Table of Contents
Previous Chapter
Table of Contents
Previous Chapter
An observer uses Timed Exposure Science Mode to acquire, process and telemeter two dimensional images from the CCD array. An observer uses a Timed Exposure Parameter Block to configure Timed Exposure Science Mode, and "Start Timed Exposure Run" and "Stop Timed Exposure Run" commands to start and stop the run.
To execute a Timed Exposure Science Run, an observer sends a "Start Timed-Exposure" command to ACIS. ACIS retrieves the parameter block referenced by the start command, dumps the parameter block(s), and proceeds to execute the science run, optionally computing the CCD bias values, and then processing the event data. As the event data is being processed and telemetered, ACIS will optionally compress and trickle the computed bias values into the telemetry stream. The observer allows ACIS to acquire, process and telemeter CCD data for a period of time, after which the observer sends a "Stop Timed-Exposure" command to ACIS. ACIS then stops the run and telemeters a Science Run Report. Optionally, the observer may then decide to recompute the bias values and send a "Compute Timed-Exposure Bias" command. ACIS will then re-compute and telemeter the bias values and then stop.
If the observer requires parameters not already provided by the ACIS parameter library, the observer must issue a "Load Timed-Exposure Parameter Block" command to load the desired parameters for the run. If the loaded parameter block requires custom event processing windows, the observer must also issue a "Load 2-D Window List" command to place the desired window settings into the instrument.
Figure 17 illustrates the interactions involved to configure and start a Timed Exposure Science Run. Figure 18 illustrates some of the Front End processor interactions during the run and Figure 19 illustrates what the Back End processor does with the generated data. Whether or not the Bad Pixel and Column Maps are applied by the Front End or Back End Processors is left for the design phase, and is not shown in the diagrams. See Section 3.2.2.3.13 on page 64 for the detailed requirements of the functionality of the Bad Pixel and Column Map applications.
FIGURE 17. Timed Exposure Setup and Start Sequence Diagram
FIGURE 18. Timed Exposure Front End Processor Sequence Diagram
FIGURE 19. Timed Exposure Back End Event Processing Sequence Diagram
FIGURE 20. Timed Exposure Pixel Processing Data Flow
An observer configures Timed Exposure Mode using a Timed Exposure Mode Parameter Block. Table 11 describes the content of a Timed Exposure Parameter Block.
TABLE 11. Timed Exposure Parameter Block Content (Continued)
----------------------------------------------------------------------------------------------------------- Req Item Description----------------------------------------------------------------------------------------------------------- 11-1 Parameter Block This item is a ground-supplied identifier which is echoed by the instru Identifier ment software in key portions of the science telemetry. 11-2 FEP CCD Selec These items specify which CCD a given FEP should process. The items tion are indexed by FEP Id. Any given CCD may be specified more than once in the set, at the cost of reducing the total number of CCDs being used. CCDs or FEPs which are not powered on (see Section 3.2.7 on page 99) are ignored and will not be used for the run. If there are no powered CCD being processed by a powered FEP, the science run will be aborted. 11-3 Subarray Readout An observer uses the start row and row count values to effectively Start Row and Row shorten the time it takes to transfer an image from the CCDs to the Count Front End Processors by sacrificing some of the imaging area. The start row and row count specifies which contiguous horizontal swath of CCD rows are transferred from the Framestore back to the FEPs. These two parameters apply to all selected CCDs. The Start Row parameter can range from 0 to 1023, where 0 corresponds to image area row 2, 1 to image area row 3, and so on. (NOTE: Image area rows 0 and 1 are never used), and the Row Count parameter can range from 1 to 1024. If Start Row plus Row Count extends beyond the image area, the software truncates the subarray row count to fit. 11-4 Primary and Sec These items control the amount of time that the Image Area of each ondary Exposure CCD is integrating data. This does not include the time taken to trans Times fer the Image Area into the Framestore. These values can range from 0 to 10 seconds in 0.1 second increments, and apply to all of the selected CCDs. They are accurate to within one SRAM Major Cycle (~10µs). The CCDs are clocked using the primary exposure time for 1 exposure, followed by the secondary exposure time for "N" exposures. "N" is determined by the Exposure Duty Cycle parameter. If "N" is 0, the sec ondary exposure time is not used. If both exposure times are used, the bias computation algorithms only use data from exposures integrated over the Secondary Exposure time. NOTE: If Image Area to Framestore transfers need to be performed at different times on each CCD, and the specified exposure time is shorter than can be supported by the number of selected CCDs, the shortest possible exposure time will be used instead of the specified time (i.e. the Image to Frame transfer time multiplied by one less than the num ber of selected CCDs). 11-5 Exposure Duty This parameter controls how many secondary exposures separate pri Cycle mary exposures. If this parameter is 0, all exposures use the Primary Exposure Time (i.e. the Secondary Exposure Time is ignored). If 1, there is one Secondary Exposure between each Primary Exposure, if 2, there are two Secondaries for each Primary, and so on. This parameter can range from 0 to 15. 11-6 Output Register This item controls the clocking mode of all of the CCD's Serial Output Clocking Mode Shift Registers. This item can specify either; Full-Mode, AC-Mode, BD-Mode or Diagnostic-Mode. See Section 2.2.2 on page 20 for a description of these modes. 11-7 Number of Over This item specifies the number of extra pixel pairs to clock out of each clock Pairs of the Serial Output Shift Registers. Timed Exposure Mode uses these pixels to determine the bias introduced to the CCDs pixel pulse height by the analog electronics. This item can specify 0 to 15 pairs of pixels. 11-8 Video Chain These items control the response of the Video Subsection in the Detec Responses tor Electronics Assembly for each CCD. This item selects 1 or 4 elec trons per Analog-to-Digital Unit (ADU). These items are indexed by FEP Id. If a single CCD is specified by more than 1 FEP entry, then the last corresponding chain response entry is used for the listed CCD. 11-9 On-chip Summing This item controls whether or not 2x2 on-chip summing is performed. Select If disabled, no pixel summing is performed on the CCD. If this is enabled, then the charge from pairs of adjacent columns and rows are summed on the CCD. Each measured pixel output from the CCD is the sum of a 2x2 pixel array from the chip. 11-10 Reference to an This is an optional item which allows a maintainer to by-pass the on- explicit DEA Load board SRAM/PRAM load builder, and load these areas directly. This reference points directly into BEP memory. If 0, then no explicit DEA load is to be used. If not 0, then the item points to the memory block which contains the load image to use. Post-launch DEA load images are uplinked into memory by the maintainer using the Write Memory function of the instrument (see Section 3.2.8.3.2 on page 108), or via a patch. This parameter is intended to be used by the maintainer of the instrument, and should be 0 during normal operations. The format of the referenced SRAM/PRAM load will be specified in the ACIS Instru ment Software Detailed Design Specification (AS-BUILT), MIT 36- 53200, and shall support different loads into different DEA boards. 11-11 Front End Selec This item specifies whether or not the Front End Processor should sup tion Algorithm ply all pixels to the Back End (Raw Mode), compute histograms of raw pixel data (Histogram Mode), or supply pixels to the Back End whose center is above the current threshold (Event-finding Mode) and whose pulse height is greater (or sometimes equal) to its immediate neighbors (3x3 local maxima). 