HEXTE FAQ: Frequently Asked Questions

Answers to the frequently asked questions are given below in reverse time order (most recent first). Suggestions and further questions are welcome - please send them to the e-mail address at the end of this page.
  1. What is the HEXTE dead-time fraction and how can I correct for it?

  2. How do I rebin a fully-sampled HEXTE response matrix for a Standard Modes (Archive Spectral Bin) spectrum?

  3. How do the values of the ClstrPosition column in HEXTE FITS files correspond to a HEXTE cluster's actual rocking positions?

  4. What are the "DEFAULT" values in RXTE Cycle 1 & 2 for the HEXTE's lower energy bound, rocking direction and rocking dwell time, and how were these values chosen?

  5. Which is cluster A and which is cluster B on the spacecraft, and how are the phoswich detectors numbered?

  6. How do I know where the HEXTE clusters' background pointings will end up on the sky?

  7. What is the difference between the HEXTE Science Modes E_8us_256_DX1F and E_8us_256_DX0F?

  8. What is the status of the malfunctioning detector in cluster B, and when did the malfunction occur?

  9. What are the features in the HEXTE background spectrum?

  10. How sensitive is the HEXTE for faint source observations?


Q: What is the HEXTE dead-time fraction and how can I correct for it?

See the HEXTE Instrument Description chapter of the NRA, and then read the report on the HEXTE deadtime from the HEXTE team.

Q: How do I rebin a HEXTE response matrix for a Standard Modes (Archive Histogram) spectrum?

Archive Spectral Bin (or Archive Histogram) spectra have a compressed PHA channel binning to save telemetry. They are stored in HEXTE raw FITS files with names beginning ``FS52...'' (Cluster A) or ``FS58...'' (Cluster B). This data product is described in the HEXTE chapter of the NRA as follows:
Archive Spectral Bin mode produces a compressed pulse height histogram (energy spectrum) for each phoswich detector every 16 s. The first 64 PHA channels are combined two-at-a-time into the first 32 archive spectrum bins with 16-bit depth per bin; the second 64 phoswich PHA channels are combined four-at- a-time into the next 16 archive histogram bins with 8-bit depth per bin; and the final 128 phoswich PHA channels are combined eight-at-a-time into the last 16 archive histogram bins with 8-bit depth per bin. This compression causes only a small loss of information; virtually all cosmic x-ray sources in the 10-250 keV range decrease in intensity with energy, while the HEXTE's FWHM spectral resolution (measured in PHA channels) also broadens with energy.
To rebin a HEXTE response matrix down to the appropriate channel sampling for Archive Spectral Bin data. one merely has to apply the standard Ftool rbnrmf to the "full" response matrix for the appropriate detector or HEXTE cluster as input, with the compressed binning specified in the file hexte_ArchSpectralBin.txt provided here. For example, to rebin the response matrix for HEXTE cluster A, just enter the following on the command line:
rbnrmf infile=hexte_pwa.rsp outfile=hexte_pwa_AS.rsp binfile=hexte_ArchSpectralBin.txt
This will result in a rebinned response matrix hexte_pwa_AS.rsp with channel binning appropriate for use with Archive Spectral Bin data.

Q: How do the values of the ClstrPosition column in HEXTE FITS files correspond to a HEXTE cluster's actual rocking positions?

A: The table below lists the possible HEXTE cluster positions and the the corresponding values of the ClstrPosition column in FITS files for the Standard Modes (Archive) data, and for the Science Modes.

Note that are two ``0.0 degree positions'' since the HEXTE clusters are moved by a cam mounted on a rotating shaft: one of the ``zero'' positions is between the +1.5 and -1.5 degree marks on this cam, and the other is diametrically opposite on the shaft between the +3.0 and -3.0 degree marks. These ``zero'' positions are essentially identical for data analysis purposes, however.

Position
(degrees)
Standard Modes
(A: FS52.., B: FS58..)
Science Modes
(A: FS50.., B: FS56..)
0 (1.5) 0 1 or 65
+1.5 1 2 or 66
+3.0 2 4 or 68
0 (3.0) 3 8 or 72
-3.0 4 16 or 80
-1.5 5 32 or 96
Values of the ClstrPosition column in HEXTE FITS files

In the HEXTE Science Modes data, two values of ClstrPosition are possible for a given cluster position. The second value listed is greater by 64, and is generated in telemetry when the cluster's motor is energized (but not necessarily moving).

This numbering scheme appears unnecessesarily complex, admittedly, but has its roots in the early history of the HEXTE's flight electronics, which were developed under a subcontract with Perkin-Elmer Corporation. HEXTE-specific Ftools are being developed to insulate users from these details.

Q: What are the "DEFAULT" values in RXTE Cycle 1 & 2for the HEXTE's lower energy bound, rocking direction and rocking dwell time, and how were these values chosen?

