Document Number:

 

Release Date:

 

LAT-TD-04631-02

March 8, 2005

Authors:

Eduardo do Couto e Silva,

Lee Steele

Supersedes:

Rev 1

 

GLAST LAT Technical Document

 

Subsystem/Office:

 

Integration and Test

Document Title:

 

GLAST LAT Instrument Data Analysis Primer

 

 

 

 

Gamma Ray Large Area Space Telescope (GLAST)

Large Area Telescope (LAT)

Integration & Test (I&T)

Instrument Data Analysis Primer


 

Change History Log

Revision

Effective Date

Description of Changes

1

27 September 2004

Initial release

2

8 March  2005

Revised Fig. 6, Fig. 12, Table 2  and text to reflect change in Recon file plane numbering scheme.

 

 

 

 

 

 

 


Contents

1.       Introduction. 9

1.1.    Hardware Overview. 9

1.2.    Overview of Data Taking During LAT Integration. 10

2.       Geometry and Numbering Scheme. 12

2.1.    ACD Geometry and Numbering Scheme. 12

2.2.    TKR Geometry and Numbering Scheme. 12

2.2.1.     CAL Geometry and Numbering Scheme. 15

3.       Readout Sequence. 17

3.1.    ACD Readout Sequence. 17

3.2.    TKR Readout Sequence. 17

3.2.1.     Mapping across TKR Physical and Electronic Space. 17

3.3.    CAL Readout Sequence. 19

3.3.1.     Mapping across CAL Physical and Electronic Space. 21

4.       Global Trigger and Dead Time. 22

4.1.    Time Delays. 23

4.2.    Dead Time. 26

4.3.    Live Time. 26

5.       Nominal Register Settings. 27

5.1.    ACD Nominal Register Settings. 27

5.1.1.     ACD Veto (hit) Threshold Discriminator 27

5.1.2.     ACD Zero-Suppression Threshold. 27

5.1.3.     ACD Time Delays. 27

5.1.4.     ACD Known Features. 27

5.2.    TKR Nominal Register Settings. 27

5.2.1.     TKR Hit Threshold Discriminator (DAC) 27

5.2.2.     Some Relevant TKR GTFE and GTRC Registers. 29

5.2.3.     TKR Time Delays. 29

5.2.4.     TKR Known Features. 29

5.3.    CAL Register Nominal Settings. 30

5.3.1.     CAL Time Delays. 31

5.3.2.     CAL Known Features. 32

6.       Calibrations. 33

6.1.    ACD Calibration. 33

6.2.    TKR Calibration. 33

6.3.    CAL Calibration. 34

7.       Event Data. 35

7.1.    Event Contributions. 35

7.1.1.     AEM.. 35

7.1.2.     TEM.. 35

7.1.3.     GEM.. 36

7.2.    Data Analysis Files. 36

7.2.1.     LDF.FITS. 37

7.2.2.     Digi.root 37

7.2.3.     Recon.root 38

7.2.4.     MC.root 38

7.2.5.     Merit.root 38

7.2.6.     SVAC.root 38

7.2.7.     Reports: Configuration and SVAC. 38

7.3.    Data Processing Steps. 38

8.       TKR Reconstruction (To Be Re-written for new TKR Recon) 40

8.1.    Clustering (TrkClusterAlg) 42

8.2.    Track Finding (TrkFindAlg) 42

8.3.    Track Fitting (TkrFitTrackAlg) 43

8.4.    Vertexing (TkrVertexAlg) 44

8.5.    Track Hypothesis for Integration and Test 45

9.       CAL Reconstruction. 46

10.     Monte Carlo Simulations. 47

10.1.      Low Energy Photon Source. 47

10.2.      Cosmic Ray Source. 47

10.3.      Thresholds and Noise. 48

11.     Data Analysis. 49

11.1.      Identification of Minimum Ionizing Particles. 49

11.2.      CALTowerGap. 50

11.3.      Timing Analysis. 50

12.     Resources. 51

Appendix  A - eLog. 53

A.1.    Shift Logbook Index 53

A.2.    Run Selection Index 54

A.3.    List Runs. 57

 