11-12 Threshold Set These items control minimum pulse-height above or below bias a pixel Points should have to be considered as an event candidate by a given FEP. Each of these items applies to a specific CCD output node being pro cessed by a FEP. These values can range from -4096 to 4095. These values are not used when the FEPs are processing raw pixels or histo grams. One set of four values is specified for each FEP. 11-13 Reference to Front This value allows a maintainer to override the standard code and data End Processor loaded into the Front End Processor, and provide a special load for the Code Science Run. This reference points directly into BEP memory. If 0, then no explicit FEP load is to be used. If not 0, then the item points to the memory block which contains the load image to use. Post-launch FEP load images are uplinked into memory by the maintainer using the Write Memory function of the instrument (see Section 3.2.8.3.2 on page 108), or via a patch. This parameter is intended to be used by the maintainer of the instrument, and should be 0 during normal opera tions. The format of the referenced FEP Load image will be specified in the ACIS Instrument Software Detailed Design Specification (AS- BUILT), MIT 36-53200. 11-14 Event List Packing This parameter is only used when the Front End Selection Algorithm is Code in "Event Finding Mode". This value is not used when the Front End is processing raw pixels or histograms. This value selects the format to use when sending processed events. Currently, the following event-list formats are supported; Faint-Mode, Faint-with-Bias, and Graded-Mode 11-15 Reference to a 2-D This parameter is only used when the Front End Selection Algorithm is Window Collection in "Raw Mode" or "Event Finding Mode". This value is not used when the Front End is processing histograms. An observer uses this item to specify a set of 2-dimensional event selection and processing windows to use for the Science Run. If no windows are specified, Timed Expo sure Mode shall process all events from the CCDs using the processing parameters specified in this block. See Table 13, "2-Dimensional Win dow List Parameter Block," on page 58. 11-16 Split Threshold These parameters are only used when the Front End Selection Algo Set Points rithm is in "Event Finding Mode". These values are not used when the Front End is processing raw pixels or histograms. These values specify the threshold above which a neighboring pixel's pulse height must be to be considered as part of an event. These values are used when grading events and computing the amplitude of an event. Each of these items applies to a specific CCD output node being processed from a particu lar FEP. These values can range from 0 to 4095, and each set is indexed by FEP Id. 11-17 Ignore Bad Pixel/ These parameters are only used when the Front End Selection Algo Column Map Flags rithm is in "Event Finding Mode". These values are not used when the Front End is processing raw pixels or histograms. These values control whether or not to discard events on the basis of the CCD's list of bad pixels and/or columns. 11-18 Grading Selection These parameters are only used when the Front End Selection Algo Bit Map rithm is in "Event Finding Mode". These values are not used when the Front End is processing raw pixels or histograms. This set of values select which event grades to accept. Accepted events whose grades are indicated in this list shall be telemetered. This bitmap consists of a 256 bits, where each bit corresponds to one of the possible grade codes. 11-19 Event Amplitude These parameters are only used when the Front End Selection Algo Lower Bound and rithm is in "Event Finding Mode". These values are not used when the Range Front End is processing raw pixels or histograms. These parameters select the minimum and maximum amplitudes of accepted events. Events whose total amplitude (function of the 3x3 grid around the event) is less than the lower bound value, or greater than or equal to the lower bound plus range are not telemetered. These field's ranges are governed by the minimum and maximum corrected event amplitudes. 11-20 Re-compute Bias This item indicates if the software must re-compute the pixel bias val Flag ues prior to starting data processing. 11-21 Trickle Bias Flag This item indicates if the software must telemeter the current bias map to telemetry after it has been computed. 11-22 Bias Algorithm This specifies the type of bias computation to be performed. This field Mode may indicate that the bias be computed using an algorithm which uses entire image frames for each of its iterations (faster, but less accurate) called "Whole Frame Mode", or uses an algorithm which incrementally operates on sections of the bias map (slower, but more accurate), called "Strip" Mode. The remaining bias parameters depend on the selected mode. There is one set of settings for each FEP, and each set is indexed by FEP Id. 11-23 Whole Frame: This specifies the number of conditioning exposures used to determine Conditioning the minimum pixel bias value. This field can vary from 0 to 200 expo Count sures. 11-24 Whole Frame: This specifies the number of approximation exposures used to locate Approximation the actual bias value above the minimum value determined during the Count conditioning pass. This field can vary from 0 to 200 exposures. 11-25 Whole Frame: Low This parameter is used to discover abnormally low pixel values after Pixel Rejection the conditioning process, and is used only if it's value is not zero. If a Threshold pixel's value is smaller than all but 1 of its adjacent neighbors by more than this value, it is replaced with the median value of its neighbors. This field's value can range from 0 to 4095. 11-26 Whole Frame: During the approximation-to-the-mean phase, this parameter is used to Event Rejection discard pixel samples which potentially contain charge from an event Threshold or charged-particle. If a sample pixel's value exceeds the current bias estimate by more than this value, it and its immediate neighbors are not used to adjust the bias estimate. This field's value can range from 0 to 4095. 11-27 Whole Frame: This parameter is used when approximating the mean to prevent sam Rejection Thresh ples whose pulse heights are uncharacteristically high from affecting old for approxima the mean estimate. Pixel samples whose pulse heights are larger than tion-to-mean the current pulse height estimate by more than this amount are not included in the estimated mean. This field's value can range from 0 to 4095. 11-28 Strip Mode: Expo This specifies the total number exposures to use per strip of pixels. This sures Count field can vary from 0 to 40 exposures per strip. 11-29 Strip Mode: Mean/ This specifies whether to base the bias value on a truncated mean or Median Selection median of the pixel samples. 11-30 Strip Mode: Mean When a mean algorithm is selected, pixel samples whose value is Sigma Rejection/ greater than this value multiplied by the RMS variance are not included Median Index in the final mean computation. The value can range from 0 to 5. When the median algorithm is selected, this parameters indexes the pulse height to use for the bias out of the sorted list. In this case, the field value can range from 0 to 4095. 11-31 Raw Histogram This item specifies how many exposures to use when building histo Exposure Count grams. This field can vary from 1 to 240,000 exposures. 11-32 Bias Compression This indicates whether or not to compress pixel bias map values prior Flag to being telemetered. 11-33 Bias Compression If bias map compression is enabled, these items specify which com Selections pression table to use for a given FEP. These selections are indexed by FEP Id. 11-34 Raw Compression This indicates whether or not to compress raw pixel data values prior to Flag being telemetered. 11-35 Raw Compression If raw pixel data compression is enabled, this item specifies which Selection compression table to use for all FEPs. 11-36 Initial Frames to This specifies the number of whole exposure frames that the FEP is to Ignore ignore at the start of a bias calibration that uses this parameter block. If this value is zero, each FEP will start processing all data pixels and overclocks received after the first vsync code. Science data processing, however, always ignores the first two whole exposures. -----------------------------------------------------------------------------------------------------------
Events produced by Timed Exposure Clocking have an originating position within the CCD's Imaging Section. Timed Exposure Mode has the capability to select and process events based on this position. This selection and processing is controlled using a 2-D Window Parameter Block. Table 12 illustrates the content within a single 2-D Window, and Table 13 illustrates the content of a list of 2-D Windows.
