next up previous contents
Next: 4 Calibration/Maintenance Up: 3.3 On-board data processing Previous: 3.3.3 Continuous Clocking Mode

3.3.4 Telemetry Formats

Within the science data allocation of the AXAF-I telemetry major frame, an asynchronous packetized transport protocol will be implemented for ACIS science data. ACIS engineering housekeeping will be acquired by the RCTU directly and placed in fixed locations within the AXAF-I telemetry major frames.


 
 
Table 3.31: ACIS Serial Telemetry Packets Grouped by System Feature
2cScience Modes
Timed Exposure (TE)
TE Obs. Record
TE Raw Exp. Record
TE Raw Pixel Record
TE Raw Hist. Data
TE Faint Record
TE Faint Data
TE Graded Record
TE Graded Data
TE Obs. Report
 
3cMaintenance  
Hardware Configuration Memory Commands
Dumped TE Bad Pixel Map BEP Memory Dump
Dumped TE Bad Column Map FEP Memory Dump
Dumped CC Bad Column Map SRAM Memory Dump
  PRAM Memory Dump
  BEP Memory Exec. Reply
  FEP Memory Exec. Reply
  Huffman Table Dump
SW Housekeeping  
Command Echo  
Software Statistics  
Fatal Error Message  
   

Table 3.31 lists the ACIS telemetry according to the features called out in the Science Instrument Software Requirements Specification. Information which is sent using the discrete hardware status bits is preceded with the word `Discrete-'. The remaining items are sent as Telemetry Packets, using the Serial Digital Telemetry stream.

ACIS telemetry reaches the ground by being embedded within the AXAF-I telemetry stream. This stream is organized by grouping into major data frames, defined in the AXAF-I to SI Interface Document (provided by TRW). The major data frame will provide an average science data rate of 24 kbps, including headers. Aspect data will not be included with this allocation. All SI data are collected in serial 8-bit words by the RCTU.


 
 
Table 3.32: Science Telemetry Frame Header Contents
Code Function Description
Major Frame Count 2 words
Time Code 6 words; provides Julian date to 1 $\mu$sec resolution.
AXAF Position 6 words; provides AXAF-I position in ECI coordinates. 2 words per axis with 3.0 km resolution.
Geocenter 4 words; direction of geocenter in azimuth and elevation from AXAF-I Line-of-Sight with 10-4 radian resolution
Attitude Code 8 words; orientation of AXAF-I in Right Ascension & Declination of the +X, +Z axes; 2 words per quantity; 10-4 radian resolution

Each 1025 byte minor frame includes a 24 word (192 bits) science data header, as listed in Table 3.32. The header is supplied by the observatory and merged with ACIS data by the CDM subsystem. Position and attitude information given in the header are defined in an Earth Centered Inertial (ECI) coordinate system with the X axis parallel to the sky position of the vernal equinox (epoch 2000; i.e. the First Point in Aries); the Z axis parallel to the Earth rotation axis; and the Y axis defined so as to complete a right-handed coordinate system.

ACIS serial telemetry data appear in two formats. Format 1 is used when the HRC is in the focal plane, and ACIS is `next-in-line.' In this configuration, ACIS telemetry is taken at 0.5 Kbps and stored in the engineering sections of the Format 1 telemetry. When ACIS is in the focal plane, Format 2 is used. In this configuration, ACIS serial telemetry is taken at 24 Kbps and is stored in the Science Data portions of the Format 2 telemetry. Table 3.33 illustrates the overall layout of Format 2's Science Frame. Sixteen Science Frames make up a Telemetry Major Frame. The areas labeled `Science Data' contain data formatted and supplied by the ACIS software. When created by the ACIS software, the science data sections are treated as contiguous. Each section is transferred to the RCTU at 128 Kbps, with appropriate delays inserted between each non-adjacent section.


 

 
Table 3.33: Science Frame Format
1||l||Frame  
Minor 1l|CCSDS
Frame 0 1l|VCDU
  1l|Header
  1l|
Bytes 1l|6
Minor 1l|CCSDS
Frame 1 1l|VCDU
  1l|Header
Bytes 1l|6
. .
. .
. .
Minor 1l|CCSDS
Frame 7 1l|VCDU
  1l|Header
Bytes 1l|6

While ACIS is operated in timed exposure science mode, there are six possible telemetry formats: raw mode event telemetry, raw histogram mode telemetry, faint mode event telemetry, faint with bias mode event telemetry, very faint mode event telemetry and graded event telemetry.

In raw event mode, contiguous regions of pixels are telemetered as a set. Table 3.34 describes the exposure record telemetry and Table 3.35 describes the content of the CCD data telemetry when in this mode. This mode allows every pixel to be examined on the ground, but it is very expensive in terms of telemetry consumption. It is anticipated to be used only occasionally, and primarily as a diagnostic for study of changes in CCD behavior.


 
 
Table 3.34: Exposure Header Content
Item Description
Science Run Start Time Start time of the science run using the ACIS time-stamp latched by the hardware at the start of the run.
Exposure Number Identifies the clocked exposure during the run. Must allow the ground to uniquely identify over 130,000 exposures during a single run (2 second exposures for 72 hours).