List of Illustrations

Figure 1: The LAT. 9

Figure 2: Principal LAT Components (Block Diagram) 10

Figure 3:  Tower Placement for Cosmic Ray Data Taking. 11

Figure 4: Grid Tower Positions for Monte Carlo Simulations. 11

Figure 5: LAT Tower Numbering and Grid Coordinate System.. 12

Figure 6: TKR Tower Numbering Scheme. 13

Figure 7: TKR Plane Physical Details (X-View) 14

Figure 8: CAL Crystal Layer Numbering and Orientation. 15

Figure 9: CAL Module Cross Section. 15

Figure 10: Zoom of Region between Two Adjacent CAL Modules (3 layers shown) 16

Figure 11: CAL Crystal Dimensions. 16

Figure 12: The Four Sides of the TKR Tower with Cables. “X” or “Y” Means Measured Coordinate. 19

Figure 13: CAL FEE Simplified Schematic Diagram.. 20

Figure 14: CAL Channel Signal Range Energy Overlap. 21

Figure 15: The Four Sides of the CAL Module with Cables. 21

Figure 16: Trigger Studies of Real Triggers vs. MC Simulations. 23

Figure 17: Conceptual Trigger Delay Adjustments Diagram (GEM Inputs) 24

Figure 18: Conceptual Trigger Delay Adjustments Diagram (GEM Outputs) 25

Figure 19: TKR FEE Readout Channel Splitting. 28

Figure 20: Shaper Output Time over Threshold. 29

Figure 21: TDS Input and Output 35

Figure 22: Data Analysis Process Flow. 39

Figure 23: Four TKR Reconstruction Steps (Block Diagram) 40

Figure 24: Four Steps of TKR Reconstruction (Illustrated) 41

Figure 25: Iterative TKR Reconstruction Algorithms (Block Diagram) 41

Figure 26: Combinatoric Pattern Recognition: ComboFindTrack Tool 43

Figure 27: Simulation of VDG Gammas – Simulated Particle Source Generation. 47

Figure 28: Particle Flux vs. Kinetic Energy for surface_muon Source. 48

Figure 29: TrkTrackLength Example of Easily Misinterpreted Data. 49

Figure 30: A Charged Particle’s Path is Parallel to the Z Axis and only Strikes Every Other Crystal 50

Figure 31: GLAST Shift Log Index 53

Figure 32: Logbook Shift Run Info. 54

Figure 33: List Runs Table. 57


List of Tables

 

Table 1: Detector Readout Channels. 17

Table 2: TKR Mapping between Physical and Electronics Space. Note that “X” or “Y” Means Measured Coordinate, T= Top Side of Tray, B=Bottom Side of Tray. 18

Table 3: CAL FEE Signal Gain Characteristics. 20

Table 4: CAL Mapping between Physical and Electronics Space. 21

Table 5: Number of Timing Delay Registers. 23

Table 6: ACD Delay Register Nominal Settings. 27

Table 7: TKR GTFE and GTRC Registers. 29

Table 8: TKR Delay Registers Nominal Settings. 29

Table 9: TKR Electronics Known Features. 30

Table 10: CAL Registers for Gain, Triggering and Data Volume. 30

Table 11: CAL Thresholds Nominal Settings. 31

Table 12: CAL Delay Registers Nominal Settings. 31

Table 13: CAL Known Features. 32

Table 14: CAL and TKR Trigger Primitive Data. 35

Table 15: GEM Event Contribution. 36

Table 16: Data Analysis Files Locations. 37

Table 17: TKR Reconstruction Clustering Methods. 42

Table 18: TKR Reconstruction Combinatoric Track Finding Methods. 43

Table 19: TKR Reconstruction Vertexing Tools. 44

Table 20: Logbook Shift Run Menu Fields. 54

Table 21: Logbook Shift Run Menu Active Buttons. 56

Table 22: List Runs Fields. 57


1.       Introduction

This document is intended to provide the LAT collaborators with sufficient information to perform data analysis during LAT integration. It is intended for users who are familiar with the LAT instrument, however a brief overview is provided.

