ADVISORY:  This document has NOT been reviewed for export control and 
therefore may be subject to ITAR regulations or requirements. 
 
Table of Contents

1.0 INTRODUCTION    3
    The MESSENGER Mission    3
    The EPPS Investigation    3
    Status of EPS Characterization and Calibration    6
    Assessment of the EPS Instrument Performance Against Requirements    7

2.0 EPS INSTRUMENT DESCRIPTION    8
    EPS Sensor Description    8
    EPS Electronic Description    16

3.0 SCOPE OF CHARACTERIZATION AND CALIBRATION    19
    Flux, differential intensity and phase space density    20
    Definition of sensor transfer function and geometric factor    20
    Goals of the EPS characterization and calibration efforts    22

4.0 EXPECTED EPS PERFORMANCE AND SIMULATIONS    22
    Collimator Performance and Geometric Factor    23
    Ion Measurements    25
    Electron Measurements    31
    Summary of Anticipated EPS Performance    34

5.0 FLIGHT EPS CALIBRATION    34
    Test Set Up    34
    Directionality    34
    Calibration Beam Runs    37
    Spacecraft Integration and Test Calibrations: TOF Performance    43
    Summary of EPS flight-model calibrations    44

6.0 ENGINEERING MODEL CALIBRATIONS    44
    Ion Counting Efficiency    45
    TOF-Spread Performance    45
    Summary    46

7.0 FUTURE PLANS    47
    Engineering Model Plans    47
    In-flight Calibration    47

8.0 REFERENCES    49

APPENDICES    50


1.0 INTRODUCTION

The Energetic Particle Spectrometer (EPS) is one of two sensors that comprise 
the Energetic Particle and Plasma Spectrometer (EPPS) package on the MESSENGER
mission to Mercury.  The other sensor is the Fast Ion Plasma Sensor (FIPS).  
EPS measures the high-energy components of the in situ electrons (> 15 keV) 
and ions (>5 keV/nucleon), while FIPS measures the low energy components of 
the ion distributions (< 20 keV / charge).  Here we document the 
characterization and calibration of the EPS sensor, and describe future plans 
for finalizing the characterization of this sensor.

The MESSENGER Mission

The MErcury, Surface, Space ENvironment, GEochemistry, and Ranging (MESSENGER)
mission will send the first spacecraft to orbit the planet Mercury.  A 
miniaturized set of seven instruments, along with the spacecraft 
telecommunications system, provides the means of achieving the scientific 
objectives that motivate the mission. The payload includes a combined 
wide-angle and narrow-angle imaging system; ?-ray, neutron, and X-ray 
spectrometers for remote geochemical sensing; a vector magnetometer; a laser 
altimeter; a combined ultraviolet-visible and visible-infrared spectrometer to
detect atmospheric species and map mineralogical absorption features; and an 
energetic particle and plasma spectrometer to characterize ionized species in 
the magnetosphere.

Launched in August 2004, MESSENGER will orbit Mercury for one Earth year 
following the following events:

Earth Flyby (1):             Jul 29, 2005
Venus Flybys (2):             Oct 23, 2006 and Jun 4, 2007
Mercury Flybys (3):             Jan 14, 2008, Oct 6, 2008, and Sep 29, 2009
Mercury Orbit Insertion:         Mar 18, 2011
Primary Science Orbit Start:         Mar 20, 2011

The EPPS Investigation

The Energetic Particle and Plasma Sensor (EPPS) Instrument Package consists of
two separate physical packages, FIPS and EPS.  Both sensors communicate to the
MESSENGER spacecraft using the Event Processor Unit (EPU) computer located 
with the EPS sensor.  Figure 1a, 1b, and 1c show the placement of the EPS and 
FIPS heads on the spacecraft and the EPS and FIPS fields-of-view (FOV).  In 
these figure the sun is generally towards the –Y(s/c) axis and the imaging 
instruments point roughly in the +Z(s/c) direction.  EPS has a 160? x 12? FOV 
with its center pointing along the +Y(s/c) axis away from the sun, one edge 
reaching within 10? of the +Z(s/c), often in the general direction of Mercury,
and the other edge pointing within 10? of the –Z(s/c) axis, often in the 
“anti-Mercury” direction.

The overall EPPS instrument (including both EPS and FIPS) has a mass 
allocation of 3.07 kg and uses an allocated maximum (operational) power of 7.8
W (4.5 W of survival heater power).














In expanded form, the science goals of the EPPS instrument are: 

1) To help determine Mercury’s internal magnetic field by identifying some 
external current sources, identifying contaminating transients in the 
environment external to Mercury, and constraining the magnetic field topology 
by identifying field lines that connect to Mercury.
2) To characterize Mercury’s exosphere neutrals by measuring the ionized 
products that arise from them and identifying ions accelerated by Mercury’s 
magnetosphere.
3) To help determine the composition of the radar-reflective materials at 
Mercury’s poles and other materials arising from Mercury’s surface.
4) To determine the electrical properties of the crust/atmosphere/ 
space-environment interface.
5) To determine characteristics of the dynamics of the Mercurian magnetosphere
and its relationship to external drivers and internal conditions.
6) To measure interplanetary plasma properties in cruise and in Mercury’s 
vicinity.

These science goals tie directly back to the objectives listed in section 2.7 
of the Program Level Requirements (PLR) document for the MESSENGER mission.  
The goals relevant to EPPS are to:
         
[3] Provide the capability to distinguish between solar and planetary sources 
of magnetospheric species.
[5] Provide in situ detection of H+ or S+ enrichment over polar regions.
[6] Provide the composition of magnetospheric ions and pickup ions in the 
solar wind from sputtered neutrals.

EPS contributes to these goals by measuring:  i) spatial distributions of 
energetic ion pressures and other moments, ii) energetic ion and electron 
intensity versus time profiles for a broad range of energy and species 
channels, iii) the major ionic species that comprise the ions intensities (H, 
He, CNO, heavies) iv) electron and ion angular distributions with respect to 
the magnetic field and the Mercury-Sun system, v) the positions and time 
variability of particle boundaries within Mercury magnetosphere, and vi) the 
energy and intensity depositions of particles onto Mercury’s surface and 
exosphere. 

The requirements are summarized in section 4.2.2.8 of the PLR which states 
“EPS shall be capable of measuring the elemental species H, He, CNO, and Fe, 
over an energy range from 10.0 keV/nuc to at least 3 MeV total energy and 
electrons from 25 keV to at least 500 keV.”

Status of EPS Characterization and Calibration

The present (May 2004) characterization and calibration of EPS is less 
complete than is usual for an instrument that has been delivered to the 
spacecraft.  The process of developing a ~1-2 kg (EPS is ~1.5 kg) instrument 
from a typically 5-10 kg class of instrument has been exceedingly challenging,
and the instrument was delivered to the spacecraft with several known 
uncertainties in its performance.  The largest problems included:

1) Total time-of-flight spread ~ 4-5 ns rather than the anticipated < 2 ns 
(the 4-5 ns is still sufficient to achieve the required clean separation of H,
He, CNO, and Fe).
2) Efficiencies of double coincident signals (TOF plus E) much lower than 
expected (The efficiency of TOF alone – i.e. Low Energy Ions, is within the 
reasonable range of expectations).
3) Some cross talk (several percent) may occur in the classification of ion 
events within the FPGA ion species classifier (It appears that when two 
different ion species are present, H and O for example, some H ions may be 
assigned with the TOF of the O ion, and vice versa).

During the pre-delivery phase, the EPS team was faced with the quandary of 
either continuing to try to understand and fix the known performance problems 
or to just accept the performance as it is and to characterize it as best as 
we can.  Thus, much work remains in characterizing the EPS instrument.

Since the instrument delivery, a flight-like engineering model has been built,
and characterization of that unit is ongoing.  We have achieved better 
understanding of the performance of the instrument and have are focusing in on
the range of selectable high-voltage, reference-voltage, and discrimination 
values that optimize the performance of the EPS.   For example, with knowledge
gained with the engineering model we tuned the flight EPS during the 
Spacecraft Integration and Test and achieved a factor of 2 improvement in the 
TOF spread highlighted in item #1 above.  Also, much higher double coincident 
efficiencies (and by inference TOF-only efficiencies) have been achieved with 
new parametric settings on the engineering model than yet tried on the flight 
model.

These engineering model characterizations will continue and the findings, in 
terms of parametric optimization of the EPS instrument, will be implemented 
and tested during the long periods of cruise (using the interplanetary 
energetic particle environment), and during planetary encounters of Earth and 
Venus, between the time of launch and the first flyby encounter of Mercury.  



Table 1.1  Final EPS Engineering Parameters and Assessment against 
requirements
ParameterCSR/PLR Requirement / ExpectationMeasured or derived from 
MeasurementsComments 
(PLR takes precedent over CSR)Engineering ParametersMass1.3 kg 
(including FIPS)EPS only: 1.5 kgIncludes cable to FIPSPower Average2 W 
(including FIPS)EPS only: 4.8W primaryIncludes Converter losses for FIPS: 
FIPS uses and additional 2 WPower PeaksamesameAdditional power needed one 
time to deploy coverData Rates: Average80/2 bps91 bpsWith DPU 
CompressionData Rates: Peak~500 bpsWith DPU CompressionSize“Hockey 
Puck”15.55 x 12.98 x 11.25 cmIncludes mounting tabs / structuresPerformance
ParametersGeometric Factor0.1 cm^2.s0.0074 cm^2 srCalculated: PDR 
Expectation: 0.01. Decrease driven by electron scattering issuesIon Energy 
Range10 keV/n to 3 MeV10 keV/n to ~3 MeVLowest energies with TOF-only, as 
expectedElectron Energies25 to 500 keV25 to 500 keV> 170 keV based on 
electron SSD response to Radiation Source.  SpeciesH, He, CNO, Fe, 
electronsH, He, CNO, electronsFe measurement expected by extrapolation of 
measurementsField-of-View160? x 12?160? x 12?Angular 
Resolution160?/6~30 deg.No explicit requirementsTOF Range0-200 ns, ?200 
ps10-200 ns, ?1500 psSufficient for required mass discr.Integration 
periodFixed: 36 s0.5 – 300 sProgrammable
Assessment of the EPS Instrument Performance Against Requirements.