TABLE 12. 2-Dimensional Window Information (Continued)
----------------------------------------------------------------------------------------------------- Req Item Description----------------------------------------------------------------------------------------------------- 12-1 CCD Identifier This selects the CCD (out of all 10) to which the window applies. 12-2 Window Position This specifies the window position in terms of the bottom-left pixel within the CCD's Imaging Section. 12-3 Window Width This specifies the width of the window in terms of Imaging Section col umns (1-1024). 12-4 Window Height This specifies the height of the window in terms of Imaging Section rows (1-1024). 12-5 Sample Cycle If processing event data, this specifies the sampling frequency of the window. If zero, no events are processed in the window. If one, then every event is processed, if 2, every other event is processed, if 3, every third, and so on. If processing raw pixel data, this specifies whether or not to send pixels who belong to the window. If zero, all pixels belonging to the window are discarded. If not zero, the pixels are telemetered. 12-6 Lower Event When processing events, these items set the lower and upper event Amplitude and amplitude bounds used to select candidate events. Events whose ampli Range tude is outside this range are rejected. When processing raw pixel data, this parameter is ignored. -----------------------------------------------------------------------------------------------------
TABLE 13. 2-Dimensional Window List Parameter Block (Continued)
--------------------------------------------------------------------------------------- Req Item Description--------------------------------------------------------------------------------------- 13-1 Parameter Block This item is a ground-supplied identifier which is echoed by Identifier the instrument software in key portions of the science telem etry. 13-2 Window Count This item identifies the number of 2-D Windows contained in the Window List. This value can range from 0 to 36. This quantity may not actually be loaded nor telemetered, but may be calculated using the length of the enclosing com mand or telemetry packet. 13-3 Windows[Win This item is an array of 2-D windows. The content of these dow Count] windows is described in Table 12. The number of windows in the array is determined by "Window Count." ---------------------------------------------------------------------------------------
Due to the expected size of SRAM and PRAM loads (greater than 2Kbytes), the ACIS software shall be capable of building the SRAM and PRAM loads needed to perform a Timed Exposure Clocking Mode, as described in Section 2.2.3 on page 22.
The following itemizes the inputs needed by the builder:
The output shall consist of a series of DEA SRAM and PRAM load commands to the DEA sequencer.
The CCD Selections specify which DEA boards need to be programmed. Since power constraints require the instrument to stagger large parallel clock transfers (such as Image Area to Framestore transfers), the builder uses knowledge of the total number of CCDs selected to prevent transfers on the different CCDs from overlapping.
The exposure times control the time between the clearing of a CCD's Imaging Section and the transfer of its rows to its Framestore, and the duty cycle controls number of iterations of each type of exposure time.
The subarray readout start and end rows control which contiguous rows of the CCD image to actually transfer out of the Serial Output Shift Registers to the DPA.
The On-chip summing selection controls how the Framestore is clocked into the Output Register and how each pixel is processed at the CCD output nodes.
The Output Register Clocking Mode specifies how to clock out the output registers.
The number of overclock pixels determines how many pixels to clock out of the output registers after each row of image data has been transferred.
The video chain response information controls how the pixel data is processed in the DEA's analog video processing section.
The optional DEA SRAM/PRAM load override instructs the builder to ignore all of the previous parameters and to load the explicitly specified SRAM and PRAM commands.
These operations address the condition where CCD's exposure time is longer than the time it takes to transfer the image from its Framestore to the DPA. In this mode, ACIS shall program the DEA's PRAM and SRAM to clock the CCDs in the following fashion:
1. Clear the Imaging Section, Framestore and Serial Output Shift Registers.
This is accomplished by clocking the Imaging Section and Framestore Rows 1026 times, and then clocking all pixels out the output registers a number of times, discarding the output. For Full-Mode and Diagnostic-Mode, (256 + 4) pixels must be shifted out of the output registers in order to clear the registers once. In AC-Mode and BD-Mode, (512 + 8) pixels must be shifted. If overclocks are specified, clock and discard the output registers an additional number of times.
2. Expose the Imaging Section for the duration specified by "Exposure Time."
The exposure effectively starts after the last row of the Imaging section has been clocked into the Framestore, and before the output registers have been flushed. On the first exposure of the run, this can be accomplished by just waiting for the specified period of time. On subsequent exposures, however, this wait period is overlapped with the steps listed below. The effective "do-nothing" time is the exposure time minus the time it takes to copy the Imaging Section to the Framestore and to transfer the image to the DPA.
3. Transfer the Imaging Section to the Framestore.
This is accomplished by clocking 1026 rows from the Imaging Section to the Framestore and from the Framestore into the output registers. This has the effect of summing any garbage data previously left in the Framestore into the output registers. This operation shall take no more than 60 msec. The goal is to perform this operation as quickly as the hardware allows.
4. Position start of subarray at row 0 of the Framestore.
This is accomplished by clocking "Start Row+2" Framestore rows into the output register. This places the first subarray row into Framestore row 0. This also has the effect of summing all rows prior to the subarray with the contents of the output registers.
5. Clock out output registers and discard.
This is accomplished by clocking the pixels out of the output registers and into the output nodes and discarding the data. For Full-Mode and Diagnostic-Mode, (256 + 4) pixels must be shifted out in order to clear the registers. In AC-Mode and BD-Mode, (512 + 8) pixels must be shifted. If overclocks are specified, clock and discard the output registers an additional number of times.
6. Clock desired number of image rows (i.e. "End Row - Start Row + 1") to the DPA
This is accomplished by clocking one row at a time from the Framestore to the output registers, and then clocking pixels from the output registers to the output nodes. As each pixel's data arrives at the output node, it is sampled, and converted to a 12 bit digital representation, which is then transmitted to the DPA. If overclock data is needed, clock and sample the output registers an additional number of times. The readout of the Framestore shall take no longer than 6.5 seconds. The goal is to perform this operation as quickly as the hardware allows.
7. Repeat starting from step 2.
NOTE: Rows after "End Row" in the Framestore are clocked and discarded as part of 3) and 5). As the Imaging Section is transferred into the Framestore, the rows left from the previous image are summed into the output registers. The summed rows are then discarded as part of 5).