Most of the information in this document is either copied from the website of Instrument Analysis Workshop presentations, or existing LATDocs documents. A list of references is provided in Resources, section 12.

1.1.         Hardware Overview

The Large Array Telescope (LAT) is an integrated instrument consisting of 16 towers set into a 4x4 grid. Each tower consists of a Tracker (TKR), Calorimeter (CAL), and Tower Electronics Module (TEM). The LAT is shown in  Figure 1. The 16 towers are surrounded by an Anti-Coincidence Detector (ACD) which is surrounded by a micro-meteorite shield.

 

Figure 1: The LAT

The following paragraphs provide a brief description of how the major components are used during pre-launch tests and are shown in Figure 2.

The ACD is mostly used to either identify charged particles for cosmic ray calibration runs, or to reject charged particles during Van de Graaff photon calibration runs.

The TKR’s function is to reconstruct the original direction of travel of either incoming photons (from 18 MeV photons from a Van de Graaff generator) or of charged particles (cosmic rays).

The CAL’s function is to measure the energy deposited by incoming charged particles, either from photon pair creation or cosmic rays.

The TEM assembles trigger primitives from the TKR and CAL (or simulated input) to determine whether or not there has been an event in the LAT. If so, it alerts the GLAST Electronics Module (GEM). It also communicates with the Event Builder Module. Trigger parameters are stored in the TEM.

The GEM (also referred to as the GLAST LAT Trigger - GLT) responds to a TEM’s message that an event has been detected and decides whether or not to generate a trigger. The GEM is an important component when performing Dead Time and Trigger analyses.

The Anti-Coincidence Detector Electronics Module (AEM) performs the same function for the ACD as the TEM does for the TKR and CAL detectors.

The Event Builder Module (EBM) communicates with the GEM, TEM and AEM.

The Global-trigger/ACD-module/Signal-distribution Unit (GASU) performs the highest logic level of event decision making, and comprises the AEM, GEM and EBM.

The Power Distribution Unit (PDU) supplies DC current to operate the electronics.

Figure 2: Principal LAT Components (Block Diagram)

NOTE: Real detectors (ACD, TKR, CAL) can be replaced by simulated input that mimics the behavior of the physical detectors beginning at the cables that readout the detectors. Although the Front End Electronics (FEE) are not simulated, a pattern can be created to generate events.

1.2.         Overview of Data Taking During LAT Integration

Data taking with cosmic ray muons will occur with 1, 2, 4, 6, 8, 10, 12, 14 and 16 Flight Modules (FMs) installed in the LAT grid. The first position filled is position #0. The second position filled is #4. 15 hours of cosmic ray data taking (plus one hour with zero suppression OFF) occurs every time towers are added to the LAT. 16 hours of data taking with Van de Graaff photons occurs for tower A. Please refer to LAT-MD-00575 for detailed information on data taking during integration. Figure 3 shows the tower positions filled in each data taking configuration. Note that shaded squares indicate a tower installed in the grid.

For each hardware configuration there will be a baseline cosmic ray data-taking run for which the hardware is configured with nominal settings (please refer to the Nominal Register Settings in section 5) for ground analysis for the integrated hardware (towers).

Figure 3:  Tower Placement for Cosmic Ray Data Taking

Monte Carlo simulations for Flight Modules will be performed for 1, 2, and 8 towers, and the fully integrated LAT grid, as shown in Figure 4.

