test, we will first execute a normal PDS run. This
will establish a baseline for the subsequent tests and will also
demonstrate that the recent problems experienced with normal PDS have
been resolved.

B. We will then execute an additional PDS run with the new software,
which can handle both the narrowband and broadband tests and which
contains some additional improvements to the old code.

C. With the telescope well-phased, we will introduce large piston
errors (~30 microns rms) and attempt to recover the correct phase with
the largest capture range (but least accurate) version of the algorithm.

D. We will acquire a raw spectrum of one of the phasing stars
through the KW filter. This is required for the simulations and
analysis because the filter is too wide to model the transmitted
spectral density as a constant.

E. The previous step will be repeated with smaller piston errors
and a smaller, but higher accuracy, capture range version of the
algorithm. Ideally, the residual piston errors remaining after
this step will be smaller than the capture range of the narrowband
PDS algorithm.

F. If time permits, we will repeat these tests with other starting
configurations.


6. Reasons why night sky is needed:

We have completed several hundred hours of numerical simulations of the
BBPDS algorithm, which enable us to predict its behavior as a function
of piston error and filter parameters (central wavelength and bandwidth).
The simulations include the effects of seeing (in the long exposure limit)
and segment aberrations. A manuscript describing the results of these
numerical experiments is in preparation. Clearly the next step is
to put the BBPDS algorithm to the true test on the actual telescope.


7. Plan for data reduction, etc:

The initial reduction of the BBPDS data is identical to that of the
existing narrowband algorithm, which has been in use for over five years.
Those aspects of the reduction and analysis which are not shared with
the narrowband algorithm have been developed and tested in the course
of the numerical simulations described above. If these experiments
are successful, a paper describing the results will be submitted to
Applied Optics. Furthermore, the scripts which have been developed
by one of us (A.P.) will be turned over to the Observatory so that
the BBPDS algorithm can be used when needed by Keck personnel.


8. Preferred dates:

Because NIRC must be on the telescope, September 13, 2003 is the only
possible night in the current proposal period.

9. Progress report:

These tests were originally scheduled for July 18, 2003. However,
virtually the entire night was lost to fog. [Only the initial
MAlign and determination of the instrument offsets were completed.]