These operations apply when the configured CCD exposure time is less than the time it takes to transfer the image from the Framestore to the DPA (NOTE: This time depends on the subarray parameters). In general, the only difference between this situation and the "Normal Exposure CCD Clocking Operations" is that there is an extra "flush" of the Imaging Section prior to the start of an exposure. In this mode, ACIS shall program the DEA's PRAM and SRAM to clock the CCDs in the following fashion:
1. Clear the Imaging Section, Framestore and Serial Output Shift Registers.
This is accomplished by clocking the Imaging Section and Framestore Rows 1026 times, and then clocking out the output registers and discarding the output. For Full-Mode and Diagnostic-Mode, (256 + 4) pixels must be shifted out of the output registers in order to clear the registers. In AC-Mode and BD-Mode, (512 + 8) pixels must be shifted. If overclocks are specified, clock and discard the output registers an additional number of times.
2. Expose the Imaging Section for the duration specified by "Exposure Time."
This is accomplished by waiting for the desired period of time. The exposure effectively starts after the last row of the Imaging Section has been clocked into the Framestore, and before the output registers have been flushed.
3. Transfer the Imaging Section to the Framestore.
This is accomplished by clocking 1026 rows from the Imaging Section to the Framestore and from the Framestore into the output registers. This has the effect of summing any garbage data previously in the Framestore into the output registers. This operation shall take no more than 60 msec. The goal is to perform this operation as quickly as the hardware allows.
4. Position start of subarray at row 0 of the Framestore.
This is accomplished by clocking "Start Row+2" Framestore rows into the output register. This places the first subarray row into Framestore row 0. This also has the effect of summing all rows prior to the subarray with the contents of the output registers.
5. Clock out output registers and discard.
This is accomplished by clocking the pixels out of the output registers and into the output nodes and discarding the data. For Full-Mode and Diagnostic-Mode, (256 + 4) pixels must be shifted out in order to clear the registers. In AC-Mode and BD-Mode, (512 + 8) pixels must be shifted. If overclocks are specified, clock and discard the output registers an additional number of times.
6. Clock desired number of image rows (i.e. "End Row - Start Row + 1") to the DPA
This is accomplished by clocking one row at a time from the Framestore to the output registers, and then clocking pixels from the output registers to the output nodes. As each pixel's data arrives at the output node, it is sampled, converted and transmitted to the DPA. If overclock data is needed, clock and sample the output registers an additional number of times. The readout of the Framestore shall take no longer than 6.5 seconds. The goal is to perform this operation as quickly as the hardware allows.
7. Clear the Imaging Section
This is accomplished by clocking 1026 rows from the Imaging Section into the Framestore.
8. Repeat starting from step 2).
NOTE: Rows after "End Row" in the Framestore and the garbage transferred from the Imaging Section to the Framestore as part of 7) are clocked and discarded during 3) and 5).
If configured to do so, the Front End Processor software on ACIS shall be responsible for computing, reporting and using the overclock levels of its CCD. Overclock data is generated by clocking the Serial Output Shift Registers after all of a row's pixel data has been clocked out. The clocking operation is under control of the SRAM and PRAM loaded into the DEA's sequencers. As the overclock data arrives from the CCDs, the Front End hardware writes the selected overclock data to a set of special buffers.
The Front End Processor software then averages the data contained in these buffers to produce an output node-specific overclock level of the CCD. The average is rounded to the nearest integer value.
(EQ 3) Overclock Level = SUM(Overclock Data) / # Overclock Elements
ACIS uses the resulting overclock levels to control the threshold values for the subsequent exposure(s). The overclock levels are also included with the event data telemetry.
The Front End Processor software shall re-compute and apply the overclock levels every exposure.
The pixel-by-pixel bias map values are computed when configured by the parameter block, instructed by a "Compute Bias" command, or when the bias maps are invalid. Prior to processing events, the Front End Processor shall use one or more exposures to compute output node-specific "Initial Overclock" values, pixel-by-pixel bias levels and store the results in the Front End's Pixel Bias Map.
(EQ 4) Pixel Bias Map Value = Computed Pixel Bias + Initial Overclock
NOTE: If the "Exposure Duty Cycle" parameter is not 0, then only exposures whose integration time corresponds to the "Secondary Exposure Time" parameter (see Table 11) are included in the bias computation.
The algorithm to use to compute the bias is determined by the "Bias Selection" parameter of the Timed Exposure Parameter Block. The chosen algorithm may be tuned using the "Bias Parameters" of the same block.
The Front End Processor software shall compute the output node-specific threshold register values according to the following formula:
(EQ 5) Threshold Register Value = Threshold Set Point + (Overclock Level - Initial Overclock)
The Front End Processor software shall set the threshold registers with the computed values prior to each exposure.
The Front End Processor contains a Pixel Threshold BitMap. This bit map indicates which pixels in the acquired CCD image have pulse heights greater than their respective output -node threshold registers plus their respective pixel bias map values. The bit corresponding to a pixel is set if:
(EQ 6) Pixel Pulse Height > Pixel Bias Map Value + Threshold Register Value
When the "Front End Selection Algorithm" field of the Timed Exposure Parameter Block indicates the Event-finding mode, the Front End Processor software shall scan this bitmap for each exposure. Any pixel marked as above its threshold shall be counted in an "above threshold" counter, and be tested as an event. This counter is telemetered with the event data. Refer to Section 3.2.2.3.10 for the requirements the Event-finding mode requirements.
The threshold map is ignored for all Front End Modes except "Event-finding Mode."
When the "Front End Selection Algorithm" field of the Timed Exposure Parameter Block indicates the Event-finding mode, the Front End processor selects events from candidates indicated by the hardware Pixel Threshold Map. If a pixel is indicated in this map, the Front End software compares the corrected pulse height (see Section 3.2.2.3.14 on page 65) of the pixel to the 8 surrounding pixels, as shown in Figure 21. In order to meet the criteria for an event, the center pixel (black square) must have a corrected pulse height greater than or equal to the previous row and the pixel immediately preceding it within its row (shaded squares), and must have a corrected pulse height greater than the pixels in the subsequent row, and the pixel immediately after it in its row (hatched squares).
FIGURE 21. 3x3 Event-finding Mode Pixel Diagram
If the pixels meet the criteria, the Front End processor forwards the 3x3 group of pixels to the Back End processor for further selection and processing.
Each Front End processor must be capable of testing at least 1000 candidate events per second. When put together, all 6 of the Front End processors must be capable of handling a total of at least 750 actual events per second.
When the "Front End Selection Algorithm" field of the Timed Exposure Parameter Block indicates "Raw Mode", the Front End processor merely passes raw pixel data to the Back End Processor as it is poured into the Front End's image buffer.
When the "Front End Selection Algorithm" field of the Timed Exposure Parameter Block indicates "Histogram Mode", the Front End processor forms one histogram for each active CCD output node (i.e. If "Output Register Clocking Mode" is Full-Mode or Diagnostic-Mode, four histograms are produced. In AC-Mode or BD-Mode, only two histograms are produced).
Each histogram contains 4096 bins, one bin for each raw pixel pulse height, where each bin is 32-bits wide. As raw pixel data is poured into the Front End's image buffer, the Front End processor reads each acquired raw-pixel pulse height and increments the location indexed by pulse height value. It proceeds to do this for a user-selectable number of exposures (as specified in the "Histogram Parameters" field). Once the desired number of exposures have been processed, or if a "Stop Timed Exposure" command is received, the Front End processor stops processing, and forwards the histogram data to the Back End processor to be telemetered.