 

Figure 4: Grid Tower Positions for Monte Carlo Simulations


2.       Geometry and Numbering Scheme

The global instrument coordinate system for the LAT is consistent with the coordinate system for the observatory. It is a right-handed coordinate system with the Y-axis parallel to the solar panel axis, the Z axis normal to the planes of the TKR, CAL, and Grid (i.e., parallel to the “bore sight”), and the X axis is mutually perpendicular to Y and Z. The positive Z-axis points from the CAL to the TKR. Particles entering the instrument at normal incidence are thus oriented along the -Z direction (please see Figure 5).

The point X=Y=0 is at the center of the Grid. The Z=0 plane is at the top face of the Grid, between the TKR and CAL units on the TKR side of the Grid. The TKR silicon plane closest to Z = 0 is at Z = +33 mm; the crystal plane closest to Z = 0 is at Z = -46 mm.

Figure 5: LAT Tower Numbering and Grid Coordinate System

2.1.         ACD Geometry and Numbering Scheme

The active elements of the ACD consist of 89 tiles and 9 ribbons. (A figure will be added later.)

2.2.         TKR Geometry and Numbering Scheme

The tracker is made up of 19 trays comprising 36 planes as shown in Figure 6.  The 36 planes are mated into 18 layers.

The TKR trays are numbered in increasing order with increasing Z. Each tray has two active planes, except the top half of the top tray (+Z) and bottom half of the bottom tray.

A tray measures in either the X or Y direction, i.e., has an X or Y view. To get X and Y information, planes from two adjacent trays are electronically combined. Mated X and Y planes are about 2 mm apart. This arrangement leaves the top-most and bottom-most planes without a partner and without silicon detectors.

A tray with detector strips physically parallel to the Y axis is an X tray: it measures the X coordinate (has an X view) and is called an X tray. Most planes have an embedded tungsten foil for g conversion: The top 12 X and Y pairs have a thin foil (3% of X0), the next four have a thick foil (18% of X0), and the bottom two X and Y pairs have no tungsten.

The active region of each TKR plane is comprised of 16 square Silicon Strip Detectors (SSDs - please see Figure 7). Each SSD has 384 conducting strips. Four SSDs are end-joined to make a ladder with the four SSDs in a given ladder joined mechanically and electrically to make 384 long strips. Four ladders per plane laid side-by-side make up a total of 1536 strips per plane. Each plane is about 360 mm by 360 mm in area.

 

Figure 6: TKR Tower Numbering Scheme.

Figure 7: TKR Plane Physical Details (X-View) 

2.2.1.      CAL Geometry and Numbering Scheme

Each CAL module is made up of 96 crystals oriented in a hodoscopic configuration of 8 layers of 12 crystals each. In contrast to a TKR plane, a CAL crystal makes its coordinate measurement along its principal axis: an X crystal has its principal axis along the X direction, as shown in Figure 8.

Each crystal has two PIN diodes at each end for reading out the signal. Each PIN diode (at either end) reads out for either the low or high energy measurement. The low energy PIN has an area four times greater than the high energy PIN.

The CAL layers are numbered from 0 – 7 in increasing order with decreasing Z. The CAL layers closest to the TKR is plane 0. CAL layers 0 has X crystals; CAL layers 7 has Y crystals. Each CAL crystal is read from each end, and each crystal end is either plus or minus: the end with the larger value of the coordinate is the “plus” end and the end with the smaller value of the coordinate is the “minus.”

Figure 8: CAL Crystal Layer Numbering and Orientation

Figure 10 shows an accurate representation of a CAL module.

Figure 9: CAL Module Cross Section

Figure 10 shows a close-up view of the displacement of CAL crystal ends of two adjacent CAL modules. It is important to note that crystal ends that face each other are at a different spacing than the closest crystals in adjacent modules that are parallel to each other.

 

Figure 10: Zoom of Region between Two Adjacent CAL Modules (3 layers shown)

The CAL crystal profile is shown in Figure 11 along with the dimensions of a CAL crystal, including its carbon fiber enclosure.