As described in the previous minor section, the calibration and 
characterization of the EPS instrument is ongoing and will continue long after
the launch of the MESSENGER Spacecraft.  However, we know enough about the EPS
performance to compare what we know about it against the MESSENGER Program 
Level Requirements and the MESSENGER Concept Study Report requirements for 
EPS.  The high level characteristics of EPS as compared to requirements are 
shown in Table 1.1.  The present understanding of the performance of the EPS 
instrument is provided in Table 6.1.  The major deviation from performance 
expectations (not requirements) is the uncertainty regarding the efficiency of
the triple coincidence TOF versus Energy ion measurements.  Should the poor 
efficiency performance on triple coincidence ion measurements persist, despite
indications with an engineering model that efficiencies will rise once the 
instrument is optimized after launch (with various discrimination and voltage 
settings), the EPS requirements are still met.  Fast ion measurements will be 
made with the double-coincident TOF-only ion measurement.  Composition 
measurement requirements will be achieved but over longer integration times 
than anticipated at the time of the MESSENGER Concept Study Report.  

2.0 EPS INSTRUMENT DESCRIPTION

EPS Sensor Description

The Energetic Particle Spectrometer (EPS) is a miniature Time-of-Flight (TOF) 
versus Energy (E) spectrometer that measures ions and electrons over a broad 
range of energies and pitch angles. Directional ion composition and energy 
spectra are measured for H to Fe from >5 keV/nucleon to ~3 MeV total energy.  
Directional electron energy spectra are measured from 25 keV to 0.6 MeV.  A 
picture of the EPS instrument on the bench is shown in Figure 2 without the 
thermal radiator that partially obscures the view of EPS in Figure 1.



Heritage. EPS was designed and manufactured at the Johns Hopkins University 
Applied Physics Laboratory (JHU/APL) based on technologies developed for many 
previous missions.  The TOF vs. E technique was first implemented for 
energetic ions with the MEPA instrument on the Ampte Mission.  It has since 
been flown on Galileo Energetic Particle Detector (EPD), and Geotail Energetic
Particle Ion Composition (EPIC) for measuring energetic ions, and on Cassini 
Magnetospheric Imaging Instrument (MIMI) and Imager for Magnetopause-to-Aurora
Global Exploration (IMAGE) High-Energy Neutral Atoms (HENA) instrument for 
measuring energetic neutral atoms. For previous energetic ion applications, 
the directionality of the ions was determined by means of mechanical and 
spacecraft rotations of well-collimated, single conical view sensors.  The new
element on MESSENGER EPS is that multiple simultaneous views (6) are obtained 
using a single time-of-flight chamber [McNutt et al., 1996; Andrews et al., 
1998].



Time-of-Flight Operation.  Figure 3 shows how EPS operates.  In Figure 3a, 
ions (blue) enter the collimator on the left, penetrate a “start” foil, pass 
thought the time-of-flight drift chamber, penetrate a “stop” foil, and then 
imbed themselves within the solid-state-detectors (SSDs) on the right.  As the
ions pass through the start and stop foils, secondary electrons are generated 
(red).  The secondary electrons are electrostatically steered onto different 
sections of a circular Microchannel Plate (MCP).  Segmented anodes collect the
MCP-amplified signals from the different sections of the MCP.  A JHU/APL 
developed time-of-flight (TOF) application-specific integrated circuit (ASIC) 
uses the start and stop signals to derive a TOF value in the ~1 to 200 nsec 
range.  A RTX-2010 microprocessor uses the TOF value and the total energy E 
value from the solid-state detector (SSD) to identify the elemental mass 
species M and the total ion energy E.  Such double-coincident (TOF plus E) 
events are called “High Energy Ions”.  That TOF versus E information can be 
used to so sort the incoming ions was proven on previous missions such as the 
Galileo EPD instrument, with data presented in Figure 4.  Ions that generate a
TOF signal but do not generate a SSD signal are also processed as “Low Energy 
Ions”.  With such low-energy ions, crude compositional information is 
retrieved using the TOF value and the MCP pulse height (PH). The pulse height 
is discriminated by EPS with 2 MCP discrimination levels over and above the 
noise-threshold discriminator.  That TOF versus PH can crudely discriminate 
between heavy and light species was demonstrated with the IMAGE HENA 
instrument (Figure 5). The TOF performance depends on the time dispersion 
associated with the secondary electron trajectories plus the dispersion 
introduced by the electronic signal processing. Simulations of that dispersion
are shown in a later section.





Foils and Materials.  The performance of EPS depends on the materials through 
which the particles must pass.  Such materials degrade particle energy and 
scatter particles in angle.  The start foil consists of 500 Å of polyamide 
covered on both surfaces with 50 Å of aluminum to ensure conductivity of the 
outer layer of the foil and enhance rejection of ultraviolet light, 
particularly Ly-?. The stop foil also consists of 500 Å of polyamide, but it 
is additionally covered by 100 Å of palladium to reduce further light (both 
visible and UV) transmission toward the SSDs.  The foils are mounted on ~90% 
transmittance stainless steel grids.  Dead-layers on the SSDs are of order 500
Å.  The cylindrical foil holder has an inner radius from the center of the 
sensor of ~2.88 cm.  The entrance and exit apertures within the foil holders 
have a height of 0.615 cm.  The entrance and exit foils each comprise two 
separate foils, and so the foil holder has a 0.127 cm blockage at just the 
very center of the sensor FOV (which translates into ~5?).  Some views of the 
mechanical configuration of EPS are provided in Appendix 12.

Directionality.  Figure 3b shows that from the top, the EPS sensor volume has 
cylindrical symmetry.  The multi-hole collimator (Figure 3c; the collimator in
Figure 3b is notional only) ideally maps particles arriving from one specific 
direction on the left uniquely past just one of the 6 start anodes on the left
and uniquely onto just one of the 6 SSDs on the right.  The direction, from 
which the ion arrives, in one angular dimension, is therefore determined by 
which SSD is activated (for “high-energy ions”) or which start anode is 
activated (for “low-energy ions”).  Because of a compromise to achieve the 
highest level of sensitivity (high geometric factor, see below) the collimator
was opened up somewhat so that the angular response has side lobes at the 
several-percent level.  Deconvolution of the collimator response will 
therefore be required to derive angular distributions with the highest 
fidelity.  

Solid-State-Detectors.  Each of the 6 SSDs has four pixels, two dedicated to 
the ion measurements described above, and two to electron measurements (Figure
6a), for a total of 24 SSD elements. For each species (ions or electrons) on 
each of the SSDs there is both a large area pixel (40 mm2) and a small area 
pixel (2 mm2). These 2 pixels allow for flexibility in the geometric factors 
(hence sensitivity and dynamic range) of the sensor.  At any one time, 12 of 
the 24 elements are used.  One may choose any combination of : i) large-ions /
large-electrons, ii) large-ions / small-electrons, iii) small-ions / 
large-electrons, or iv) small-ions / small-electrons (for all 6 SSD’s 
simultaneously). Two views of the SSD mounting structure are shown in Figure 
6b.  The surfaces of the SSD are positioned ~3 mm behind the stop-foil 
mounting structure.  



Electron Measurements.  A thin layer (flashing) of 1 ?m of Aluminum covers the
SSD elements dedicated to electron detection. This dead layer stops protons 
with energy less then 200 keV; while electrons loose less then 10 keV energy 
by the interaction with this additional layer. Electrons are identified in EPS
by the presence of an energy signal in one of the electron SSD elements and by
the absence of start and stop pulses. Energetic electrons have low efficiency 
for production of secondary electrons when passing through thin foils (< 1%). 
Under such conditions, a signal of 50keV in the electron detector can be 
generated either by a ~250 keV proton that did not generate either a start or 
a stop pulse (with a probability of perhaps < 25%), or by a 60 keV electron. 
The TOF vs E and TOF vs. PH spectra collected in the adjacent ion SSD element 
(without flashing) are used during ground data analysis for checking and 
correcting for the proton contamination within the electron SSDs


Collimator.  The EPS collimator consists of 4 concentric, thin (20 mils) 
titanium sheets perforated with a regular pattern of holes (Figure 7).  The 
holes are drilled so that, if one imagines straws passing through each of the 
set of 4 holes that pass through the 4 concentric sheets, all of the straws 
would focus onto a single point right at the center of the cylindrical 
structure of the sensor half way between the start and stop foils.  That is, 
even while the walls of the collimator have cylindrical symmetry, the 
hole-patterns themselves have spherical symmetry.  Even the sizes of the holes
have spherical symmetry; that is, the sizes of the holes in the 4 concentric 
titanium sheets are proportional to the distance from the focal point in the 
center of the sensor.  This design has two advantages over commonly used 
collimators: i) the geometric factor for particles striking the front foil of 
the sensor is essentially the same as the geometric factor of the entire 
sensor (thus minimizing the accidental rates that arise from high counting 
from the start foil), and ii) the primary particles are kept well away from 
surfaces within the sensor that can scatter particles.  Figure 7c shows a 
front-side view of the collimator, as it would appear to a detectable particle
approaching the entrance aperture.  The inner radii of the 4 cylindrical 
sheets that constitute the collimator are:  4.50, 4.12, 3.58, and 3.20 cm.  
The holes in the sheets have diameters 2.34, 2.24, 1.66, and 1.68 mm and are 
arrayed in a hexagonal pattern of 4 rows (Figure 7c).  The centers of the 
holes in the outermost and innermost rows view 3.83? and 1.92? away from 
central plane.  There are 33 holes per row spanning the ~160? FOV. Each set of
4 aligned holes has an absolute maximum geometric full-width acceptance cone 
of 8.8?, and so that 2 x 2.38? + 8.8? = 13.56? is the absolute full-width of 
the viewing fan.  The full width at half maximum is roughly 12.2?.  

If the holes within the collimator walls are relatively small, then the 
mapping of particles that come from a singe direction (and illuminate the 
entire instrument) to a single position at the plane of the solid state 
detectors is well behaved.  As mentioned in a previous section, the holes in 
the EPS collimator are larger than those that would be optimum for angular 
response in order to achieve greater sensitivity.  Thus, the angular response 
has side lobes, as described in a later section.

Electron / ion fields of view and detector numberings.  The nominal total 
field-of-view (FOV) of EPS is 160? x 12?.  Because the electron and ion SSD’s 
are side-by-side, the total electron or ion FOV in the long dimension is about
1/12 smaller (~13? smaller) or about 147?.  And, the centers of the ion and 
electron FOV’s are shifted with respect to each other by ~13?.  In Figure 1, 
let us define two angles:  “alpha” is the angle from the +Y(s/c) axis and 
within the Y(s/c)-Z(s/c) plane (with “plus” angles viewing towards the +Z(s/c)
axis); “beta” is the angle for rotations away from the Y(s/c)-Z(s/c) plane.  
With these definitions, the total FOV of EPS is roughly:  (-80? < alpha < 
+80?) and (-6? < beta < +6?).  The ion FOV is (-67? < alpha < +80?) and (-6? <
beta < +6?).    The electron FOV is (-80? < alpha < +67?) and (-6? < beta < 
+6?).     