A: In addition to the Science Mode (eg. E_8us_256_DX1F), which determines the HEXTE telemetry data format for each cluster, proposers must also specify the following for each HEXTE cluster:
  1. Lower Energy Bound: 5-50 keV,
  2. Beamswitch (cluster rocking direction): +1.5, +/-3.0, +1.5, -1.5, +3.0, -3.0, or 0.0 degrees
  3. On-source dwell time: 16, 32, 64, or 128 s.
Alternatively, users could request "DEFAULT" for any of these parameters, leaving the HEXTE team free to choose appropriate values. The "DEFAULT" values for XTE Cycle 1 observations are given below, along with the reasons for their choice:
1. DEFAULT Lower Energy Bound = 12 keV
Although the HEXTE can detect x-rays at energies as low as 5 keV, an analysis of in-orbit checkout data shows that, as expected, the HEXTE's sensitivity falls off rapidly below 15 keV due to a decreasing effective area and an increasing background count rate. The latter results in a much higher telemetry rate in Event List mode when the lower energy bound is below about 10 keV. Since the PCA instrument is more sensitive than the HEXTE below ~25 keV, the extra telemetry would not be scientifically useful.

The HEXTE's advertised energy range is 15-250 keV, in order to maintain a useful overlap with the PCA (2-60 keV). Since the HEXTE's spectral resolution is ~3 keV FWHM at 15 keV, the HEXTE team have set the lower energy threshold to a nominal 12 keV so that almost all pulses generated by 15 keV photons will be recorded.

2. DEFAULT beamswitch (rocking direction) = +/-1.5 degrees
In the absence of contaminating sources nearby, the best HEXTE background measurements are to made by beamswitching alternately to either side of the source.

Note that it is the observer's responsibility to check the source and background fields of view for contaminating sources, and to request changes in the pointing and HEXTE rocking directions accordingly (see below).

3. DEFAULT on-source dwell = 16 s
This is the HEXTE's fastest rocking dwell time, and the least efficient, since 4 s of off-source data is lost each cycle due to the clusters' motion. However it should minimize systematic errors in background subtraction. Once the HEXTE team have analyzed the power spectra of temporal background variations for a variety of observing conditions, they may elect to increase the on-source dwell time to 32 s.

Q: Which is cluster A and which is cluster B on the spacecraft, and how are the phoswich detectors numbered?

A: Users are also referred to a diagram of the XTE spacecraft. The {x} axis looks towards the source, and the long dimension of the spacecraft contains the {z}axis, which points away from the ASM.

                       /|                                  
                      / |                                  X   
                     /  |                                  ^  Y
                    /   |                                  | /
     _______       /   /|                                  |/
    /      /      *   / |                           Z < ---+
   /______/       |  /  |                                
       |      _ _ |_/___|______________                          
        \    /H/|XTE B//_(o)(o)_______/|                   
         \  /-/ |----//0 / 1/ 2/ 3/ 4/ |    =====             
          \/H/  |E_A//__/__/__/__/__/  |    \   /|           
           |/   |         P  C  A   | =======|_| |        
           /   /|                   | ===       \|        
          /   / |   |||||     ||||| | =/     ASM         
         *   /  |   |||||     ||||| | /                 
         |  /   |___________________|'
         | /   /|       |
         |/   / |       |               XXX   XXX  TTTTTTTTT  EEEEEEEEE
         |   /  |    ___|____           XXX   XXX  TTTTTTTTT  EEEEEEEEE
         |  /   |   /   |   /            XXX XXX      TTT     EEE
         | /   /   /_______/              XXXXX       TTT     EEEEEEE
         |/   /                           XXXXX       TTT     EEEEEEE
         |   /                           XXX XXX      TTT     EEE
         |  /                           XXX   XXX     TTT     EEEEEEEEE
         | /                            XXX   XXX     TTT     EEEEEEEEE
         |/
         '
  				       
   ASCII Art: Jay Sedler, XTE Mission Operations Center.
RXTE spacecraft coordinate system, showing HEXTE clusters A and B, and the five PCUs (numbered 0-4) of the PCA. The RXTE observatory look-direction is along +X; this is constrained to be at least 30 degrees away from the sun vector. The "roll offset", or angle between the X-Z plane and the sun vector, is further constrained to lie within +/-5 degrees. For most pointed observations (i.e. no rasters or scans) the roll offset is 0.0 degrees.
Cluster A is on the {-y} side of the spacecraft, cluster B on the {+y} side, the same as the star trackers. Relative to the spacecraft's {x,y,z} axes, for cluster A the "+" rotation direction is towards {-z}, and for cluster B the "+" rotation direction is towards {-y}.

[HEXTE clusters top view]

View of HEXTE clusters A (left) and B (right) along the direction of incoming x-rays (-x), showing the orientation of their rotation (rocking) axes relative to the XTE spacecraft's Y and Z axes. The phoswich detectors in each cluster are numbered as their data appears in telemetry. The orientation of their hexagonal collimator cells is shown beneath each cluster.

Q: How do I know where the HEXTE clusters' background pointings will end up on the sky?

A: Each HEXTE cluster can "rock" off source by either 1.5 or 3.0 degrees in the following manner: How these rocking positions translate into RA and Dec is determined by the XTE spacecraft roll angle, which is constrained by the relative orientation of the source and the sun. Therefore, the RA and Dec of the background fields may only be calculated once the observation date is known (to within 1 week).