Figure 11: CAL Crystal Dimensions


3.       Readout Sequence

Table 1 lists the number of readout channels (active elements) for the ACD, TKR and CAL in a 1, 2, 4, 8, and the full LAT configuration. Note that the ACD front-end PCBs actually have 216 channels, but because each tile is read by 2 PMTs that are assembled with a logical OR, the number of tiles that can be read out is actually 108. With a total number of  97 tiles and ribbons, some channels are not used.

Table 1: Detector Readout Channels

 

ACD

TKR

CAL

 

Tiles/Channels

Ribbons/Channels

Planes/ Channels

Crystals/ Channels

1 Tower

 

 

36 / 55,296

96 / 384

2 Tower

 

 

72 / 110,592

192 / 768

4 Tower

 

 

144 / 221,184

384 / 1536

8 Tower

 

 

288 / 442,368

768 / 3072

LAT

89/178

8/16

576 / 884,736

1536 / 6144

3.1.         ACD Readout Sequence

To be written.

3.2.         TKR Readout Sequence

Data from each silicon plane is read out by 24 GTFE ASICs located in a Multi-Chip-Module (MCM), controlled by 2 GLAST Tracker Readout Controllers (GTRC). There are 2 (GTRC 0 and GTRC 1) per plane or 4 per tray, except trays 0 and 18 (please see Figure 19). The two GTRCs are situated at the edge of the plane. GTRC 0 (RC0) is defined as being closest to where strip 0 is located. RC1 is defined as where strip 1535 is located. The default mode is to read from both ends. The readouts of a TKR are carried through eight cables (please see Figure 12).

If any shaper-out signal of a channel in a GTFE chip is over threshold, a trigger request (TREQ) signal is issued and transferred to the TEM, which then checks the trigger status. If a trigger condition (3-in-a-row) is satisfied, the TEM sends the trigger request acknowledge (TACK) signal to all planes to latch hit strip data into GTFE event buffers. Each GTFE has 4 event buffers. Note that it is not the TREQ but the TACK signal that starts the Time over Threshold (ToT) counter in the GTRC.

The TEM sends the READ-OUT command and transfers event data from GTFE event buffers to GTRC event buffers. A GTRC event buffer is limited to 64 hit-strips. The GTRC has 2 buffers. The TEM sends the TOKEN signal and transfers event data from the GTRC event buffers to the TEM one plane at a time. The GTRC waits to send data until the process of the READ-OUT command finishes, and the ToT counter terminates. The ToT counter saturates at 1000 clock cycles (= 50 μs).  In a case where the ToT counter overflows, the GTRC will start to send data at the overflow point (1000 clock cycles).

3.2.1.      Mapping across TKR Physical and Electronic Space

The TKR mapping scheme is shown in Table 2. This table maps the TKR physical information in LDF, digi and recon files. Note that GTRC value is an address. These numbers are unique taken in combination with the cable controller. Figure 12 shows the four faces of the TKR. A typical user does not need to know the information in electronics space. This table is intended to serve as a reference for hardware-oriented people. Before using the Electronics Space information, always verify that it is the latest information in the TEM manual.

Table 2: TKR Mapping between Physical and Electronics Space. Note that “X” or “Y” Means Measured Coordinate, T= Top Side of Tray, B=Bottom Side of Tray

Physical Space (data analysis)

Electronics Space (command the instrument)

Tray  #

Digi file

Plane

Recon file

LDF file

bi-Layer

View

GTCC (Cable)

