Документ взят из кэша поисковой машины. Адрес
оригинального документа
: http://www.naic.edu/~cima/cima_backends.html
Дата изменения: Thu Mar 1 19:02:27 2012
Дата индексирования: Mon Oct 1 22:43:30 2012
Кодировка:
Поисковые слова: п п п п п п п п п р п р п р п р п р п р п р п р п р п р п р п р п р п р п р п р п р п р п р п р п р п р п р п п р п п р п п р п п р п п р п п р п
|
Backends Available with CIMA
Backends available with CIMA
Different backends are supported by CIMA depending on the selected
receiver and observation type (spectral line or pulsar). CIMA has
traditionally supported two backends: the interim correlator
for spectral line single-pixel receiver observations and the
WAPPs for any kind of observation. The WAPPs behave quite
differently in spectral line and in pulsar mode, with the differences
being so big that the observer has to select which type of
obsercvation already at start-up of CIMA. In spectral line mode, the
data from all WAPPs are collected into a single CIMAFITS file with a
dump rate not faster than 1 second, while in pulsar mode, each WAPP
produces an individual data file written in the raw WAPP pulsar format
which allows for high speed sampling. Also, other backends like the
ASP and machines supported by the 30 MHz input can be run in parallel
with the WAPPs. The GALFA spectrometer has not been supported from
CIMA and to take GALFA data with CIMA, it has thus been necessary to
take data with the WAPPs in parallel.
The newer versions of CIMA have expanded capabilities. The
single-pixel receiver WAPP dual board observing mode was added in
version 2.3 as well as the possibility to run observations with GALFA
without the WAPPs, while version 3.0 adds the new Mock spectrometer
for ALFA observations.
The interim correlator
The interim correlator can only be used for spectral line single-pixel
receiver observations and it can not be used with any of the observing
modes developed for ALFA. Most of the observing modes are using
implementations in the 'datataking' program and once CIMA has set
everything up for an observation, it hands over control to the
'datataking' program (called 'VX-works' in CIMA) and waits until it
has completed the observation. The interim correlator can not be used
in parallel with any other backend. It is the only backend used in the
'Calibration' window in CIMA.
The interim correlator is the only backend that can be run without
CIMA, since most observing modes are implemented and executed by the
'datataking' program which also handles all telescope control and
IF/LO set-ups.
The WAPPs
The WAPPs (Wideband Arecibo Pulsar Processors) are four boxes which
each run two correlator boards capable of handling 100 MHz each. They
can thus be used for up to 800 MHz of bandwidth coverage with single
pixel receivers (in dual board mode) or 100 MHz per beam for ALFA. The
possibility to use the dual board mode for single pixel receivers was
added in CIMA version 2.3 - with older versions of CIMA, the WAPPs can
only handle a maximum of 400 MHz. Originally built for pulsar
observations, the WAPPs were later adapted also for spectral line
observations. The history has left its traces though, since the WAPPs
behave very differenly in pulsar mode and in spectral line mode, and
this big difference is the reason why all CIMA observers have to
specify whether they are going to run spectral line or pulsar
observations already when launching CIMA. Pulsar mode and spectral
line mode both have their own sets of observing modes in CIMA.
In pulsar mode data can be taken with high dump rates (writing up to
25 MB/sec for each WAPP). The file format is simple with a data header
written just before the observation starts followed by data without
any meta data. Dynamic information like telescope positions are thus
not available in the data (except once in the header). The files have
a maximum size of 2 GB and all files needed for a scan are
preallocated before the scan is started. Each WAPP writes the data to
their own disks one file at a time - dual board observations
(including ALFA) thus have the two data streams interleaved. Each scan
produce its own set of files.
In spectral line mode, data is dumped no faster than once a second
(and not slower than once per 15 seconds). The data streams from the
WAPPs are sent to WAPPDATA which merges them and writes the data to a
single CIMAFITS-file in the /share/wappdata area. Data from many scans
can be added to the same file. The file is changed either by a request
from the observer or when the header table gets full.
Starting with CIMA version 3.1, the WAPPs can also be run in
continuous data taking mode for spectral line observations. In this
mode, the WAPPs are kept running and recording data also between the
scans. It is possible to write an entire observation as one continuous
data stream to one file or to create one file per observing loop.
The WAPPs can be run in parallel with the Mock spectrometer for ALFA
observations and with ASP and 30 MHz input machines for single-pixel
pulsar observations.
The Mock spectrometer
The Mock spectrometer is a versatile machine consisting of no less
than 14 boxes which each run two correlator boards capable of handling
172 MHz each. The designed data rate for writing is 50 MB/sec per
box. The Mock spectrometer can so far only be used for ALFA
observations, and the 14 boxes are then divided up in two groups which
apart from a few common parameters can be configured
independently. The two groups can thus work almost as two independent
spectrometers for ALFA. The Mock spectrometer is the third backend
that is completely controlled from CIMA and it was added to CIMA with
version 3.0.
The Mock spectrometer can be used for all observing modes supported by
the WAPPs for ALFA and it can be run in parallel with the WAPPs. Data
taking does not start at exactly the same time when run in parallel
with the WAPPs, since the Mock spectrometer needs more time to start.
The data format is the new CIMAFITS-2 format which works both for slow
spectral line data and high-speed pulsar data. One or more files is
generated for each correlator board and each scan. The maximum file
size is selectable. There is an alternative FITS-format for pulsar
observations called PSRFITS, which soon will be available. Data can
also be output in PDEV-format which was an early engineering format
that is not recommended.
Starting with CIMA version 3.1, the Mock spectrometer can also be run
in continuous data taking mode for spectral line observations. In this
mode, the Mock spectrometer is kept running and recording data also
between the scans. It is possible to write an entire observation as
one continuous data stream to one set of files or to create one set of
files per observing loop.
The GALFA spectrometer
The GALFA spectrometer can not be controlled from CIMA and does not
interact with CIMA at all. It is started up independently and is then
running in a continuous data taking mode. In CIMA version 2.2 and
earlier, it was necessary to also run the WAPPs since they are telling
CIMA when each individual scan is finished and when it is time to move
on. Starting with CIMA version 2.3, there is no longer a need to run
the WAPPs, since there is now a GALFA only option in CIMA. There is
still no communication between CIMA and GALFA, but CIMA is now capable
of self keeping track of time.
Other backends
When the WAPPs are used for single-pixel receiver observations in
pulsar mode, CIMA has support for the ASP and machines supported by
the 30 MHz input. However, they can't be run alone, unless You just
set things up with CIMA and manually wait while they are taking
data. There is no observing mode that can be run without the WAPPs or
the Mock spectrometer.
You can return to the main CIMA page by clicking here.
This page is administered by Prakash Atreya
( patreya (a) naic . edu ) and was last updated on 11 December 2008.