With the ion and electron segments, there are a total of 12 SSD elements 
active at any one time.  The numbering scheme for these detector elements 
ranges between 0 and 11, with the even SSD elements corresponding to electrons
and the odd SSD elements corresponding to ions.  The numbering does not change
with the selection of large or small detectors for either ions or electrons.  
The complete specification of an EPS detector element would be, for example:  
“SSD element 3 small”.  In Figure 1, the “0” detector (an electron detector) 
is the one that looks most closely aligned with the –Z(s/c) axis.

The numbering of the TOF Start-Anodes ranges between 0 and 5.  An ion or 
electron that passes right over Start-Anode “0” (only the ion “stimulates” 
this start anode) strikes either SSD 10 or SSD 11.  Thus, the Start-Anodes 5, 
4, 3, 2, 1, and 0 map to SSD’s (0, 1), (2, 3), (4, 5), (6, 7), (8, 9), and 
(10, 11), respectively.  



EPS Electronics Description

The layout of the EPS electronic boards is shown in Figure 8 and an EPS Block 
diagram is shown in Figure 9.

SSD Detector Pulse Conditioning and Measurement. The SSD charge amplifier used
(the “front end amp in Figure 10) is an Amptek A250F followed by a 2.4 
microsec shaping stage utilizing an AD8032. An A/D converter accepts the 
shaped pulses with a maximum input of 2.5V. The shaping circuit also outputs a
fast timing pulse that is used for “time-to-peak” digitization purposes. A PHA
SSD pulse amplitude is obtained both at the time that the fast pulse is first 
detected (the so-called baseline measurement), and at the time of the peak in 
the shaped pulse.  Both the baseline and the peak amplitude are captured 
within the PHA event packets.   The energy measurement is based on the 
difference between the peak and the baseline measurement (unless a commanded 
choice is made to use only the peak value).  Only twelve of the 24 A250F’s are
powered at any given time based on the setting of the “analog switch bank” 
(either the large pixel or the small pixel preamps).  The conversion between 
SSD electronic energy and PHA bin number is:  Energy (keV) = 0.57387 * bin#.



SSD Detector Thresholds. Eight D/A converters are used to set the threshold of
groups of three discriminators. The DAC grouping is shown below.  An event 
that is registered within the SSD is used to generate two separate pulses, a 
fast pulse used for timing (ergo the “timing” discriminator) and a slow pulse 
used for pulse-height analysis (the “shaped” discriminator).  The “timing” 
referred to here should not be confused with time-of-flight measurement.


1. timing discriminator (Electron channels 0,2,4)
2. shaped discriminators (Electron channels 0,2,4)
3. timing discriminator (Ion channels 1,3,5)
4. shaped discriminators (Ion channels 1,3,5)
5. timing discriminator (Electron channels 6,8,10)
6. shaped discriminators (Electron channels 6,8,10)
7. timing discriminator (Ion channels7,9,11)
8. shaped discriminators (Ion channels 7,9,11)

These settings are communicated to the ground as part of the housekeeping data
carried with the EPS high priority, as shown in Table 2 of Appendix 4.  

Detector Channel Control. All 12 detector channels have fully independent 
circuitry that feeds a common FPGA-based control circuit. By masking its 
discriminator firings via a software-programmable register, a noisy channel 
may be ignored.

Time-of-Flight System. Micro-channel plates (MCPs) are used as time-of-flight 
start and stop pulse detectors. Seven anodes (six start and one stop) go 
directly to the Time-of-Flight board via coax cables. Each of the six start 
anodes is split into two isolated signal paths:

1. Input to a discriminator (with adjustable threshold) and used to identify 
which anode fired
2. Recombined with other start anode signals, amplified and used as start 
pulse for the TOF chip via a Constant Fraction Discriminator (CFD)

The stop anode signal is also split into two isolated signals:

1. One signal forms the stop pulse for the TOF chip via a CFD
2. The other signal is fed to two level discriminators which are used to 
indicated whether the particle was “light” (H or He), “medium”, or “heavy” (C,
N, O, Fe)

The threshold settings are communicated to the ground as part of the 
housekeeping data carried with the EPS high priority, as shown in Table 2 of 
Appendix 4.  The time-of-flight (TOF) value is digitized into discrete bins 
with the conversion: TOF (ns) = 0.16276 * bin#.

Rate Counters. A total of 35 rate counters are implemented in a 
field-programmable gate array (FPGA). All counters are 16-bit counters, except
START, STOP and VALID TOF, which are 24 bits deep. These counters are read out
directly from the FPGA in response to 38 unique address-decoded read cycles 
from the processor.  In addition, the RTX2010 Event Processor generates 7 
additional counters.  The hardware and software counters are documented in 
Appendices 1 and 2.

Event Logic.  During normal operation, EPS generates 3 different data 
products, “Electron Data”, “Low Energy Ion Data (LEID)” and “High Energy Ion 
Data (HEID).”  A schematic diagram of the logic producing valid events is 
provided in Figure 10.  EPS can be put into a mode (the diagnostic mode) 
where, instead of generating LEID and HEID it generates ion spectra based 
solely on the ion SSD energy events (TOF and PH are ignored).  Here the “SSD 
Ion Data” is generated in a fashion exactly parallel to the process that 
generates the Electron Data.

It is possible to command a decimation of the electron throughput in order to 
prevent the electron rates from overwhelming the ion rates within the Event 
Processing Unit.

Event Packets. The physical information in a valid event is stored in a block 
of words (16 bit) called event packets. These events are stored by the FPGA 
controller into FIFO and are read out by the RTX2010 Event Processor Unit.  
Four different event packets have been defined: 1) Electron Packets, 2) High 
Energy Ion Packets, 3) Low Energy Ion Packets, and 4) Diagnostic Packets.  The
information contained in these event packets, as package for telemetry to the 
ground for selected events, is shown in Appendix 7.  

Software Event Processing.  The EPPS software obtains event data through two 
separate FIFOs, one for EPS and one for FIPS. Event packets consist of a 
number of event words. The total number of words in a packet depends on the 
type of event, as described in the last subsection. The first word within each
event packet contains a synchronization pattern and an ID that specifies the 
event type. The synchronization pattern is used to detect the beginning of a 
new packet.

The acceptable throughput for EPS event processing is of order 5 kHz.  During 
high counting situations the processed events are designed to “poll” the 
incoming events and thus maintain the correct distribution of ion energies, 
species, and directionality; and separately, the correct distribution of 
electron energies and directionality.  The correct particle intensities are 
derived by ratioing the appropriate hardware counters with the appropriate 
software counters.  Note that when the rate of incoming particles exceeds the 
capabilities of even the hardware, such simple ratioing does not suffice.  
Part of the characterization and calibration activities is to determine the 
so-called “Rate-versus-Rate” corrections (so called R-R corrections) 
associated with extremely high counting situations.  

One way that the EPU processes event packets is to generate spectra or 
histograms for the various event types by using lookup tables to determine the
corresponding bin within each spectrum to increment. These lookup tables are 
accessed based on some combination of the energy, time-of-flight and/or 
mass-to-charge ratio depending on whether the particle is from the EPS or FIPS
system. For example, with EPS the TOF versus E space shown in Figure 4 is, for
each of the 6 view directions, divided into 22 different histogram bins to 
generate the EPS High Energy Ion data.  Also with EPS, the TOF versus Pulse 
Height space shown in Figure 5 is divided into 16 different bins to generate 
the Low Energy Ion data.   The bins to be incremented within the electron 
spectra are determined based on the electron event energies.  Eight different 
energy ranges are used.  The channel definitions for Electrons, Low Energy 
Ions, High Energy Ions, and Diagnostic Mode Ions are provided in Appendix 8.

The EPPS instrument software system can also be commanded into a "test mode" 
in which it accepts commands from and outputs selected science and status 
information through an RS-422 test port.  A test-port may also be employed to 
supply a source of simulated event data for use in instrument checkout.

The accumulation, formatting and reporting of EPS event and housekeeping data 
are synchronized to two different time intervals. These are referred to as the
N1 and N2 time intervals. The N1 time interval is nominally 300 seconds and is
an integral multiple of the N2 time interval. The N2 time interval has a 
nominal value of 30 seconds. Either or both of the intervals can be changed 
via command. However, the value for N1 should always be a multiple of the N2 
value. (If N1 is not a multiple of N2, the medium priority data becomes 
garbled and uninterruptible).  Among the data collected at these intervals are
the 35 hardware rate counter registers. To prevent these registers from 
overflowing, each of the intervals N1 and N2 is subdivided into 10 
equal-length subintervals. Thus the basic integration period corresponds to 
N2/10 seconds.  The EPS-event first-in-first-out (FIFO) is flushed at the 
conclusion of each N2/10 interval to keep the events that occur during 
different time intervals separate.

Hence all data products for the EPS subsystem are collected at the N2/10 
second rate. Values that are reported every N2 seconds are produced by summing
the 10 consecutive N2/10 samples. Values that are reported every N1 seconds 
are produced by summing over as many N2/10 subintervals as there are in the 
N1-second time period.

EPS Telemetry.  The High-Priority, Medium-Priority and Low-Priority data 
formats for EPS are documented in Appendices 4, 5, and 6.  High-Priority and 
Low-Priority data is collected at a cadence of N1 (last subsection; often 
300s).  Medium-Priority data is collected at a cadence of N2 (30s or less).  
The Pulse-Height-Analysis-event formats imbedded in these data products are 
documented in Appendix 7.   Appendices 4 and 5 also document in red the 
changes that result when EPS is put into the Diagnostic Mode for measuring Ion
Spectra (and Electron Spectra) while ignoring the TOF system.  Finally, the 
status of the EPPS instrument, including EPS is communicated through a 
separate EPPS Status Packet as documented in Appendix 3.  

Pulse-Height Analysis Priority Scheme. 

In order to prevent the most common particle species from overwhelming the 
Pulse-Height-Analysis (PHA) events that are sent to the ground, a rotating 
priority scheme is used for selecting species groups.  The priority scheme is 
documented in the EPPS Flight Software Specification Document (MESSENGER 
Document JHU/APL 7389-9041).

3.0 SCOPE OF CHARACTERIZATION AND CALIBRATION

EPS measurements are intended to generate the information needed to derive the
charged ion and electron differential intensities (I [cm-2 sr-1 s-1 keV-1]) 
and phase space densities (PSD[s3 cm-6]).  The significance of I and PSD are 
that, for many particle transport processes within planetary magnetospheres 
and interplanetary space, they satisfy the collision-less Boltzmann equation, 
derived from Liouville’s Theorem for collision-less transport, for energy 
preserving, and for non-energy preserving processes, respectively.  Our 
purpose here is to develop the quantitative procedures for converting the 
count rates (R[counts s-1]) reported by EPS in the many predefined channels 
described in preceding sections, or in channels defined on the ground using 
the pulse-height-analysis (PHA) data, into estimates of I and PSD for the 
various defined ranges of energies, particle species, and arrival angles.  We 
begin by defining terms.