GTRC

18/B

35

17

Y

4,5

8

17/T

34

17

X

6,7

8

17/B

33

16

X

2,3

8

16/T

32

16

Y

0,1

8

16/B

31

15

Y

4,5

7

15/T

30

15

X

6,7

7

15/B

29

14

X

2,3

7

14/T

28

14

Y

0,1

7

14/B

27

13

Y

4,5

6

13/T

26

13

X

6,7

6

13/B

25

12

X

2,3

6

12/T

24

12

Y

0,1

6

12/B

23

11

Y

4,5

5

11/T

22

11

X

6,7

5

11/B

21

10

X

2,3

5

10/T

20

10

Y

0,1

5

10/B

19

9

Y

4,5

4

9/T

18

9

X

6,7

4

9/B

17

8

X

2,3

4

8/T

16

8

Y

0,1

4

8/B

15

7

Y

4,5

3

7/T

14

7

X

6,7

3

7/B

13

6

X

2,3

3

6/T

12

6

Y

0,1

3

6/B

11

5

Y

4,5

2

5/T

10

5

X

6,7

2

5/B

9

4

X

2,3

2

4/T

8

4

Y

0,1

2

4/B

7

3

Y

4,5

1

3/T

6

3

X

6,7

1

3/B

5

2

X

2,3

1

2/T

4

2

Y

0,1

1

2/B

3

1

Y

4,5

0

1/T

2

1

X

6,7

0

1/B

1

0

X

2,3

0

0/T

0

0

Y

0,1

0

Figure 12: The Four Sides of the TKR Tower with Cables. “X” or “Y” Means Measured Coordinate.

3.3.         CAL Readout Sequence

The CAL crystals are normally read from both ends through a total of four cables – each crystal is read out by two cables, one + and one -. Each crystal end has two PIN diodes, one large and one small, for low and high energy, respectively. Each crystal end (left and right, or + and -) has its own FEE pre-amplifier electronics assembly for a total of 192. Both low and high energy signals go through a pre amp and shaper then a Track and Hold gain multiplier and into an analog multiplexer and finally to the Analog to Digital Converter (ADC). Each row of crystals has a GCRC for each end for a total of 16. (There are 8 rows and each row is read out at both ends.) The 16 GCRCs feed four cables (and four GCCCs). The cables carry +X, -X, +Y and -Y signals to the GCCCs. The two X  GCCCs combine +X and -X information to produce X0, X1, X2, X3 signals; the Y GCCCs do the same to produce Y0, Y1, Y2, and Y3. A simplified FEE schematic diagram is shown in Figure 13. The calibration charge injection signal is fed to the front end of these pre amps.

Figure 13: CAL FEE Simplified Schematic Diagram

Table 3 shows the gain stages.

Table 3: CAL FEE Signal Gain Characteristics

Channel

Diode Discrimination

Pre Amp Gain

Track and Hold Gain

Signal Out of Track and Hold

Relative Gain

Low Energy

X 6

X 64

X 1

LEX1

64

X  8

LEX8

512

High Energy

X 1

X 1

X 1

HEX1

1

X 8

HEX8

8

Figure 14 shows how the energy ranges overlap. During cosmic ray data taking on the ground most of the events fall within the low range diodes.

Figure 14: CAL Channel Signal Range Energy Overlap

3.3.1.      Mapping across CAL Physical and Electronic Space

Table 4 maps the locations of layers and crystal ends between the electronics space and the physical space. A typical user does not need to know the information in electronics space. This table is intended to serve as a reference for hardware-oriented people. Before using the Electronics Space information, always verify that it is the latest information in the TEM manual.

Table 4: CAL Mapping between Physical and Electronics Space

Physical Space (Used for data analysis)

Electronics Space

(Used to command the instrument)

Layer # / Layer Type

Digi File Layer

Recon File Layer

LDF File

 

 

 

GCCC (Cable)

GCRC

0 / X

0

0

0,2

0

1 / Y

1

1

1,3

0

2 / X

2

2

0,2

1

3 / Y

3

3

1,3

1

4 / X

4

4

0,2

2

5 / Y

5

5

1,3

2

6 / X

6

6

0,2

3

7 / Y

7

7

1,3

3

The CAL readout cables and the associated GCCC are shown in Figure 15.

