Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.adass.org/adass/proceedings/adass96/reprints/gilesa.pdf Дата изменения: Wed Jan 14 23:26:34 1998 Дата индексирования: Tue Oct 2 11:42:48 2012 Кодировка: Поисковые слова: семинары

Astronomical Data Analysis Software and Systems VI ASP Conference Series, Vol. 125, 1997 Gareth Hunt and H. E. Payne, eds.

Real Time Science Displays for the Proportional Counter Array Experiment on the Rossi X-ray Timing Explorer
A. B. Giles1 Code 662, NASA Goddard Space Flight Center, Greenbelt, MD 20771 E-mail: barry@rosserv.gsfc.nasa.gov Abstract. The Rossi X-ray Timing Explorer (RXTE ) spacecraft contains a large Prop ortional Counter Array (PCA) exp eriment which produces high count rates for many X-ray sources. Telemetry from RXTE is returned via the NASA Tracking and Data Relay Satellite System (TDRSS) which provides a stream of nearly continuous real time data packets. This allows opp ortunity for some serious real time interpretation and decision making by the exp eriment controller, duty scientist, and Guest Observer (GO), if present. The GOs also have the option of arranging for the remote display of programs at their home institution. This pap er briefly describ es the available Science Monitoring subsystem display options.

1.

Introduction

The PCA is one of three exp eriments on RXTE , and was develop ed at the Goddard Space Flight Center (GSFC). The RXTE exp eriment software effort was widely distributed, and the Instrument Teams (IT) were contracted to deliver much supp orting code for their hardware exp eriment. During 1992, it was decided to use C++ for all of the software associated with the Science Op erations Facility (SOF), but this decision was revised in mid-1993. At this time the Guest Observer Facility (GOF) elected to stay in line with the well established practices of the Office of Guest Investigator Programs (OGIP) and High Energy Astrophysics Science Archive Research Center (HEASARC) activities within NASA at GSFC. The SOF was committed to using C++ and continued with this approach. The SOF build deliveries were driven by the inexorable requirement to keep up with the spacecraft's aggressive schedule, and the need to supp ort various tests and mission simulations. The GOF development was under far less pre-launch pressure. Since almost all GOF code is written in fortran using FITS and FTOOLS, there could b e no commonality with the SOF ob ject-oriented C++ environment. This fundamental decision meant that the original requirement for an integrated system with extensive analysis capability in the SOF, was no longer feasible. Resources did not allow the duplication of the functionality of the analysis tools that the GOF was required to produce.

1

also Universities Space Research Asso ciation

389

© Copyright 1997 Astronomical Society of the Pacific. All rights reserved.


390

Giles

The PCA C++ code produced for the SOF was develop ed using Ob ject CenterTM from CenterLine Software Inc. The GUI used is TAE+, which is a commercially available NASA product. Individual graphs were produced using the Athena Tools plotter widget set. Code was delivered to the following SOF subsystems: Command Generation (CG), Mission Monitoring (MM), Health and Safety (HS), and Science Monitoring (SM). In this pap er we discuss only the SM subsystem contributions. 2. Design Requirements

The PCA exp eriment contains five similar detectors, each of which produces identical housekeeping data packets for the HS displays. These packets contain many detector parameters and some X-ray rates, but no sp ectral information. Most science data packets are generated by the Electronic Data System (EDS), which is provided by the Massachusetts Institute of Technology (MIT). This sophisticated data selection and compression system allows many pre-programmed data modes to b e run simultaneously in the six Event Analyzers (EAs) devoted to the PCA. Extensive details on the PCA/EDS combination can b e found in the RXTE NRA (1995). Further details of the RXTE spacecraft and mission can b e found in Swank et al. (1994) and Giles et al. (1995). The requirements for the SM displays are summarized in Table 1, in approximate order of increasing scientific interest.

Table 1. PCA Science Monitoring Requirements. Basic things to examine and look for in real time. Do sp ectra look typical? Are they free of spikes, gaps and noise? Are the sp ectra from all Xe layers in all detectors similar (total of 30)? Are all the internal calibration sp ectra normal? Has the source b een detected? How does the source intensity vary with time? What sort of sp ectral shap e does the X-ray source have? How does the sp ectrum vary with time? How do the source intensity and sp ectrum translate into telemetry loading? What is the source hardness ratio, and how does it vary with time? Are "slow" time scale p eriodic or ap eriodic features visible? Are "faster" time scale features visible in the on-b oard EDS modes?