For each exposure added to the histogram, the Front End Processor shall compute the overclock values for the exposure, and maintain the minimum, maximum, mean and variance of the computed overclock values. Once the histograms are complete, the Front End processor shall forward the computed minimum, maximum, mean and variance values to the Back End processor, which includes the information in the telemetry stream.
NOTE: If the "Exposure Duty Cycle" is not zero, then only exposures whose integration time is defined by "Secondary Exposure Time" are accumulated into the histograms (see Table 11).
The Front End Processors shall be able to process all pixels from every other 2.65 second exposure. It is a goal to be able to keep up with each 2.65 second exposure.
The Back End Processor shall contain a list or bitmap of bad pixels and columns from each CCD. This Bad Pixel and Column Map is edited using maintenance commands from the ground (see Section 3.2.7 on page 99). The Bad Pixel and Column Map entries provide the information described in Table 14 and Table 15.
TABLE 14. Bad Pixel Entry Information (Continued)
----------------------------------------------------------------------------------------------- Req Item Description----------------------------------------------------------------------------------------------- 14-1 CCD Id The Bad Pixel Map must contain enough information to identify the CCD to which the bad pixel entry applies. 14-2 Pixel Row and Column Id The Bad Pixel Map must contain enough information to identify which pixel in the CCD is bad. -----------------------------------------------------------------------------------------------
TABLE 15. Bad Column Entry Information (Continued)
------------------------------------------------------------------------------------------- Req Item Description------------------------------------------------------------------------------------------- 15-1 CCD Id The Bad Column Map must contain enough information to identify the CCD to which the bad column applies. 15-2 Column Id The Bad Column Map must contain enough information to identify which col umn in the CCD is bad. -------------------------------------------------------------------------------------------
When the "Front End Selection Algorithm" field of the Timed Exposure Parameter Block indicates the Event-finding mode, and the "Ignore Bad Pixel Map Flag" is de-asserted, the software shall test any events generated by the Front End to ensure that the center pixel of the 3x3 event is not listed in the Bad Pixel and Column Map. If the center pixel is in the list, no further processing on the event takes place. If the center of the event is not in the list, software accepts the event for further selection and processing.
Bad pixels or pixels belonging to a bad column which surround the center pixel of an event shall be treated as having a pulse-height of 0.
The Bad Pixel and Column maps are ignored for all Front End Modes except "Event-finding Mode."
To obtain the corrected pulse height of a pixel, the ACIS Science Instrument Software subtracts the corresponding output node overclock level, computed pixel bias value (i.e. the stored bias map value minus the global initial overclock) from its raw pulse height
(EQ 7) Corrected Pixel PH = raw pixel ph - pixel bias map value - (overclock - initial overclock)
The amplitude of an event is the sum of the corrected pulse heights of the appropriate pixels in the group. The pixels whose pulse heights are to be included are shown in Figure 22, and are as follows:
FIGURE 22. Event Amplitude Pixel Diagram
When the "Front End Selection Algorithm" field of the Timed Exposure Parameter Block indicates the Event-finding mode, the Back End Processor shall compute the overall amplitude of 3x3 events received from the Front End, and compare the result with the "Event Amplitude Lower Bound and Range" specified in the Timed Exposure Parameter Block. If the value is less than the lower bound, or greater than or equal to the upper bound (lower bound plus range), the software shall increment a telemetered event amplitude rejection counter and discard the event. If the value is within range, the software shall accept the event for further selection and processing.
Event Amplitude selection only applies when the "Front End Selection Algorithm" field indicates "Event-finding Mode."
When needed for event selection or telemetry, the Back End software shall compute the "grade" of a 3x3 pixel event. To accomplish this, the software compares the corrected pulse heights of the 8 outer edge pixels of the event to the "Split Threshold" specified in the Timed Exposure Parameter Block. Each pixel corresponds to a "bit" in the grade code. If a pixel's pulse height is greater than or equal to the threshold, the corresponding bit is marked as a `1'. If a pixel's pulse height is less than the threshold, the bit is marked as a `0'. Since there are 8 pixels on the edge of a 3x3 event, this leads to an 8 bit grade code. There are 256 possible "grades" for a 3x3 event. The actual pixel position to grade code bit assignments will be defined in the design.
Figure 23 illustrates an example 3x3 event, with pixel pulse heights drawn as vertical bars, and the split threshold drawn as a grid. The pixels whose pulse height is above the threshold are shown to the right as lightly shaded areas. The since the center pixel does not contribute to the grade, it is shown using a darker shade. The diagram then shows the mapping of the pixels to grading bits, and from grading bits to the resulting Grade Code.
FIGURE 23. 3x3 Event Grading Illustration
When the "Front End Selection Algorithm" field of the Timed Exposure Parameter Block indicates the Event-finding mode, and the "Grading Selection Bitmap" does not indicate that all possible grades are to be accepted, the Back End software shall compute the grade of an event and compare the computed code to the list of desired grades. If the computed grade is not in the list, the grade discard counter is advanced and the event is ignored. If the event's grade is in the list, the event is accepted for further selection and processing.
Grade-based event selection only applies when the "Front End Selection Algorithm" field indicates "Event-finding Mode," and when the "Grading Selection Bitmap" indicates that only a subset of grades are to be processed (i.e. if all grades are accepted, no grade-based selection operation is performed).
The Back End software shall use 2-D windows to select events and pixels to report. If an event's center pixel, or an individual pixel's position is within the bounds of a 2-D window, it is selected according to the window. For a given event or pixel, each window in the window list is checked in the order they are presented in the list. If the event/pixel position is within the boundaries of the window, the event/pixel is processed by the window. If the position is outside the bounds of the window, the next window is checked. This continues until the list is exhausted.
If no window list is specified, or if the event/pixel does not intersect with any of the configured windows, the event/pixel is, by default, accepted for processing. This behavior can be thought of as acting as a default window which encompasses the entire CCD and whose property is to accept all events/pixels.
The possible selection modes for the windows are as follows:
For example, to accept only events which occur within a small region of a CCD, two windows are needed, as shown in Figure 24. The first window (white square) in the list specifies the region to accept, and has a processing mode which accepts all events. The second window (shaded square) encompasses the entire CCD, and has a processing mode which rejects all events. Events whose position intersects the smaller first window are accepted by the window. All remaining events will intersect only the second window, and are rejected.
FIGURE 24. Example Clipping Technique
Another example to perform the inverse of the previously described operation, as shown in Figure 25. Here, we want to reject events in the smaller window, and accept all other events. This scenario only requires a single window. This window identifies the region from which to reject events. Events which intersect this window are then rejected. All other events will be handled by the default CCD window, whose scope is the entire CCD, and whose property is to accept all events.
FIGURE 25. Example Inverse Clipping Technique
Figure 26 illustrates an fictitious layout which uses several windows.