Figure 15: The Four Sides of the CAL Module with Cables


4.       Global Trigger and Dead Time

The Trigger inputs from the ACD, CAL and TKR are processed by the GEM. The GEM decides whether to read out an event, or not, based on all inputs received. Therefore trigger primitives can be issued but may lead to no data latched if the GEM decides that all necessary conditions have not been met.

Any of the three detectors (ACD, CAL and TKR) can issue a trigger request (TREQ) but data is only latched from the detector buffers if the GEM logic is satisfied that a trigger acknowledge (TACK) is warranted.

The ACD signals are usually used as a veto for charged particles. However, the CNO signal can be used to trigger on heavy nuclei (mostly used for calibrations). During ground testing we can only test the CNO signal through charge injection. Another useful concept one has to keep in mind is that Regions Of Interest (ROI) can be defined using the ACD tiles. So a number of tiles can be logically grouped for trigger/veto purposes.

The TKR trigger logic requires at least one hit above a predetermined threshold in three consecutive XY planes (six planes).

The CAL trigger logic requires at least one hit above a predetermined threshold in any of the crystals.  One can have a low energy trigger and a high energy trigger. For ground testing with cosmic rays the nominal CAL low-energy trigger is already too high and one must lower the discriminator thresholds or use the TKR trigger, instead.

During ground testing there will be a pattern generator (software) that will produce several trigger rates from a few Hz to tens of kHz. These solicited triggers will be overlaid with the nominal trigger conditions (e.g. TKR trigger) to study the behavior of the trigger and data flow system.

To trigger on a set of trigger inputs coming from different detectors, they must all fall within the same time coincidence window, but the system must not be busy (already reading an event). Note that during ground testing, in order to cope with rates greater than 1.5 kHz, one may have to prescale (discard) events. Note that an otherwise “triggerable” event may not be latched because it was prescaled away, or because the instrument was busy. Trigger information the GEM reports:

·        Number of triggered events not passing the prescalers (Prescale count – 24 bits)

·        Number of triggered events passing the prescalers, but lost due to the LAT being busy (Discarded counts – 24 bits)

·        Number of triggered events read out (Sent count – 16 bits)

For the TKR, the hit threshold is also the trigger threshold. The threshold is that six consecutive planes must be fired. Nominal trigger rate on the ground (no ACD) is roughly 25 Hz (TBR) for each TKR tower for cosmic ray analysis.

Trigger studies are performed as illustrated in Figure 16. Actual tracker trigger primitives are received by the TEM, which makes a trigger request, and then by the GEM, which evaluates the request and either does, or doesn’t, issue a trigger acknowledge. Trigger request information does not include which strip was fired (and initiated the trigger request). Therefore, it is useful to compare real trigger requests with simulated trigger requests.

When a tracker requests a trigger, the possibility exists that the charge will not be held when the data is latched. This should be investigated during timing studies.

Without Flight Software (FSW) there are possible timing problems, such as difficulty in relating GEM time counters to the Online 60Hz and 20 MHz event time stamps. The 20 MHz time stamp is applied when the event is moved from the Event Builder, and this time may increase as events queue up. Above 300 Hz the GEM Delta event time is usable. It is the time between event (n-1) and event n. It is a 16-bit counter that saturates at 3.2 ms.

 

Figure 16: Trigger Studies of Real Triggers vs. MC Simulations

4.1.         Time Delays

All three subsystems have preamps that put out shaped pulses that peak at different times after the entrance time (t0) for the particle into the LAT.

Any one of the three detectors can cause the GEM to issue a TACK which latches the buffered data of all three detectors regardless of whether or not they issued a trigger request. The purpose of the time delays is to ensure that the trigger window is open at the correct time to receive data from all the detectors at the peak of the input pulses for a given detector. Table 5 shows the different delays for the ACD, CAL and TKR. The table is for the entire ACD but just one of the 16 tower modules.

Table