Flux, differential intensity and phase space density

The rate R[particles/s] that traverse an area A can be given by the Flux F 
[cm-2 s-1] with:

R = A * F                                        (1)

or by the intensity i [cm2 s-1 sr-1]

R = A * ? i cos(û) d?                                    (2)

where ? is solid angle and û is the angle to the area normal.

Often used is the quantity differential intensity I [cm2 s-1 sr-1 keV-1], 
defined as the number of particles of a given species S with energy between E 
and E+?E that traverse the area A during the time t, where

R(S, E) = I(S, E) * A * ?? * ?E                            (3)
    
In three dimensions, with ? being the polar angle and ? the azimuthal angle of
a polar reference system:

d3R(S, E,???) = I(S, E, ????) * A(????) cos(û) * t * dE cos ? d? d?           
(4)

where we note that the area A and the unit vector û pointing normal to A are 
both functions of ? and ?.

Note that for non-relativistic energies, I(S, E, ????) is related to the phase
space density PSD (number of particles in the configuration space element d3R 
and with velocity between v and v+ d3v) by the simple relationship

PSD[s3 cm-6] = I[cm2 s-1 sr-1 keV-1] * m/v2                        (5)

For relativistic energies, the phase space density (identified with lower case
letters here: psd) is defined in terms of particle momentum p (psd[s3 cm-6 
kgm-3]), and the relationship to differential intensity I is: 

psd[s3 cm-6 kgm-3] = I[cm2 s-1 sr-1 keV-1] / p2                    (6)

Definition of sensor transfer function and geometric factor

Using (4), the rate R[particles/s] of particles of species S, in the energy 
band ?E around mean energy E, angular band ?? around mean polar direction ?, 
and angular band ?? around the mean azimuthal direction ?, measured by the 
instrument can be expressed as:

R(S, E)  = ??? ?d? ?d?   I(S, E, ????) * A(????) cos(û) * dE cos (?? d? d?    
   (7)

One may take the kernel approach to relating R to I by allowing I(S, E, ????) 
to be a Dirac ?-function (mono-energetic, infinitely narrow beam).  Then:

dR(S, E, ????)  = dI(S, E??????) * T(S, E, ????)                    (8)

where G(S, E, ????) [cm2 sr keV] is the kernel, also known as the “transfer 
function” of the instrument.  Then:

R(S,E) = ??? ? d? ?d?   I(S, E, ????) * T(S, E, ????) cos(??d? d?            
(9)

It is typical for the first estimate to assume that the particle intensities 
are constant over the energy-species-angle bandwidths of any one channel.  The
estimated intensity Ikj associated with a specific detector k, which at a 
specific time views a range of angles centered in a specific direction (?k, 
?k), and which is associated with a specific energy channel j, which measures 
particle species Sj over an energy range that stretches from Ej to Ej + ?Ej, 
is given by equation (9) as:

Rkj = ???j  Rk(Sj, Ej) dE = ??Ej  Ik(Sj, E) * T(Sj, E)) dE  ~  <Tkj> Ik(Sj, 
Ej) ?Ekj    (10)

where Rkj is the channel count rate and where <Tjk> is called here the 
Transfer Factor Hkj.  It is convenient to separate Hkj into a counting 
efficiency “?kj” and a factor that relates strictly to geometry, the so-called
Geometric Factor Gkj.   That is:

Ik(Sj, Ej) ~ Rkj / (?kj Gkj ?Ekj)                                (11)

Therefore, the first goal of calibration is to obtain 
detector-channel-averaged transfer factors Hkj (or the equivalent efficiency 
factor ?kj since Gkj is easy to calculate), the detector-channel energies Ekj 
and the detector-channel energy band pass ?Ekj.  

More quantitative work may require a full inversion of the kernel equation 
given in Equation (9).  Here we are usually aided by the fact that to some 
level of approximation the Transfer Function T is separable in the form:  

T(S, E, ????) =  K(S, E) * P(????)                              (12a)

and hopefully even            

T(S, E, ????) =  K(S,E) * C(?) * D(?)                        (12B)

Note that even with these separations, an exact inversion of the integral is 
rarely possible, and we can compute only the coefficients of some tailored 
expansion of K, C, and D, as for example in spherical harmonics (using 
Legendre polynomials). The accuracy of these coefficients depends on both the 
raster coverage of the measurements and on the calibration.



Goals of the EPS characterization and calibration efforts

The goal of EPS characterization and calibration efforts, then, is to relate 
the rates, Rkj, of the on-board or ground-defined channels to the in situ 
particle intensities I(S, E, ????)  by developing a hierarchy of information 
about the EPS Transfer function T(S, E, ????).  Specifically, we wish to 
derive for each channel:

Ekj,  ?Ekj,  ?kj Gkj j,  Ckj(?),  Dkj(?),  Kkj(S, E), Pkj(????),  and finally,
Tkj(S, E, ????)

Clearly, the further to the right we move into this list, the higher the 
fidelity the characterization and calibration efforts must be. 

Typical calibration procedure consists of stimulating the instrument with 
energetic particles, first from radiation sources, and then from accelerator 
beams (fair approximation of a delta function) and recording the response of 
the different channels. For EPS we must scan following variables:

* Species and Mass: e, H, He, O, Ar (proxy for heavier species, e.g. Fe)
* Energy: 1keV – 5MeV (~40 energies for 10 points/decade)
* Polar angle: +/- 10 deg (12 degrees nominal FOV)
* Azimuthal angle: +/- 90 deg (160 degrees nominal FOV

Note that even given this level of discreetness, complete characterization to 
the level of establishing Tkj(S, E, ????) for all channels requires: 5 masses 
* 40 energies * 21 polar angles * 180 azimuthal angles = 756,000 calibration 
points!  Clearly it is not possible to run this many beam values without 
significant infrastructure and automation.  Therefore, we are dependent to a 
substantial degree on the separability of the Transfer Function as described 
in Equation (12).  An important goal of the characterization and calibration 
efforts is, in fact, to establish that degree of separability. Simulations 
play an important role in establishing expectations for the instrument, as 
discussed in the next section.

4.0 EXPECTED EPS PERFORMANCE AND SIMULATIONS

Described here are the calculated expectations regarding the performance of 
the EPS instrument.  We describe various calculations and simulation runs that
provide a higher level of fidelity in predicting the performance of the EPS 
instrument than those based on experience with previous instruments.  These 
simulations constitute the benchmarks against which subsequent ground and 
flight calibrations are to be compared.  A table of expected EPS performance 
characteristics is provided at the end of this section (Table 4.1)

Numerical simulation provides an efficient and effective mean of verifying and
quantifying the flight instrument response.  Hence, a range of simulation 
tools are developed in order to characterize EPS.  The results from these 
simulation tools allow us to calculate the end-to-end expected performance of 
the instrument which one can use to compare with in-flight and calibration 
data.  Different sections of EPS are simulated using different tools to gain 
better understanding on each part of the instrument.  Three major sections are
simulated separately: 1) Collimator; 2) TOF section, and 3) Solid State 
Detector (SSD).

Collimator Performance and Geometric Factor

Collimator Geometric Factor.  Figure 7 shows the collimator of EPS, as 
described in section 2.  It consists of four concentric half circular plates 
that have holes aligned with a common point of origin at the center of the EPS
TOF telescope.  The size and number of collimator holes define the geometric 
factor G of the instrument.  The many-holes collimator design minimizes the 
scattering of ions and electrons at the collimator while restricting the 
field-of-view (FOV) of the instrument.  The geometric factor of a single set 
of aligned holes, with outer-hole and inner-hole diameters of 2.34 and 1.68 
mm, with holes separated by 1.3 cm, is 0.00056 cm2 sr. With 33 x 4 holes, the 
geometric factor G for the entire collimator is 0.074 cm2 sr.  Assuming that 
each species on each SSD is allotted 1/12th of that amount, the Gkj value is 
roughly 0.0062 cm2 sr.  Gkj for the Low Energy Ions (TOF-only) is roughly 
twice that value, or ~0.012 cm2 sr.  Note that the needed Transfer Factor Hkj 
depends also on the counting efficiency ?kj which depends, in turn, on species
and instrument mode. 

Collimator Side Lobes.  This geometric factor calculation given above does not
take into account the situation of particles coming through different set of 
holes (i.e. side lobes) that are not incorporated a detector with flight-like 
dimensions. An example of such side lobes is shown in Figure 11. Here, a Monte
Carlo simulation and an independent laboratory measurement of a collimator 
were performed. The laboratory measurement (blue line; performed with laser 
light and a 1 mm wide photo diode) is compared with the simulation (shown in 
pink), which incorporated a detector with flight-like dimensions. The real and
the simulated detections were performed at the position where the SSD 
detectors are mounted.  Both curves have been normalized to transmission of 
100% at 0? incidence.  The simulation results are in excellent agreement with 
the actual measurement with the collimator.  The collimator used for this 
comparison is a prototype with slightly different characteristics from that 
used for flight.  The prototype validated the simulation model, which in turn 
was used to refine the flight model characteristics.

A two-dimensional view of a simulation of the true flight-collimator 
transmittance characteristics is shown in Figure 12.  Here the view is as it 
would be to a tiny bug sitting on the surface of one of the SSDs.  The main 
side lobe resides about ?28.3 degrees on either side of the principal peak 
along the “alpha” direction along the main axis of the sensor (The principal 
peak is shown with saturated color values and positioned at [0, 0]).  Each 
side lobes represent ~4 % of the transmitted flux, and the full width at half 
maximum (FWHM) of the principal peak is ~7.5?.  Because Figure 12 is the 
response for a single position on a SSD, the instrument-level Ckj(?),  Dkj(?),
and Pkj(????) functions are calculated by convolving the square SSD sensor 
area (as transformed to a an angle) with the function displayed in Figure 12. 




Electron Scattering. Simulations using GEANT4 software (Agostinelli et al., 
2003) were also performed to assess the collimator performance for suppressing
electron scattering effects.  Figure 13 shows the configuration used in the 
simulation along with some sample electron trajectories and secondary products
(scattered electrons, x-rays).  Simulations like this demonstrate that the 
electron counts detected in detectors adjacent to detector directly targeted 
by the collimator are just several percent of the counts observed in the 
targeted detector.  The engineering prototype of EPS will be tested with 
electrons after the planned upgrade of the Johns Hopkins APL particle 
accelerator facility before about 2007.