The Back End software produces raw pixel telemetry when the "Front End Selection Algorithm" field of the Timed Exposure Parameter Block indicates a "Raw" mode. In this mode, contiguous regions of pixels are bit-packed into telemetry packets and telemetered as a set. For each CCD readout, zero or more Raw Mode Data Telemetry packets are produced, followed by one Raw Mode Exposure Record packet.
Table 20 describes the content of the CCD data telemetry when in this mode and Table 17 describes the exposure record telemetry.
TABLE 16. Raw Mode Data Telemetry Content (Continued)
----------------------------------------------------------------------------------------------------- Req Field Description----------------------------------------------------------------------------------------------------- 16-1 Packet Synch This field marks the start of the ACIS packet within the telemetry stream. 16-2 Packet Word This field indicates the number of 32-bit words in the packet. Length 16-3 Format Tag This field indicates the type of data contained in the telemetry packet, in this case a Timed Exposure Raw Mode data packet. 16-4 Packet This field is used to sequentially number all ACIS telemetry packets. Sequence Num ber 16-5 CCD/FEP Iden These identify which CCD and FEP produced the raw pixel data. tifiers 16-6 Data Packet Since data from different CCDs can be interleaved, and also mixed with Number other kinds of packets, this field sequentially numbers the data packet for a given CCD exposure. 16-7 Compression These fields indicate whether or not the raw data has been compress, and Selection Flag if so, by which compression table. and Table Selec tion 16-8 Pixel Position This specifies the row and column of the first pixel in the packet in a col lection of raw pixel data. 16-9 Packed or com This consists of an array of window-selected raw CCD pixel data. Pixels pressed collec filtered out by the window processing are dropped. This forces the ground tion of raw 12- to use knowledge of the chosen 2-D windows to reconstruct the image. bit CCD pixel The packed pixels are output row by row, starting from the first row and overclock clocked out of the CCD, with the overclock data for a given row sent data immediately after the pixel data for the row. Overclock pixels are always sent, even if the row is completely filtered out by the 2-D windows. Unused bits at the end of the telemetry packet (needed to fill out to the nearest 32-bit word boundary) will be zero. -----------------------------------------------------------------------------------------------------
TABLE 17. Raw Mode Exposure Record Telemetry (Continued)
--------------------------------------------------------------------------------------------------------- Req Field Description--------------------------------------------------------------------------------------------------------- 17-1 Packet Synch This field marks the start of the ACIS packet within the telemetry stream. 17-2 Packet Word This field indicates the number of 32-bit words in the packet. Length 17-3 Format Tag This field indicates the type of data contained in the telemetry packet, in this case, a Timed Exposure Raw Mode Exposure Record packet. 17-4 Packet Sequence This field contains the packet sequence number. Number 17-5 Science Run Start This identifies the start time of the science run using the ACIS time- Time stamp latched by the hardware at the start of the run. 17-6 Timed Exposure This item is a copy of the Timed Exposure Parameter Block Identifier Parameter Block used to configure the current run. Identifier 17-7 Window Block This item is a copy of the Window List Parameter Block Identifier used Identifier to configure the current run. If no Window List was specified, this field will be 0xffffffff. 17-8 CCD/FEP Identi These identify which CCD and FEP produced the exposure. fiers 17-9 FEP Timestamp This is the value of the FEP time-stamp counter when the exposure arrived from the CCD. 17-10 Exposure Number This uniquely identifies the clocked exposure during the run. This field must allow the ground to uniquely identify over 2.6 million exposures during a single run (0.1 second exposures for 72 hours). 17-11 Pixel Count This specifies the total # of pixels sent for the exposure from the CCD. ---------------------------------------------------------------------------------------------------------
The Back End software produces "Histogram Mode" telemetry when the Front End Processor is in "Histogram Mode," as indicated by the "Front End Selection Algorithm."
Table 18 describes the content of the CCD data telemetry and Table 19 describes the exposure record telemetry when in this mode. Zero or more data packets, followed by one Histogram Record packet are sent for each active CCD.
TABLE 18. Histogram Data Content (Continued)
------------------------------------------------------------------------------------------------------ Req Field Description------------------------------------------------------------------------------------------------------ 18-1 Packet Synch This field marks the start of the ACIS packet within the telemetry stream. 18-2 Packet Word This field indicates the number of 32-bit words in the packet. Length 18-3 Format Tag This field indicates the type of data contained in the telemetry packet, in this case a Timed Exposure Histogram data packet. 18-4 Packet This field contains the packet sequence number. Sequence Num ber 18-5 CCD/FEP Iden These identify which CCD and FEP produced the histogram. tifiers 18-6 Data Packet Since data from different CCDs can be interleaved, and also mixed with Number other kinds of packets, this field sequentially numbers the data packet for a given CCD exposure. 18-7 Output Node This identifies from which output node the histogram was formed. Identifier 18-8 Starting Bin This item is the index of the first histogram bin contained within the Number packet. 18-9 Array of histo This consists of an array of adjacent histogram bins. Each bin contains a gram counts count of the number of pixels from the specified output node whose pulse height corresponded to the bin index. ------------------------------------------------------------------------------------------------------
TABLE 19. Histogram Record Content (Continued)
--------------------------------------------------------------------------------------------------------- Req Field Description--------------------------------------------------------------------------------------------------------- 19-1 Packet Synch This field marks the start of the ACIS packet within the telemetry stream. 19-2 Packet Word This field indicates the number of 32-bit words in the packet. Length 19-3 Format Tag This field indicates the type of data contained in the telemetry packet, in this case a Timed Exposure Histogram Record packet. 19-4 Packet Sequence This field contains the packet sequence number. Number 19-5 Science Run Start This identifies the start time of the science run using the ACIS time- Time stamp latched by the hardware at the start of the run. 19-6 Timed Exposure This item is a copy of the Timed Exposure Parameter Block Identifier Parameter Block used to configure the current run. Identifier 19-7 CCD/FEP Identi These identify which CCD and FEP produced the histogram. fiers 19-8 First Exposure This uniquely identifies the first exposure acquired into the histogram Number during the run. This field must allow the ground to uniquely identify over 130,000 exposures during a single run (2 second exposures for 72 hours). 19-9 Last Exposure This uniquely identifies the last exposure acquired into the histogram Number during the run. This field must allow the ground to uniquely identify over 130,000 exposures during a single run (2 second exposures for 72 hours). 19-10 Output Node This identifies from which output node the histogram was formed. Identifier 19-11 Number of Expo This identifies the total number of exposures acquired in the histogram. sures Processed Since one or more exposures may be dropped or ignored, this may be dif ferent than the "Exposure Number" listed above. 19-12 Overclock Values These contain the minimum, maximum, mean and variance of the over clock values computed while the histogram was been formed. ---------------------------------------------------------------------------------------------------------
The Back End software produces Faint Mode event telemetry when the "Event List Packing Code" field of the Timed Exposure Parameter Block indicates a "Faint" packing mode. This mode applies only when the Front End Processor is in "Event-finding mode," as indicated by the "Front End Selection Algorithm." In this mode, sets of distinct events are telemetered as 3x3 arrays of pixels. Table 20 illustrates the content of the Faint-Mode event list data packet, and Table 21 illustrates the content of an Exposure Record used with Faint-Mode.