Given the practicalities following the SOF/GOF split, the PCA team chose to emphasize the following asp ects in SOF displays: real time or near real time graphical displays, visual impact and clarity, multiple options within a predefined set of choices, limited analysis capability but exp ort of data to external tools like IDL, and supp ort for only a few of the many EDS modes, concentrating on Standard Modes 1 and 2. Rememb er that detailed analysis is not intended with this real time system. If GOs are present at the SOF during their observations and need to do higher level tasks, such as background subtraction


Real Time Science Displays for the RXTE PCA Exp eriment

391

or sp ectral fitting in near real time, the GOF's Fits Formatter (XFF) system can b e used. A SOF copy of XFF can b e run to produce temp orary FITS files on an estimated 15­30 minute time frame. The precise time scale dep ends on the data flow from the NASA PACOR system into GSFC and "fast" FITS files made in this way may b e very incomplete. Normally, a 24 hour p eriod elapses b efore all late packets are assumed to have arrived. Once the FITS files are created, they can b e studied using the GOF FTOOLS. 3. Science Monitoring Displays

The set of display programs develop ed to address these monitoring requirements is detailed in Table 2. These displays can all b e run individually or, more commonly, selected from a main GUI interface.

Table 2. Main group STD Mode 1 Temp oral Science

PCA Science Monitoring Displays. Description

Sub Group

STD Mode 2 Sp ectral Science

Others

Cal. Sp ectra Raw 256 bin sp ectra every 128 s Light Curve (LC) 8 LC's. Each 1024в0.125 s p oints Power Sp ectra (PS) Power sp ectrum of STD Mode 1 LC PS History 2-D color coded p ower, freq. vs. time Layer LC's Select keV range from STD Mode 2 Energy LC's Select 4 keV ranges from STD Mode 2 Sp ectra Xenon and Propane sp ectra every 16 s. Diagnostic Rates 29 rates (в 5 PCU's) every 16 s. Base Summation Derived from STD Mode 2, Xe signals added raw or with gain/offsets applied Recommend Base summation mapp ed to the 6 energy bands of the GOF RECOMMD program Colors vs. Time 3 definable keV bands using Base Mode Sp ectral History 2-D color coded intensity, keV vs. time EDS FFT On-b oard EDS FFT sp ectra mode EDS Delta Time On-b oard EDS Delta Time Binned Mode Event Selection A tool to examine basic Event Mode Slew Detection A derivative of Event Selection

The main GUI allows multiple instances of programs to b e started, e.g., it is common to run four light curve options at the same time--0.125, 1, 8, and 16 s temp oral resolution. These displays then span 128 s, 1024 s, 2.28 hours, and 18.2 hours resp ectively. Standard Modes 1 (mainly temp oral) and 2 (mainly sp ectral) are always present. These displays have too many features and options for a detailed discussion (see User Guide, Rhee 1995; Design Guide, Giles 1995). A document containing sample screens for the PCA real time SM subsystem is


392

Giles

referenced on the WWW RXTE SOF page. New options are carefully tested b efore transfer to the SOF configuration controlled environment. 4. Op erations

RXTE was a fixed price program that was completed within budget and on time with resp ect to goals set 4 years prior to the planned launch. RXTE was launched on 30th Decemb er 1995 and has already produced a vast amount of high quality timing data on a wide variety of X-ray sources. The various PCA monitoring displays have proved very effective in supp orting the mission and have allowed the sort of interactive decision making that was hop ed for. The duty scientist and exp eriment controllers monitor the observation in progress, to try and ensure that it is proceeding as planned, and that modifications to the observing modes are not required. RXTE can also b e slewed rapidly to p oint at new Targets Of Opp ortunity. The PCA instrument team has p ermanent access to the SOF data flow, to monitor their exp eriment using the same display programs. Guest observers need not b e present at GSFC--they routinely monitor their observation in real time from their home institution using remote displays of the same suite of programs running in the SOF. RXTE has already made many public observations and this is exp ected to continue in the future. The SOF are in the process of providing a mechanism for real time public data to b e seen by anyone in the world via a WWW interface to the PCA display programs. Acknowledgments. much comment and input The coding for the PCA Vikram Savkoor, Hwa-ja and Arun Simha who wer contract for NASA. References RXTE 1995, 1st RXTE NASA Research Announcement, January Giles, A. B. 1995, PCA Science Monitoring, Design Concepts, Screen Functions, SOC User Interface, Version 6.1, Decemb er Giles, A. B., Jahoda, K., Swank, J. H., & Zhang, W. 1995, Publ. Astron. Soc. Aust., 12, 219 Rhee, H. 1995, PCA Science Monitoring User's Guide, Version 4.1.1, August Swank, J. H., et al. 1994, in NATO ASI Series C, vol. 450, The Lives of the Neutron Stars, eds. M. A. Alpar, U. Kiziloglu, & J. van Paradijs, (Dordrecht: Kluwer), 525 The design of these software systems has b enefited from from my fellow scientists in the PCA Instrument Team. SOF subsystems was p erformed in varying degrees by Rhee, Ramesh Ponneganti, David Hon, Aileen Barry, e all with the Hughes STX Corp oration working under