Ion Measurements 

Electron Optics.  As discussed in section 2, secondary electrons generated by 
primary ions in a start and a stop foil are used to generate timing signals to
measure the ion’s time-of-flight (TOF) within EPS.  They are also used to 
determine the directionality of the low energy ions that do not stimulate the 
SSDs.   Figure 14 shows sample simulated secondary trajectories within EPS 
from the foil to MCP, derived using the commercial COULOMB software (see web 
site: “www.integratedsoft.com/products/coulomb/”).  Here, a pie-shaped cutout 
of the cylindrical EPS sensor is shown (cylindrical symmetry is assumed for 
the electrostatic fields).  The electron optics for both the start foil/MCP 
and stop foil/MCP are identical, hence Figure 14 shows a good representation 
for both start and stop electron trajectories.  Electrons are generated at the
foil (right of the figure) when ions (not shown) transverse through the foil. 
These low energy electrons (typically <10 eV) are immediately affected by the 
accelerating voltages within EPS (the high energy ions are not significantly 
affected by the accelerating voltages) and are steered towards the 
Microchannel Plate (MCP; the flat surface on the top end in the figure) as 
evident in the figure.  An important issue is the dispersion in the arrival 
times of the secondary electrons as a function of foil position and the angle 
of emittance of the secondary electrons from the foil.  Such dispersion adds 
to the error in the measurement of time-of-flight.  Figure 15a shows the time 
dispersion associated with varying angles for electrons emitted from the 
center of the foil.  In Figure 15b the dispersion shown for any one column are
associated with varying the emission position over the ~6 mm vertical extent 
of the foil.  The combination of position and angle yields a time dispersion 
of roughly 1.5 – 2 ns.  Combining the dispersions associated with the start 
and stop detections in a root-mean-square sense yields a total time dispersion
error of roughly 2.1 – 2.8 ns.  Combining that value in a root-mean-square 
sense with the ~1.5 ns electronic dispersion yields a total time dispersion 
that resides between 2.5 and 3.2 ns.






Ion energy losses.  As ions transverse the TOF section of EPS through the 
front foil (~10 ?g/cm2), back foil (~19 ?g/cm2), dead layer of the solid state
detector (~550 Å), and eventually stop and deposit their total energy in the 
SSD, they will lose energy and scatter depending upon the ion’s initial 
energy, mass and the medium through which they pass.  We have incorporated 
data from the SRIM (Stopping and Range of Ions in Matter; Biersack and 
Haggmark, 1980) particle-interaction-with-matter code in our simulation to 
simulate realistically the energy loss and scattering as ions transverse 
through different materials in EPS.  

TOF measurements.  When ions penetrate through the front foil, a distribution 
of ion velocities is created, as calculated using SRIM.  This distribution of 
ions is then used to calculate the distributions of ion TOFs.  The uncertainty
due to secondary electron dispersion and electron noise together with SRIM 
data are all incorporated in our simulation.  Figure 16 shows the simulated 
response of EPS TOF spectra as a function of particle initial energy for four 
species (H, 3He, 16O, and 56Fe).  At low energy (~10s of keV), ions lose 
significant amounts of energy and scatter significantly when going through the
front foil.  These effects explain the spread in TOF measurement at low 
energy.  However, at higher energy (~100s of keV), the TOF spreads are mostly 
consequences of the uncertainties in the TOF measurement from both the 
electronics jitteri (~1.5ns), and secondary-electron dispersion in the TOF 
optics as discussed in previous section (except for very heavy ions, e.g., 
Fe).



Total energy measurement.  If an ion has sufficient energy left once it 
transits the front and back foil, it ends up in the SSD.  Depending on the 
ion’s final energy and mass when it reaches the SSD, it can penetrate through 
the dead layer of the SSD and produce an electronic signal to be measured.  
Figure 17 shows the simulated measured energy by EPS as a function of the 
ion’s initial energy before entry into EPS.  The measured energy shows the 
cumulative effect of the energy spread due to energy lost throughout the EPS 
(start foil, stop foil, dead layer, and electron hole pair production).  The 
energy loss is greatest for ions with the greatest nuclear charge, here Fe.  



TOF versus Energy.  Finally, Figure 18 shows the result of combining the 
energy measurement with the TOF measurement on a particle-by-particle basis.  
Notice that even with a ~3 ns time dispersion, the major elemental species of 
interest (H, He, CNO, Fe) are discriminated except at the lowest energies for 
CNO and Fe, where the error in the energy measurements dominates.  

Efficiencies.  The efficiency for detection of an ion within the SSD’s is 
roughly ~100% (except at the very lowest energies where energy straggling can 
position the energy below the low energy threshold).  The efficiency for 
obtaining a TOF measurement is estimated using the efficiency of generating 
secondary electrons in both the front and the rear foil. To emit a secondary 
electron, such electrons must be generated close enough (distance “?”) to the 
surface of the foil so that the electron can escape before it is 
re-assimilated. Thus, very roughly, it is expected that the efficiency for the
generation of a secondary electron is proportional to the amount of energy per
unit distance (dE/dX [keV/micron]) that an ion deposits as it goes through 
this outer thin layer of the foil.  The canonical number of secondary 
electrons generated out of each surface as a proton with 10’s of keV energy 
encounters the foil is between 0.5 and 1 (Frischkorn et al., 1983; D. W. 
McEntire, Private communication, 2004; estimated here as 0.75).  Since we 
require a simultaneous start and stop electron, the efficiency of proton 
detection at 10’s of keV is (1-Exp[-0.75])2, where 1-Exp[-0.75] is the Poisson
probability of having at least 1 or more electrons emitted when the mean 
emission number is 0.75.  Thus, the efficiency is roughly 22%.  Other energies
and species may be roughly scaled with this number using tabulated dE/dX 
values.  For example, at 50 keV total energy the dE/dX values for protons and 
oxygen ions are 120 and 250 keV/micron, respectively (we ignore for now the 
energy losses suffered by the ion in getting to the position in either foil 
where the secondary electron is generated).  Thus, since the average number of
secondary electrons for oxygen will by ~0.75 x (250/120), or 1.56, Poisson 
statistics tells us that the probability of detecting an oxygen ion with 50 
keV energy is (1-Exp[-1.56])2 or 62%.  Tables in Appendix 8 will provide 
calculations of expected efficiencies for the end energies (each end of the 
energy bandpass) calculated on the basis of a dE/dX scaling.  In those tables,
energy losses at intermediate positions within the instrument (i. e. at each 
of the two foils) are folded into the calculations.  Not folded into those 
calculations are the differences between forward and backward secondary 
electron generation nor is the probability of the primary ion being scattered 
within the start foil to the extent that it misses the stop foil and the SSD 
array.  The efficiency calculations in the Tables in Appendix 8 will be 
complete by 3/1/05, several months prior to the EPPS in-flight checkout 
period.  






Electron Measurements

The EPS electron measurement strategy depends on the use of the aluminum 
flashing on the electron SSD.  We therefore need to understand the effect of 
that flashing on both the ion and electron measurements within the electron 
SSDs.  Based on simulations with GEANT-4, Figure 19 shows the effect of Al 
flashing on proton measurements, and Figure 20 shows the effect on electron 
measurements, for a varying thickness of Al flashing.  The baseline spectrum 
assumed for both protons and electrons is a power-law spectrum with a spectral
index of 3; that is, Intensity = A E-3, where “A” is a constant.  With the 
flashing thickness utilized with EPS, 1 micron Al, the proton intensities are 
severely depleted for energies < 250 keV.  With that flashing thickness, the 
electron intensities are maintained to energies down to about 20 keV.  The 
dropoff of electron intensities at the highest energies occurs because of the 
electron penetration of the 500 micron detectors.  That drop-off is rounded 
because of electron scattering within the SSD, as demonstrated with an 
examination of the detailed trajectories of individual electrons with the 
GEANT-4 simulations.  There may be periods of time when the ion counts 
overwhelm the electron measurements.  The most important issue is not that the
electrons may be occasionally contaminated, but whether or not the electrons 
are contaminated in a way that we are fooled into thinking the ion counts are 
really electron counts. This problem is intrinsic to the technique used here 
to measure the electrons.  Extensive experience with similar measurement made 
by the EPIC instrument on Geotail demonstrates that, because of the clean 
proton measurements (below and above 250 keV) made in roughly the same 
direction, it is unlikely that we will be fooled into thinking that ions 
counts are actually electron counts.  



In Figure 20 the “1-micron” flashing spectrum lies below the nominal input 
spectrum primarily because of scattering of electrons as they traverse the 
flashing and the SSD itself.  The table of electron channels in Appendix 8 
will show the nominal consequences of such scatterings on the mean energies 
and efficiencies associated with each channel, as will be calculated with 
GEANT-4.  



Electron scattering off collimators and other sensor surfaces is always an 
issue with the measurement of electrons.  The GEANT-4 simulations of the EPS 
instrument discussed in the section on the EPS collimator assure us that such 
scattering generates signals in non-illuminated SSD that are no more than 
several percent of the signal within illuminated SSDs.




Table 4.1  EPS Expected Performance
Geometric Factor (Instrument Total)0.074 cm2 srGeometric Factor (Electron 
Pixel)0.0062 cm2 srGeometric Factor (Low E Ion Pixel)0.012 cm2 srGeometric
Factor (High Energy Ion Pixel)0.0062 cm2 srGeometric Factor (Diagnostic 
Ions)0.0062Electron Efficiency vs. EnergyRoughly 100%.  Details in Appendix
8 (TBS 3/1/05)Low Energy Ion Efficiency vs. En x M> 20% anticipated.  
Details in Appendix 8 (TBS 3/1/05)High Energy Ion Efficiency vs  En x M>20% 
anticipated Details in Appendix 8 (TBS 3/1/05)Diagnostic Ion 
Efficiencies~100% Energy CoverageSee Table Below?E / E capability at 50 
keV16%?E / E capability at > 100 keV< 8 %Energy / TOF Channel 
resolutionElectronic channels in Appendix 8.  Primary particle 
characteristics TBS 3/1/05TOF dispersive spread2.5 to 3.2 nsMass Species 
SeparationSee Table BelowAngular Coverage (total)160?  x 12? Angular 
Coverage (Electrons)147?  x 12?Angular Coverage (Low Energy Ions)160?  x 
12?Angular Coverage (High Energy Ions)147?  x 12?Angular Coverage 
(Diagnostic Ions)147?  x 12?Angular Pixels (Electrons)12? x 27? (13? with 
13.6? gap)Angular Pixels (Low Energy Ions)12? x 27? (full ~27? 
coverage)Angular Pixels (High Energy Ions)12? x 27?  (13? with 13.6? 
gap)Angular Pixels (Diagnostic Ions)12? x 27?  (13? with 13.6? gap)Electron
Scattering Angle ContrastSeveral percentCollimator Side-Lobes: Position(??,
??) = (?28.3?, ?0?)Collimator Side-Lobe Fraction/Lobe~4% per lobeEnergy 
Coverage:TechniqueElectrons15 keV – 0.6 MeVSingles / Foil TechniqueLE 
Protons5 keV – 1 MeVTOF + PH DiscriminationLE CNO15 keV – 3 MeVTOF + PH 
DiscriminationHE Protons15 keV – 2.8 MeVTOF vs. EnergyHE He25 keV – 2.8 
MeVTOF vs. EnergyHE CNO60  keV – 3 MeVTOF vs. EnergyDiagnostic Protons15
keV – 2.8 MeVSinglesEnergy / TOF ChannelsElectronic channel definitions.
In Appendix 8.   Primary particle characteristics TBS 3/1/05.
Summary of Anticipated EPS Performance. 