TABLE 20. Faint Mode Exposure Event Data (Continued)
--------------------------------------------------------------------------------------------------------- Req Field Description--------------------------------------------------------------------------------------------------------- 20-1 Packet Synch This field marks the start of the ACIS packet within the telemetry stream. 20-2 Packet Word This field indicates the number of 32-bit words in the packet. Length 20-3 Format Tag This field indicates the type of data contained in the telemetry packet, in this case a Timed Exposure Faint-Mode data packet. 20-4 Packet Sequence This field is used to sequentially number all ACIS telemetry packets. Number 20-5 CCD/FEP Identi These identify which CCD and FEP produced the event fiers 20-6 Data Packet Num Since data from different CCDs can be interleaved, and also mixed with ber other kinds of packets, this field sequentially numbers the data packet for a given CCD exposure. The remainder of the packet contains zero or more of the following: 20-7 Position Identifier This identifies the row and column of the center pixel of the event 20-8 Uncorrected These items are the measured (uncorrected) pulse heights of the 9 pixels Pulse Heights. of the event's 3x3 array. ---------------------------------------------------------------------------------------------------------
TABLE 21. Faint Mode Exposure Record Content (Continued)
--------------------------------------------------------------------------------------------------------- Req Field Description--------------------------------------------------------------------------------------------------------- 21-1 Packet Synch This field marks the start of the ACIS packet within the telemetry stream. 21-2 Packet Word This field indicates the number of 32-bit words in the packet. Length 21-3 Format Tag This field indicates the type of data contained in the telemetry packet, in this case, a Faint-Mode Exposure Record packet. 21-4 Packet Sequence This field is used to sequentially number all ACIS telemetry packets. Number 21-5 Science Run Start This identifies the start time of the science run, as specified by the DEA Time latched ACIS time-stamp. 21-6 Timed Exposure This item is a copy of the Timed Exposure Parameter Block Identifier Parameter Block used to configure the current run. Identifier 21-7 Window Block This item is a copy of the Window List Parameter Block Identifier used Identifier to configure the current run. If no Window List was specified, this field will be 0xffffffff. 21-8 Bias Start Time This identifies the start time of the most recently computed bias map, relative to when the DEA interface latched the ACIS time-stamp at the start of the bias computation. 21-9 Bias Parameter This item is a copy of the Parameter Block Identifier used to configure Block Identifier the most recent bias map computation. 21-10 CCD/FEP Identi These identify which CCD and FEP produced the exposure fiers 21-11 FEP Timestamp This is the value of the FEP time-stamp counter when the exposure arrived from the CCD. 21-12 Exposure Number This uniquely identifies the clocked exposure during the run. This field must allow the ground to uniquely identify over 130,000 exposures dur ing a single run (2 second exposures for 72 hours). 21-13 Number of This specifies the number of events being telemetered in the exposure. Events Teleme This field must accommodate at least 64K events. tered 21-14 Number of Pixels These specify the total number of pixels from a CCD whose pulse above threshold heights were above their respective spatial thresholds during the expo sure. 21-15 Number of These specify the number of events which were discarded due to their Events Discarded amplitude. This field must accommodate the total number of pixels by Amplitude within the CCDs. 21-16 Number of These specify the number of events which were discarded due to their Events Discarded "grade." by Grade 21-17 Number of These specify the number of events which were discarded by the pro Events Discarded cessing windows. by Window 21-18 Overclock Levels These identify the overclock levels used for the thresholding during the exposure which produced the events. 21-19 Bias Map Parity This specifies the total number of pixels disabled due to parity hits. Hit Count ---------------------------------------------------------------------------------------------------------
The Back End software produces Faint Bias Mode event telemetry when the "Event List Packing Code" field of the Timed Exposure Parameter Block indicates a "Faint-Bias" packing mode. This mode applies only when the Front End Processor is in "Event-finding mode," as indicated by the "Front End Selection Algorithm." In this mode, sets of distinct events are telemetered as 3x3 arrays of pixels along with the corresponding pixel bias values. Table 22 illustrates the content of the Faint-Bias Mode event list data packet, and Table 23 illustrates the content of a Faint-Bias Mode Exposure Record.
TABLE 22. Faint-Bias Mode Exposure Data Content (Continued)
---------------------------------------------------------------------------------------------------- Req Field Description---------------------------------------------------------------------------------------------------- 22-1 Packet Synch This field marks the start of the ACIS packet within the telem etry stream. 22-2 Packet Word Length This field indicates the number of 32-bit words in the packet. 22-3 Format Tag This field indicates the type of data contained in the telemetry packet, in this case a Timed Exposure Faint-Bias Mode data packet. 22-4 Packet Sequence Number This field is used to sequentially number all ACIS telemetry packets. 22-5 CCD/FEP Identifiers These identify which CCD and FEP produced the event 22-6 Data Packet Number Since data from different CCDs can be interleaved, and also mixed with other kinds of packets, this field sequentially num bers the data packet for a given CCD exposure. The remainder of the packet contains zero or more of the following: 22-7 Position Identifier This identifies the row and column of the center pixel of the event 22-8 Uncorrected Pulse Heights These items are the measured (uncorrected) pulse heights of the 9 pixels of the event's 3x3 array. 22-9 Pixel Bias Map Values These are the computed pixel bias values associated with the event's 9 pixels. ----------------------------------------------------------------------------------------------------
TABLE 23. Faint-Bias Mode Exposure Record Content (Continued)
------------------------------------------------------------------------------------------------------- Req Field Description------------------------------------------------------------------------------------------------------- 23-1 Packet Synch This field marks the start of the ACIS packet within the telem etry stream. 23-2 Packet Word Length This field indicates the number of 32-bit words in the packet. 23-3 Format Tag This field indicates the type of data contained in the telemetry packet, in this case, a Faint-Bias Mode Exposure Record packet. 23-4 Packet Sequence Number This field is used to sequentially number all ACIS telemetry packets. 23-5 Science Run Start Time This identifies the start time of the science run, as specified by the DEA latched ACIS time-stamp. 23-6 Timed Exposure Parameter This item is a copy of the Timed Exposure Parameter Block Block Identifier Identifier used to configure the current run. 23-7 Window Block Identifier This item is a copy of the Window List Parameter Block Iden tifier used to configure the current run. If no Window List was specified, this field will be 0xffffffff. 23-8 Bias Start Time This identifies the start time of the most recently computed bias map, relative to when the DEA interface latched the ACIS time-stamp at the start of the bias computation. 23-9 Bias Parameter Block Iden This item is a copy of the Parameter Block Identifier used to tifier configure the most recent bias map computation. 23-10 CCD/FEP Identifiers These identify which CCD and FEP produced the exposure 23-11 FEP Timestamp This is the value of the FEP time-stamp counter when the exposure arrived from the CCD. 23-12 Exposure Number This uniquely identifies the clocked exposure during the run. This field must allow the ground to uniquely identify over 130,000 exposures during a single run (2 second exposures for 72 hours). 23-13 Number of Events Teleme This specifies the number of events being telemetered in the tered exposure. This field must accommodate at least 64K events. 23-14 Number of Pixels above These specify the total number of pixels from a CCD whose threshold pulse heights were above their respective spatial thresholds during the exposure. 23-15 Number of Events Dis These specifies the number of events which were discarded carded by Amplitude due to their amplitude. This field must accommodate the total number of pixels within the CCDs. 23-16 Number of Events Dis These specify the number of events which were discarded due carded by Grade to their "grade." 23-17 Number of Events Dis These specify the number of events which were discarded by carded by Window the processing windows. 23-18 Overclock Levels These identify the overclock levels used for the thresholding during the exposure which produced the events. 23-19 Bias Map Parity Hit Count This specifies the total number of pixels disabled due to parity hits in the corresponding bias map location. 23-20 Initial Overclock This is the initial overclock added to the pixel bias map val ues. (NOTE: This should have the same value for the length of the entire run). -------------------------------------------------------------------------------------------------------
The Back End software produces Graded event telemetry when the "Event List Packing Code" field of the Timed Exposure Parameter Block indicates a "Graded" packing mode. This mode applies only when the Front End Processor is in "3x3 Mode," as indicated by the "Front End Selection Algorithm." In this mode, sets of distinct events are telemetered using reduced event amplitude information and event "Grade" (see Section 3.2.2.3.16 on page 66). The content of an Exposure Record in this mode is identical to that used by Faint Mode Telemetry (see Table 21). The event data content, however, is different. Table 24 illustrates the content of the event lists telemetered when in this mode.