 
 
Table 3.35: Raw Mode Data Telemetry Content
Item Description
CCD Identifier CCD (& node ID) from which the raw pixel data were taken.
Window Identifier Identifies 2-D window used to select this set of pixels. (2-D windows may not be included in final version of ACIS software.)
Exposure Overclock values Overclock values read from the CCD while the pixel data were being acquired.
Array of raw 12-bit CCD pixel data Array of raw CCD pixel data. The bottom left position of the first pixel corresponds to the intersection of the identified window and the sub-array readout selection. The pixels are organized by row and then by column, where the number of rows and columns in the set is the intersection of the identified window and sub-array read-out parameters.

In raw histogram mode, each DEA output node has an associated histogram which counts pixels as a function of raw pulse height. Pixels produced during a single exposure whose positions are processed by a given node are counted in the corresponding histogram. After all pixels from the exposures have been processed, all the histograms are telemetered. Tables 3.34 and 3.36 describe the exposure record telemetry and the contents of the CCD data telemetry when in this mode, respectively. This mode is also primarily diagnostic, but is far less consumptive of telemetry, as only the histogram is telemetered to the ground.


 
 
Table 3.36: Raw Histogram Mode Data Telemetry Content
Item Description
CCD Identifier CCD (& node ID) from which the raw pixel data were taken.
Node Identifier Identifies node used to select this set of pixels.
Array of histogram counts Array of 4096 bins. Each bin contains a count of the number of pixels within the node and whose pulse height corresponded to the bin index.

In faint mode event telemetry, sets of distinct events are telemetered as 3x3 arrays of pixels. Table 3.37 illustrates the content of an exposure record, Table 3.38 shows the contents of the event lists telemetered when in this mode.


 
 
Table 3.37: Faint Mode Exposure Record Content
Item Description
Science Run Start Time Start time of the science run, as specified by the DEA latched ACIS time-stamp.
Exposure Number Identifies the clocked exposure during the run. Allows the ground to uniquely identify over 130,000 exposures during a single run (2 second exposures for 72 hours).
Number of Events Telemetered Number of events being telemetered in the exposure. Must accommodate at least 64K events.
Number of Pixels above threshold Number of pixels from a CCD whose pulse heights were above their respective spatial thresholds during the exposure.
Number of Events Discarded due to Bad Pixel and Column Map Number of events which were discarded because they were listed in the Bad Pixel and Column Map.
Number of Events Discarded by Pulse Height Number of events which were discarded due to their pulse height. Must accommodate the total number of pixels within the CCDs.
Number of Events Discarded by Grade Number of events which were discarded due to their `grade.' If grade selection is not performed, these counters need not be telemetered.
Number of Events Discarded by Window Number of events which were discarded by the processing windows.


 
 
Table 3.38: Faint Mode Event Telemetry Content
Item Description
CCD Identifier Identifies which CCD the event occurred.
Position Identifier Row and column of the center pixel of the event.
Uncorrected Pulse Heights Measured (uncorrected) pulse heights of the 9 pixels forming the 3x3 neighborhood of a suspected X-ray event.
Overclock Level Overclock level used for the thresholding during the exposure which produced the event.
Spatial bias offset Spatial bias used for thresholding for this event.

Graded event telemetry also provides information about the 3x3 array of pixels around all suspected X-ray events. But in this mode, (as opposed to the `Faint' mode), instead of all nine pixel values only a combined net pulse height is telemetered and an event ``grade'' which describes the relative distribution of pixels in the array which exceed a split event threshold (see Section 3.3.2 for the definition of event grades). The content of an exposure record in this mode is identical to that used by faint mode telemetry (see Table 3.37 above). The event data content, however, is different. Table 3.39 illustrates the content of the event lists telemetered when in this mode.


 
 
Table 3.39: Graded Event Telemetry Content
Item Description
CCD Identifier CCD the event occurred.
Position Identifier Row and column of the center pixel of the event.
Event Amplitude Amplitude of the event.
Grade Code Identifies the Grade Code (see Sect. 3.3.2).
Corner Pulse Heights Potential function of the corrected pulse-heights of the 4 corner pixels in the 3x3 event array.

While ACIS is operated in continuous clocking science mode, a set of three possible telemetry formats exists: continuous raw mode event telemetry, continuous faint mode event telemetry, and continuous graded mode event telemetry.

In continuous raw event mode, contiguous regions of pixels are telemetered as a set. For each configurable number of rows, a ``Continuous Raw Mode Record'' is formed and telemetered. Table 3.34 identifies the exposure record telemetry; whereas Table 3.40 lists the elements telemetered for each pixel.


 
 
Table 3.40: Continuous Raw Mode Event Data Telemetry
Item Description
CCD Identifier CCD from which the raw pixel data were taken.
Node Identifier Node used to process this set of pixels.
Overclock values Overclock values read from the CCD while the pixel data were being acquired.
Array of raw 12-bit CCD pixel data Array of raw CCD pixel data.