Table 4.1 summarizes the expected performances of the EPS instrument based on 
the calculations and simulations described in this chapter.  

5.0 FLIGHT EPS CALIBRATION

Here we document the calibration of the flight instrument that is now (29 May 
2004) mounted onto the spacecraft.  These calibrations primarily established 
the basic functionality of the instrument, but left open a number of questions
that are being addressed with a flight-like engineering model (Sections 6).  
Final calibration will be established in flight with a plan that takes 
advantage of much that we have learned, and will learn, with the engineering 
model (Section 7).  The calibrations were performed using a degraded (energy 
spread) Am-241 alpha-radiation source and an ion beam accelerator facility 
recently established at JHU/APL.  The JHU/APL calibration facilities are 
described in Appendix 9.  At present, the calibration sources and the test 
facilities generate only ions.  The functionality of the electron sensors in 
EPS have, to date been established with protons with sufficient energy to 
penetrate the Al flashing on the electron sensors.  From the detector 
functionality perspective, protons serve just as well as electrons to 
stimulate the electron SSDs and demonstrate that the electron signal 
processing chains operate well for the range of SSD energy depositions needed 
to characterize the ~25 to 500 keV electron energies of interest.  The 
question that remains following such proton calibrations of the electron 
channels is whether the scattering property of the electrons gets in the way 
of establishing the directionality of the electrons.  Our full-up, detailed 
geometric simulations of the electron performance of EPS using GEANT4 show 
that scattering randomizes the directionality of less than several percent of 
the electrons.  

Tests Set Up

For the accelerator beam tests, the output of the accelerator described in 
Appendix 9 was attached to a ~1 m diameter and ~1 m tall cylindrical vacuum 
chamber evacuated to < 10-6 torr.  The EPS instrument was mounted to a 
rotating turn-table that allowed the allowed the beam to be scanned over 5/6 
of the ~160? alpha dimension. (Rotation was possible only in one dimension.) 
The response of the instrument to beta-angle variations (i. e. over the ~12? 
FOV dimension) was not investigated with the accelerator beam.  Also, during 
the tests where the turn-table was free to rotate, the instrument tended to 
overheat because of poor thermal coupling with the surrounding structures.  
Therefore the angular response of the instrument was investigated only for 
basic functionality.  For most of the tests reported here, the angular 
position was frozen so that a liquid-nitrogen cooling system could be attached
to the EPS sensor.  

Directionality

Calibration runs like those depicted in Figure 21 document the ability of the 
EPS sensor determine the directionality of the incoming charged particles.  
Figure 21a shows the ability of the discrete “start” anodes to determine the 
directionality of the incoming particles for the Low Energy Ion events (no SSD
signals).  For this particular test, the collimator on this flight unit was 
not installed (Some tests are run without the collimator because the small 
hole pattern of the collimator interferes with the pencil-like characteristic 
of the accelerator beam in a fashion that makes it difficult to quantify the 
response of the detector.)  The beam was confined to shining on one small 
portion of the start foil by the narrowness of the ion beam.  Figure 21b shows
the ability of the discrete SSD’s to determine the directionality of those 
high-energy ions that stimulate the ion SSD’s.  



Issues that need to be addressed includes the level of electronic cross-talk 
between the adjacent discrete start anodes and the level of cross-talk between
the different pixels of each SSD wafer (each SSD wafer has 4 pixels).  The 
cross-talk between start anodes will comprise some combination of electronic 
noise and cross-anode drift of secondary electron trajectories.  These two 
effects will be sorted out with the engineering model and our simulation model
of secondary electron trajectories (Section 4).  Cross-talk between SSD pixels
will be quantified with the engineering model.  The engineering model is 
constructed to be sufficiently flight-like to reproduce the flight model 
characteristics.  



Calibration Beam Runs

Throughout most of the calibration activities where EPS was tested within the 
JHU/APL Calibration facilities, various problems were diagnosed (noise, FPGA 
programming, etc.) and were fixed.  Following the finalization of the EPS 
configuration for delivery and following the introduction of cooling hardware 
within the calibration facility to keep the EPS head at a temperature of ~0? 
C, a series of calibration runs were performed that constitute our best 
knowledge of the performance of the EPS instrument at the time of delivery.  
Those calibration runs are listed and described in Appendix 10.  Unless 
otherwise indicated, the figures discussed here are taken from the calibration
runs documented in Appendix 10.  

Energy Resolution.  Figures 21 through 25 show the energy and time-of-flight 
analysis performed for calibration runs #7, #1, #22, and #26, representing 
beams with 170 keV H+, 80 keV H+, 170 keV O+, and 110 keV O+.  FWHM energy 
resolutions achieved are: 12, 10, 12, and 15 keV.  These values are within a 
factor of 2 of expectations (Table 4.1) and are dominated by noise pickup from
the power supplies, so it is thought.  These results represent satisfactory 
performance and meet the requirements of the EPS system. 





Time-of-flight Resolution.  The FWHM of times-of-flight for the 4 examples 
shown in Figures 21 through 25 are ~4, ~4, ~6, and ~13 ns.  We have found 
that, intrinsically, the TOF spread ranges between ~4 and 6 ns for the EPS 
instrument as delivered to the spacecraft.  When EPS was delivered to the 
spacecraft we did not understand why the TOF spread was substantially broader 
than anticipated (< 2 ns).  We now understand that there are two causes: 1) 
the dispersive spread in the secondary electrons is broader than anticipated (
~3 ns; Section 4), and 2)  the performance of the Constant Fraction 
Discriminators (CFD) that analyze the timing pulses requires a higher bias 
voltage than previously appreciated in order that they follow the 
fast-changing MCP pulses.  As documented below, the CFD bias is adjustable, 
and it is believe that a ~3ns TOF spread will be achieved by EPS in flight.  







Time-of-Flight Versus Energy.  Figures 26 through 30 illustrate the 
characteristics of simultaneous Energy and TOF measurements.  Figure 26 shows 
a GSEOS screen capture of a 170 keV proton beam measurement, the same run used
to obtain Figure 22.  Figure 27 places a similar measurement (Run #5, 140 keV 
protons) in the context of the theoretical expectations for protons and helium
ions.  The figure demonstrates that, despite the somewhat broader energy and 
TOF spreads than those anticipated, relatively clean separation of the major 
mass species is anticipated.  The bottom panel of Figure 28 shows the TOF 
versus Energy characteristics as revealed using an energy-degraded Am-241 
alpha source.  That same data is replotted (as individual points) on the top 
panel of Figure 28, and represented as green crosses.  With the TOF spread of 
~5 ns, clean mass discriminations between H and He above ~1 MeV total energy 
may be problematic. With the anticipated improvements, discrimination to 
several MeV should be possible.  



The square red and green symbols represent the peak intensities of ion beams 
unambiguously identified as hydrogen and helium ions.  These peaks fall 
relatively close to theoretical expectations.  Figure 29 shows additional 
points.  Our experience with the JHU/APL accelerator was, at the time of these
calibrations, limited, and we did not always know the species and charge 
states of the various components that came out of the JHU/APL ion accelerator.
  In some cases for the additional points plotted in Figure 29, charge states 
were assigned in a fashion that cannot be proven.  .  The figure suggests, but
does not prove, that EPS identifies the different major mass species according
to expectations.  



Cross Talk?  There is the suggestion in the calibration data that there exists
some cross talk within the FPGA sorting of the different particle events.  
Some GSEOS screen captures like that shown in Figure 30 suggest that, when 
multiple species are present, some fraction of the TOF of one species may be 
paired with energy of another species.   Extensive pulser tests have failed to
identify this potential failure mode, and so, at this point in time, the issue
remains open.  Since the data like those shown in Figure 30 were accumulated, 
we have discovered that the Constant Fraction Discriminators (CFD) have been 
operated with non-optimal settings, and future tests with the engineering 
model will be performed to evaluate the presence - or absence - of such cross 
talk. 

Efficiency.  Efficiency of ion counting has been a continuing problem with the
EPS instrument during the early phases of the calibration efforts.  A rough 
estimate of ion TOF-only and TOF versus E counting is provided in the 27th and
28th rows of the Table in Appendix 10.  These two estimates are obtained by 
dividing the “valid TOF” events and the “ion events”, respectively, by the SSD
Fast Counts.  Note that for runs with multiple peaks, this estimate can easily
be a factor of 2 in error.  Once CFD discrimination levels were adjusted 
downward beginning with Run #9, reasonable TOF-only efficiencies were 
achieved. Note, for example, the efficiencies of 16.5% and ~23% during the 
proton Runs #15 and #23, respectively.  We noted in Section 4 that 
efficiencies of order 22% were expected for protons.  However, the “ion event”
efficiencies, depending on simultaneous TOF and energy measurements, reside in
the neighborhood of a fraction of 1 percent.  



As described below, post-delivery testing of the EPS engineering model 
indicates that high-efficiency operation requires higher MCP voltages than 
were achieved during the pre-delivery testing to yield a higher gain out of 
the MCP.  It appears that pre-delivery testing required multiple secondary 
electron emissions from the start and stop foils. Overall sensor efficiency is
highly sensitive to this condition.  A less than 10% probability of generating
sufficient secondary electron signal on the start and stop anodes yields a < 
1% efficiency for generating TOF-vs-electron events.  Still puzzling is the 
relatively high efficiencies achieved with the TOF-only events.  All of the 
results presented in this section regarding resolutions and mass separations 
are based on demonstrably valid TOF-versus-Energy events, not on TOF-only 
events.