TABLE 24. Graded Event Telemetry Content (Continued)
---------------------------------------------------------------------------------------------------------- Req Item Description---------------------------------------------------------------------------------------------------------- 24-1 Packet Synch This field marks the start of the ACIS packet within the telemetry stream. 24-2 Packet Word This field indicates the number of 32-bit words in the packet. Length 24-3 Format Tag This field indicates the type of data contained in the telemetry packet, in this case a Timed Exposure Graded Event Mode data packet. 24-4 Packet Sequence This field is used to sequentially number all ACIS telemetry packets. Number 24-5 CCD/FEP Identi These identify which CCD and FEP produced the event fiers 24-6 Data Packet Num Since data from different CCDs can be interleaved, and also mixed with ber other kinds of packets, this field sequentially numbers the data packet for a given CCD exposure. The remainder of the packet contains zero or more of the following: 24-7 Position Identifier This identifies the row and column of the center pixel of the event 24-8 Event Amplitude This identifies the amplitude of the event (see Section 3.2.2.3.14 on page 65). 24-9 Grade Code This identifies the Grade Code of the event. The computation of an event's Grade Code is described by Section 3.2.2.3.16 on page 66. 24-10 Corner Pulse This is the mean, rounded to the nearest integer, of the corrected pulse- Height Mean heights of those corner pixels in the 3x3 event array which are below their corresponding split threshold in two's complement. ----------------------------------------------------------------------------------------------------------
When configured to re-compute and telemeter the bias maps, the instrument software shall trickle the pixel bias maps from each of the active Front End Processors, into the telemetry stream. The algorithm used shall ensure a 10% minimum utilization of the telemetry stream. The algorithm is permitted to consume under-utilized telemetry. The pixel bias map values may be compressed in a number of rows using a compression scheme which utilizes patchable, pre-defined CCD-specific compression tables.
TABLE 25. Pixel Bias Map Telemetry Content (Continued)
---------------------------------------------------------------------------------------------------------- Req Item Description---------------------------------------------------------------------------------------------------------- 25-1 Packet Synch This field marks the start of the ACIS packet within the telemetry stream. 25-2 Packet Word This field indicates the number of 32-bit words in the packet. Length 25-3 Format Tag This field indicates the type of data contained in the telemetry packet, in this case a Timed Exposure Pixel Bias Map packet. 25-4 Packet Sequence This field is used to sequentially number all ACIS telemetry packets. Number 25-5 Bias Start Time This identifies the start time of the bias run, as specified by the DEA latched ACIS time-stamp. 25-6 Timed Exposure This item is a copy of the Timed Exposure Parameter Block Identifier used Parameter Block to configure the current run. Identifier 25-7 CCD/FEP Identi These identify which CCD and FEP produced the bias map fiers 25-8 Bias Packet Since bias data from different CCDs can be interleaved, and also mixed Number with other kinds of packets, this field sequentially numbers the data packet for a given CCD exposure. 25-9 Initial Overclock This specifies the initial overclock value used to correct subsequent expo sures for changes in average overclock. 25-10 Compression These fields indicate whether or not the data has been compress, and if so, Selection Flag by which compression table. and Table Selec tion 25-11 Row Identifier This specifies which row of the pixel bias map is being sent first within the packet. Since the rows are being sent in reverse order, subsequent rows in the packet are those consecutively closer to the Framestore. 25-12 Row Count This specifies the number of rows being sent. 25-13 Pixels Per Row This specifies the number of pixel bias values per row The remainder of the packet contains zero or more of the following: 25-14 Compressed This is a bit-packed array of compressed Bias map data, starting from Bias Map Data "Row Identifier" and encompassing "Row Count". In order to send the bias map rows near the most likely focal point first, the rows are sent in reverse order, starting with those farthest from the Framestore. The col umn order is NOT reversed. ----------------------------------------------------------------------------------------------------------
In the process of producing events during an exposure. the Front End Processors detect and forward pixel bias map parity errors to the Back End Processor. The Back End accumulates these errors into one or more parity error telemetry packets, and posts these packet to telemetry as they fill, or at the end of each exposure if a partially filled data packet remains. The content of these packets is illustrated in Table 26.
TABLE 26. Bias Map Parity Error Telemetry Content (Continued)
------------------------------------------------------------------------------------------------- Req Item Description------------------------------------------------------------------------------------------------- 26-1 Packet Synch This field marks the start of the ACIS packet within the teleme try stream. 26-2 Packet Word Length This field indicates the number of 32-bit words in the packet. 26-3 Format Tag This field indicates the type of data contained in the telemetry packet, in this case a Timed Exposure Parity Error data packet. 26-4 Packet Sequence Number This field is used to sequentially number all ACIS telemetry packets. 26-5 CCD/FEP Identifiers These identify which CCD and FEP produced the event 26-6 Data Packet Number Since data from different CCDs can be interleaved, and also mixed with other kinds of packets, this field sequentially num bers the parity error packet for a given CCD exposure. The remainder of the packet contains zero or more of the following: 26-7 Position Identifier This identifies the row and column of the bias error. 26-8 Corrupted Value This a copy of the corrupted pixel bias map value. -------------------------------------------------------------------------------------------------
The diagnostic extensions of this mode are described in more detail by Section 3.2.5 on page 97. The extensions specific to Timed Exposure Mode include:
Next Chapter