In continuous faint mode event telemetry, a series of distinct events is telemetered as 3x1 arrays of pixels. Table 3.41 illustrates the content of a continuous faint mode record, and Table 3.42 describes the contents of the event lists telemetered when in this mode.


 
 
Table 3.41: Continuous Faint Mode Record Content
Item Description
Science Run Start Time Start time of the Science Run, relative to the time-stamp latched at the start of the run.
Continuous Faint Mode Record Number Uniquely identifies a stream of data within the run. Must allow the ground to uniquely identify over 130,000 records during a single run (2 second exposures for 72 hours).
Number of Events Number of events being telemetered as part of this event stream. Must accommodate at least 64K events.
Number of Pixels above Threshold Number of (possibly summed) pixels whose pulse heights were above the threshold.
Number of Events Discarded due to Bad Column Map Number of events which were discarded because they were listed in the Bad Column Map.
Number of Events Discarded by Pulse Height Number of events which were discarded due to their pulse height. Must accommodate the total number of pixels within the CCDs.
Number of Events Discarded by Grade Number of events which were discarded due to their "grade." If grade selection is not performed, these counters need not be telemetered.
Number of Events Discarded by Window Number of events which were discarded by the processing windows.


 
 
Table 3.42: Continuous Faint Mode Data Telemetry
Item Description
CCD Identifier Identifies in which CCD the event occurred.
Row Identifier Which (possibly summed) row the event was acquired. Used to obtain a high-resolution time-stamp of the event.
Column Identifier Which (possibly summed) column the event was acquired.
Uncorrected Pulse Heights Measured (uncorrected) pulse heights of the 3 pixels of the event's 3x1 array.
Overclock level Overclock level used for the thresholding when the event was acquired.
Spatial Bias Offset If computed, this identifies the spatial bias offset used for thresholding when the event was acquired.

In continuous graded mode event telemetry, sets of distinct events are telemetered using reduced pulse height information and event grade. The content of a continuous graded event record in this mode is identical to that used by continuous faint mode telemetry. The event data content, however, is different. Table 3.43 illustrates the content of the event lists telemetered when in this mode.


 
 
Table 3.43: Continuous Graded Mode Event Data Telemetry
Item Description
CCD Identifier Identifies in which CCD the event occurred.
Row Identifier Which (possibly summed) row the event was acquired. Used to obtain a high-resolution time-stamp of the event.
Column Identifier Which (possibly summed) column the event was acquired.
Center Pixel Pulse Height Corrected pulse height of the 3x1 array's center pixel.
Grade Code Grade Code of the event.

Programmability of the telemetry modes is covered by the `Software Users Manual' by Jim Francis. Such programmability is implemented by uploading a patch to the operating software, so the capability to support required changes is nearly unlimited.

Event timing information can be related to the AXAF-I clock with an accuracy of one period of the ACIS pixel. The original ICD requirement of 1 $\mu$sec was waived to allow the substition of a 100 kHz clock. As currently defined this accuracy is 15 $\mu$sec. The first 32 bits of the Science Data written following the receipt of the Science Frame Pulse is the value of a timestamp counter specially sampled and inserted into the serial output stream by ACIS hardware. Its value represents the number of pixel clock cycles since it was last reset or wrapped. This counter will wrap about once per eight hours. ACIS shall use this same counter to timestamp the start of its Science Runs. This, coupled with the Spacecraft-supplied timestamp information in the Science Header of each frame, specifies the start time of a run. All science data sent by ACIS will be indexed relative to the start of the run, using clocking-mode driven counters (such as exposure counts or data set counts).

Timing with respect to the AXAF-I central clock is received via a science frame synch signal provided by the Spacecraft Interface Unit and utilized by the BEP. The BEP acts as the ACIS Master Clock Generator and synchronizes all ACIS software functions to this signal. (Note the science header telemetry is derived from the same master oscillator [operating at 14.848 MHz] so the telemetry is also synchronous with this clock.)

ACIS provides a Master Timestamp Counter, also resident on the BEP. It is reset only on power-up of the ACIS experiment. Thereafter it is free running, incrementing with each cycle of the science frame, and thus rolling over its 32-bit capacity in 41920 seconds (roughly 700 minutes). When it rolls over it broadcasts a reset pulse to Slave Timestamp Counters running on the FEPs, thus assuring synchronization of the FEPs which generate the `Beginning of Frame' signals at the start of every CCD frame.

The full details of how event timestamps are applied on the ground can be found in Appendix D of the Flight Software Requirements Specifications. This document (MIT Document Number 36-01103) can also be found at the MIT Web site http://acis.mit.edu/sw.html.

Microprocessor diagnostics are not currently supported in the flight software. If they are required they must be uploaded from the ground to enable true diagnostics.


next up previous contents
Next: 4 Calibration/Maintenance Up: 3.3 On-board data processing Previous: 3.3.3 Continuous Clocking Mode

John Nousek
11/21/1997