Spacecraft Integration and Test Calibrations: TOF performance

Some understanding of the wider-than-expected TOF spreads was achieved with a 
combination of tests with the engineering model calibrations and flight-model 
during the spacecraft Integration and Test (I & T) activities.  Post delivery 
engineering model testing revealed that the performance of the Constant 
Fraction Discriminators (CFD’s) that process the start and stop signals from 
the MCP anodes are more sensitive to bias voltage than anticipated.  At low 
biases the CFDs apparently do not work fast enough give the speed of the MCP 
pulses.   During spacecraft I&T, an Am-241 radiation source was positioned so 
that it illuminated the entrance aperture of the flight EPS.  Using bias 
settings for the CFD’s commensurate with those values used during the 
accelerator beam calibrations yielded TOF spreads of order 4 ns.  When the 
bias voltage was increased substantially, TOF spreads of order 2 ns were 
achieved.  We now understand with the engineering model tests that followed 
the I&T activities (as described in the next section), that the TOF spread is 
somewhat sensitive to the CFD discrimination values, and so we do not believe 
that < 2 ns TOF spread will be achieved once the instrument settings are fully
optimized.  However, we believe that we now understand the sources of the TOF 
spreads, and that the expected performance presented in Table 4.1 (somewhere 
between 2.5 and 3.2 ns) will be achieved.  Thus, we anticipate achieving clean
mass discriminations to the highest energies addressed with the EPS system, ~3
MeV.

Summary of EPS flight-model calibrations

Ion-beam calibrations of the flight EPS instrument demonstrate that energy 
resolution (FWHM) of between 10 and 15 keV is achieved, which, while being a 
factor of 1.5 to 2 larger than is achievable with SSD sensors and electronics 
of the sort used here, is nonetheless adequate for meeting the MESSENGER EPS 
requirements.  For most of the calibration runs reported here, the observed 
TOF resolution (4-6 ns) was of the order of a factor of 2 wider than 
anticipated and could limit the ability of EPS to discriminate the major 
species H, He, and the CNO group at energies larger than ~1 MeV.  However, 
during the I&T period calibrations of the EPS flight instrument, data acquired
indicate that the anticipated TOF resolution of between 2.5 and 3.2 ns will be
achieved in flight, so that clean separation of the major species (H, He, CNO 
group) will be achieved over the entire range of energies covered by EPS: 
>5keV/nucleaon to ~3 MeV total energy.  

The major outstanding issues following calibration of the EPS flight model 
include: 1) low ion counting efficiencies, and 2) possible cross-talk in the 
identification of different mass species.  Major progress has been made in 
addressing the first issue with the engineering model, as described in the 
next section.  The second issue awaits attention.  Given that the cross-talk 
flaw is not revealed with extensive testing of the species identification 
firmware using pulsers, it is, perhaps, a reasonable expectation that the 
cross-talk issue will disappear once all other sensor parameters, including 
MCP gain (next section) and CFD performance, are optimized.  

6.0 ENGINEERING MODEL CALIBRATIONS   

A flight-like engineering model of EPS has been developed and is now complete.
 Detailed calibrations of this model will begin in November of 2004 and will 
be completed by ~1 June 2005..  Some calibrations have already taken place 
with this unit at various stages of its construction.  The differences between
the flight EPS and the engineering model used for the test to date are that 
the engineering model has had only 2 SSD’s installed, and it utilized 
so-called “rimmed” MCP’s rather than the rimless MCP used for flight.  The 
rimless plates are used for flight because they are considered more likely to 
survive handling and launch environments.  Rimmed plates, on the other hand, 
are likely to outgas much more quickly than are the rimless plates, give the 
mounting configuration used with EPS.  Two issues have been addressed to date:
1) the efficiency of ion counting, and 2) TOF spread performance.  



Ion Counting Efficiency

For the engineering model of EPS, a critical issue has turned out to be the 
length of time that the unit has spent outgasing.  For this test on ion 
counting efficiency, the MCP were allowed to outgas for several weeks within a
high vacuum environment.  After this outgasing, MCP plate voltages were raised
to ~2300 volts, and the CFD thresholds were lowered to values (60 Hex) 
substantially below those achieved with the calibration runs documented in 
Appendix 10, all without the MCP start and stop rates reaching unreasonable 
values.  Using an Am-241 alpha-particle radiation source, counting 
efficiencies of TOF-versus-energy events were as high as 35% (based on a ratio
of triple coincident “ion events” and the SSD counting events).  Further such 
testing will be performed beginning in Novermber of 2004 and will be complete 
by ~June 2005.  However, the engineering model has demonstrated to date that 
the ion counting efficiencies commensurate with the theoretical expectations 
described in Section 4 are achievable with EPS.  

TOF-Spread Performance

To investigate the factors that control the TOF spread, the engineering model 
was exercised by varying several of the EPS parameters, specifically the MCP 
plate voltage, the CFD discrimination values, and the CFD bias voltages.  In 
this case the engineering model was configured with flight-like rimless MCP 
plates, and again, the unit was allowed to outgas for several week within a 
high vacuum environment.  Sample results are shown in Table 6.1 below.

      Table 6.1 Engineering Model TOF spread performance.
MCP voltageCFD Discrim.CFD BiasTOF FWHM(V)(mV)(V)(ns)191012 
3.64.22030123.62.62030323.61.252030323.751.0

Clearly the TOF-spread performance is affected profoundly by MCP voltage and 
associated MCP gain, the CFD discrimination values, and the CFD bias voltage. 
TOF-spreads of order ~1 ns (bottom 2 rows) are unlikely to be achieved for an 
optimized EPS if, at the same time one requires high efficiency performance.  
The larger CFD discrimination values (e. g. 32 mV) are likely to be too large 
to achieve efficient counting.  The 2.6 ns value is very encouraging at a very
reasonable CFD discrimination value.  The actual TOF spread may be a little 
larger if the full breadth of the entrance aperture is illuminated with the 
source.  However, once full optimization is achieved, it is highly likely that
our advertised TOF spread performance of 2.5 to 3.2 ns will be achieved, and 
probably surpassed.  






Table 6.1 EPS Performance Parameters.
Geometric Factor (Instrument Total)0.074 cm2 srGeometric Factor (Electron 
Pixel)0.0062 cm2 srGeometric Factor (Low E Ion Pixel)0.012 cm2 srGeometric
Factor (High Energy Ion Pixel)0.0062 cm2 srGeometric Factor (Diagnostic 
Ions)0.0062Electron Efficiency vs. EnergyRoughly 100% at mid energies.  
Details in Appendix 8 (TBS 3/1/05)Low Energy Ion Efficiency vs. En x 
MRoughly 20%.  Details in Appendix 8 (TBS 3/1/05)High Energy Ion Efficiency 
vs En x MMeasured at several percent.  Higher values anticipated with tuning,
based on Engineering model. Details in Appendix 8 (TBS 3/1/05)Diagnostic Ion 
Efficiencies~100%Energy CoverageSee Table Below?E / E capability at 50 
keV10 – 15 keV?E / E capability at > 100 keV10 – 15 keVEnergy / TOF 
Channel resolutionElectronic Channels in Appendix 8.  Primary particle 
characterisitics TBS 3/1/05TOF dispersive spreadDemonstrated: 4 ns
Expected < 2.5 to 3.2 nsMass Species SeparationSee Table BelowAngular 
Coverage (total)160?  x 12? Angular Coverage (Electrons)147?  x 12?Angular
Coverage (Low Energy Ions)160?  x 12?Angular Coverage (High Energy 
Ions)147?  x 12?Angular Coverage (Diagnostic Ions)147?  x 12?Angular 
Pixels (Electrons)12? x 27? (13? with 13.6? gap)Angular Pixels (Low Energy 
Ions)12? x 30? (full ~27? coverage)Angular Pixels (High Energy Ions)12? x 
27?  (13? with 13.6? gap)Angular Pixels (Diagnostic Ions)12? x 27?  (13? 
with 13.6? gap)Electron Scattering Angle ContrastSeveral percentCollimator 
Side-Lobes: Position(??, ??) = (?28?, ?0?)Collimator Side-Lobe 
Fraction/Lobe~4% per lobeEnergy Coverage:TechniqueElectrons15 keV – 0.6 
MeVSingles / Foil TechniqueLE Ions5 keV – 1 MeVTOF OnlyLE CNO15 keV – 3 
MeVPH-based mass discrimination may be limited given TOF dispersionHE 
Protons15 keV – 2.8 MeVTOF vs. EnergyHE He25 keV – 2.8 MeVTOF vs. 
EnergyHE CNO60  keV – 3 MeVTOF vs. EnergyDiagnostic Protons15 keV – 2.8 
MeVSinglesEnergy / TOF ChannelsElectronic channels in Appendix 8.  
Primary particle characteristics TBS 3/1/04.
Summary

The engineering model has proven itself to be exceedingly useful for achieving
improved understanding of the performance of the EPS instrument.  When it is 
constructed into a fully flight-light configuration, it will be extensively 
tested to establish the optimum parameters for trading off such performance 
characteristics as ion counting efficiencies, TOF performance, anode 
cross-talk, SSD cross-talk, etc.  Here we update Table 4.1 concerning the EPS 
performance characteristics, making those changes that are appropriate to what
we have learned with the calibrations of both the flight EPS and the 
flight-like engineering model.  The few changes from Table 4.1 are indicated 
in blue.  

7.0 Future Plans

Final calibration and characterization of the EPS instrument will require two 
thrusts: 1) continued characterization and testing of the flight-like 
engineering model, once completed, and 2) extensive in-flight calibrations.

Engineering Model Plans

The engineering model is essentially complete with a full complement of SSD’s 
(Earlier engineering model tests have used only 2 sets of SSD’s) and rimless 
MCP’s.  It will then be extensively exercised in a bell jar with radiation 
sources to find the right set of parameters for optimum efficiency, TOF and 
Energy measurement performance (MCP voltage, CFD bias, CFD discriminators, 
Anode discriminators, SSD fast and shaped discriminators, etc.).  It will be 
tested extensively for anode-to-anode and SSD pixel-to-pixel cross-talk.  The 
existences and variations of such cross-talk will be folded into the 
optimization of the various settings.  When the optimization versus parametric
settings are well understood, the unit will be moved to the APL accelerator 
beam chamber for final characterization.  In addition, to better characterize 
the relative efficiencies of different mass species, we will investigate 
possible cross-talk in the valid-event identification circuitry that may occur
when multiple energies and species are measured at the same time.  