The HEXTE team have provided a software program, HEXTErock, which calculates the RA and Dec centers of the HEXTE clusters' background fields given the source position and the UT date. The WWW version at the link above also produces a sky plot. Just as for the on-source pointing itself, it is the user's responsibility to use the literature or archival data to check their HEXTE background positions for contaminating sources, and then if necessary select an alternative rocking pattern for the affected HEXTE cluster(s), such as 3.0-degree and/or one-sided rocking. (3.0-degree rocking is preferred in such cases, since this allows 2 background positions to be sampled and averaged).

Q: What is the difference between the HEXTE Science Modes E_8us_256_DX1F and E_8us_256_DX0F?

A: E_8us_256_DX0F is the HEXTE Science Mode mentioned in the NRA Technical Appendix (and earlier versions of the HEXTEmporize software) as the preferred mode for faint-source observations (<100 count/s per HEXTE cluster). This mode provides 4 bytes of information for each event: 2 bytes for the time stamp (8us sampling), 1 byte for the detector ID and 1 for spectral information (the Pulse Height Analyzer channel, or PHA), which is a measure of the scintillation pulse strength, related to the incident x-ray's energy. For each event pulse, the HEXTE on-board processor also measures one additional attribute - the pulse SHAPE analyzer channel (PSA), which is a measure of the scintillation pulse's rise time. This is used to distinguish "good" scintillation events in the NaI crystal from background or partial energy loss events in the CsI.

A PSA selection "window" is applied to all events by the on-board processor, and the HEXTE team have attempted to choose upper and lower PSA window values for the best possible background rejection. In E_8us_256_DX0F, the pulse shape value is used for this discrimination, but is then discarded.

In the mode E_8us_256_DX1F, however, each event's pulse shape measurement is retained in the telemetry. Thus the mode E_8us_256_DX1F contains an extra (fifth) byte for each event: the PSA byte.

Why care about the PSA byte?

In future it may be possible for users to improve background event rejection even further by applying joint PSA/PHA selection criteria to their data. For this reason (and since the HEXTE team is still experimenting with different on-board event selection schemes during Cycle 1) it was decided to include the PSA byte in all faint source observations, so for instance:

  • E_8us_256_DX0F (4 bytes) -> E_8us_256_DX1F (5 bytes), and
  • E_2ms_256_DX0D (3 bytes) -> E_8us_256_DX1D (4 bytes).
  • In each case, the additional byte per event increases HEXTE telemetry, and thereby the total size of the HEXTE Science Event dataset. Apart from this, and the future possibility of advanced PHA/PSA selection mentioned above, the reduction/analysis sequence remains the same.

    Q: What is the status of the malfunctioning detector in cluster B, and when did the malfunction occur?

    A: Please read the report on the loss of spectral capability of detector PWB2.

    Q: What are the features in the HEXTE background spectrum?

    A: From Duane Gruber:

    The line features you see are all real, and result from activation of radioactive daughters by cosmic rays and SAA protons. The ~30 keV line is due to Iodine and Thallium daughters decaying by K-capture: after the K-electron is eaten by the nucleus the electron cloud fills the vacancy, with emission of characteristic x-rays. There is a smear of activities between 60 and 80 keV:

  • a 58 keV nuclear inelastic scattering of neutrons on Iodine127
  • K x-rays (28.5-32.5 keV) + a nuclear gamma-ray adding to 67 keV from Iodine125
  • 73-87 keV K x-rays from the Pb collimator
  • K x-rays (28.5-32.5 keV) from Iodine123 + a nuclear gamma-ray adding to 191 keV from Iodine123
  • The slow increase in the continuum from 100 to 200 keV I'm not completely sure of, but I think it could be the Compton continuum from the 191 keV line. At 200 keV we get full energy capture in the 3mm crystal for only a smallish fraction of events, maybe 10 percent. Gory detail is available in my HEAO 1 paper: Radioactivity observed in scintillation counters... in AIP Conference Proceedings 186, "High Energy Background Radiation in Space" (1987, eds C. Rester & J. Trombka), p232.

    [HEXTE background spectrum]

    In-orbit background spectrum for a HEXTE cluster (sum of 4 detectors)

    Q: How sensitive is the HEXTE for faint source observations?

    A: From Philip Blanco:

    Please refer to the Feasibility Chapter in the HEXTE documentation for details. Below I have reproduced the plot of the HEXTE's continuum sensitivity as a function of energy. The effective area, background and FWHM resolution data from which this curve was produced may be obtained as an ASCII text file. However, if you plan to use these data, please let us know where and how.

    [GIF]
    HEXTE faint-source continuum sensitivity per FWHM resolution element, for a 3-sigma detection in a 200 ks observation. This assumes 16-second source/background beamswitching and an average live-time fraction of 60%. Appropriately scaled by sqrt(exposure time), this curve is valid for exposures at least up to 500 ks, where systematic background subtraction effects may appear.


    HEXTE: High Energy X-ray Timing Experiment


    Send questions or comments to Philip Blanco:
    pblanco@ucsd.edu