In-flight Calibration

Described below are the general plans for performing in-flight calibrations, 
as documented in the previously issued EPS Calibration Plans document 
(MESSENGER Energetic Particle and Plasma Spectrometer/Energetic Particle 
Spectrometer (EPPS/EPS) Calibration Plan [7389-9060]).  However, given the 
limited calibrations achieved with the actual flight instrument before 
delivery, these calibrations have increased in importance.  The single most 
important goal of these in-flight calibrations is to determine the efficiency 
of ion counting as a function of mass species.  This goal is easily achievable
for protons.  At times the interplanetary environment will be filled with 
energetic protons of solar origin and energized by interplanetary processes.  
Since we know that protons dominate such particles, cycling back and forth 
between the diagnostic ion mode (SSD measurements only) and the full 
triple-coincident High Energy Ion mode (TOF vs. E) will reveal the counting 
efficiency of the High Energy Ion Mode for protons.  The one modification from
the previously published plan is the need for more time (at least 6 months 
operating within interplanetary space to assure that sufficient intensities of
energetic ions are encountered.  

Determining the counting efficiency for the heavier ions is not so 
straightforward.  We will rely, in the first instance on scaling according to 
the dE/dX parameter and scaling from the proton results using calibrations of 
previously flown TOF-versus-Energy instruments.  Figure 31, for example, shows
calibration results for the Galileo Energetic Particle Detector 
TOF-versus-Energy sensor.   Operation of the EPS during the Earth encounter 
will also be exceedingly valuable because the environment is fairly well 
cataloged.  Earth-environment calibrations of EPS are now included in the 
official testing plan for the Earth encounter occurring ~1 year after launch. 
Finally operating EPS in the general environment where other interplanetary 
assets reside would also be exceedingly valuable, specifically in the vicinity
of Earth where such spacecraft as ACE and Wind are operating.  Our more 
general goals for performing in-flight calibrations are documented here.



Interplanetary Space.  Energetic particles in interplanetary space will be 
used to test the mass classification algorithm and verify the content of the 
various look-up tables. Observations by EPS during the instrument checkout 
give the first opportunity to cross check the pitch angle magnetic field 
information with particle signatures.  It is therefore important that the 
magnetometer be operational during the EPPS checkout period.   Comparison of 
particle distribution axes of symmetry with the magnetic field will be 
attempted with these data to validate the angular response of the instrument. 
Success will depend on the properties of the energetic particles at the time 
of observation.  

Because the interplanetary environment is variable, it is highly desirable to 
have the EPS instrument operating substantially longer than that planned for 
the checkout period.  Interplanetary energetic particles tend to vary on the 
time scale of a solar rotation: 27 days.  It would be highly desirable to 
operate the EPS sensor (with the magnetometer) for a cumulative time period of
at least 6 x 27 days prior to the first Mercury flyby in order to optimize the
chances of obtaining substantial, calibrating fluxes over the full EPS energy 
range. 

Venus Flybys.  The magnetosphere of Venus is characterized by the sporadic 
appearance of energetic particles. For example, previous missions have 
observed energetic particle acceleration at the Cytherean bowshock within the 
upstream solar wind.   The EPS sensor will be operated during the Venus fly-by
periods to collect and identify energetic particles around the planet. At 
Venus we will check out operational modes and parameter settings.

Mercury Flybys.  The Mercury fly-bys provide the only opportunities prior to 
the orbital phase of the mission to sample a magnetic field and plasma that 
are not solar wind or shocked solar wind.  The increased degree of ordering 
provided by the Mercury magnetic field allows superior inter-comparison of 
magnetometer and particle distribution symmetry axes.  The Mercury Fly-bys 
also offer the best opportunity to carry out dress rehearsals for the orbital 
phase in terms of operating modes and instrument parametric states.

Calibration Monitoring.  In flight, detectors degrade and gains change.  
Microchannel Plate (MCP) sensor gain changes are monitored by periodic 
measurements of the pulse height distributions from each of the 7 anodes using
built-in capabilities of the instrument.  MCP biases can be varied to correct 
shifts in gain and restore detection efficiency. 

Solid State Detector (SSD) performance is monitored using the Time-of-Flight 
(TOF) versus Energy (E) displays of the Pulse-Height-Analysis event data 
sampled at all times throughout the mission.  Degradations are apparent and 
quantifiable by noting shifts in the TOF vs. E species tracks.  These shifts 
are normalized to the ground calibration states so that shifts in energy 
band-passes can be determined and the corresponding look-up table be updated. 


8.0 REFERENCES

Agostinelli et al., GEANT4- a simulation toolkit, NIM(A), 506, 250-303, 2003.

Andrews, G. B., R. E. Gold, E. P. Keath, D. G. Mitchell, R. W. McEntire, R. L.
McNutt, Jr., and N. P. Paschalidis, Compact particle detector for space 
measurements: prototype performance, Proc. SPIE Conference 3442, Missions to 
the Sun II, 3442, 1998.

Biersack, J. P. and L. Haggmark, NIM, 174, 257, 1980.

Frischkorn , H. J., et al., Total yield and escape depth of electrons from 
heavy ion solid interactions, IEEE Transactions on Nuclear Science, NS-30, 
1983.

McNutt, Jr., R. L., D. G. Mitchell, E. P. Keath, N. P. Paschalidis, R. E. 
Gold, and R. W. McEntire, A compact particle detector for low-energy particle 
measurements, Proc. SPIE International Symposium, Optical Science, 
Engineering, and Instrumentation, Missions to the Sun, 2804, 217-227, 1996. 


Appendix 1:  EPS Hardware Counters

Address OffsetRate Counter Address OffsetRate CounterAddress OffsetRate 
Counter0x007Fast 00x014Shaped 10x021Anode 20x008Fast 10x015Shaped 
20x022Anode 30x009Fast 20x016Shaped 30x023Anode 40x00AFast 
30x017Shaped 40x024Anode 50x00BFast 40x018Shaped 50x025Electron 
Event0x00CFast 50x019Shaped 60x026Ion Event0x00DFast 60x01AShaped 
70x027Start (16 LSBs)0x00EFast 70x01BShaped 80x028Start (8 
LSBs)0x00FFast 80x01CShaped 90x029Stop (16 LSBs)0x010Fast 
90x01DShaped 100x02AStop (8 MSBs)0x011Fast 100x01EShaped 
110x02BValid TOF (16 LSBs)0x012Fast 110x01FAnode 00x02CValid TOF (8 
MSBs)0x013Shaped 00x020Anode 1
Appendix 2.  EPS Software Counters.


Name of software rate counterDescription1Electron-count number of  events
processed during the accumulation interval2High energy ion  number of  
events processed during the accumulation interval3Low energy ion number of 
events processed during the accumulation interval4Electron Pileup number of
events discarded due to pileup condition5Electron Multi-hits number of 
events discarded due to multiple hits6High energy ion pileup number of 
events discarded due to pileup condition7High energy multi-hit number of 
events discarded due to multiple hits




Appendix 3.  EPPS Status Packet Format
NameLength (bits)ValueDescriptionVersion3000Designates a source 
packetType10Designates a telemetry packetSecondary?11Secondary header 
presentApID11010 0000 0101Application process IDGrouping211 = 
NoneGrouping flagsSequence Count14Unsigned IntegerContinuous sequence 
count for each Application Process IDLength1632Secondary Header + Data 
(bytes) –1Secondary Header32Unsigned IntegerTime tag (MET)  Status 
Int.161 – 65535 (0=Off)Status interval (seconds)Macro Blocks16Unsigned 
IntegerNumber of macro blocks freeTLM Vol16Unsigned IntegerTelemetry 
volume produced (KB)Watch Addr16Unsigned IntegerMemory watch addressWatch
Mem8Unsigned IntegerWatched memory (pg. no)Watch Data2*8Unsigned 
IntegerWatched memorySW Version8Unsigned IntegerSoftware version 
numberAlarm Id8Latest Alarm IdAlarm Type10 = Persistent
1 = TransientLatest alarm typeAlarm Count7Unsigned IntegerCount of 
alarmsCmd Exec8Unsigned IntegerCommands executedCmd Reject8Unsigned 
IntegerCommands rejectedMac Exec8Unsigned IntegerMacro commands 
executedMac Reject8Unsigned IntegerMacro commands rejectedMacro 
Id8Unsigned IntegerId of most recent macro executedMacro Learn10 = Not 
learning
1 = LearningMacro learn modeMonitor Response1 0 = DIS
1 = ENMonitor responseWrite Enb10 = DIS
1 = ENMemory write enablespare50SpareEPPS Status infoNvariableEPPS 
specific status information (see following table)













EPPS Status Information Format
NamebitsValueDescriptionHVPS current 16integer valueExpansion Bus 
address 0x006HVPS voltage 16integer valueExpansion Bus address 0x008Bias 
current16integer valueExpansion Bus address 0x00ABias voltage16integer 
valueExpansion Bus address 0x00CEPS status word16integer valueExpansion 
Bus address 0x000+5 volt monitor10log compress 16 to 10 bitsI2C read of 
LVPS ADC on mux channel 0-5 volt monitor10log compress 16 to 10 bitsI2C 
read of LVPS ADC on mux channel 1+12 volt monitor10log compress 16 to 10 
bitsI2C read of LVPS ADC on mux channel 2-12 volt monitor10log compress 16
to 10 bitsI2C read of LVPS ADC on mux channel 3+5 current monitor10log 
compress 16 to 10 bitsI2C read of LVPS ADC on mux channel 9-5 current 
monitor10log compress 16 to 10 bitsI2C read of LVPS ADC on mux channel 
10+12 current monitor10log compress 16 to 10 bitsI2C read of LVPS ADC on 
mux channel 11-12 current monitor10log compress 16 to 10 bitsI2C read of 
LVPS ADC on mux channel 12EPS temperature10log compress 16 to 10 bitsI2C 
read of LVPS ADC on mux channel 7LVPS temperature10log compress 16 to 10 
bitsI2C read of LVPS ADC on mux channel 13Primary current monitor10log 
compress 16 to 10 bitsI2C read of LVPS ADC on mux channel 14Switched primary
current monitor10log compress 16 to 10 bitsI2C read of LVPS ADC on mux 
channel 1515V monitor10log compress 16 to 10 bitsFIPS extended address map
93-15V monitor10log compress 16 to 10 bitsFIPS extended address map 945V 
monitor10log compress 16 to 10 bitsFIPS extended address map 95-5V 
monitor10log compress 16 to 10 bitsFIPS extended address map 963.5V 
monitor10log compress 16 to 10 bitsFIPS extended address map 972.5V 
monitor10log compress 16 to 10 bitsFIPS extended address map 98FIPS Vref 
I2C Temp10log compress 16 to 10 bitsFIPS extended address map 99FIPS BH 
BOT I2C Temp10log compress 16 to 10 bitsFIPS extended address map 9aFIPS 
BH TOP I2C Temp10log compress 16 to 10 bitsFIPS extended address map 
9bFIPS DS-HV I2C Temp10log compress 16 to 10 bitsFIPS extended address map
9cMCP-HV monitor10log compress 16 to 10 bitsFIPS extended address map 
9dPA-HV monitor10log compress 16 to 10 bitsFIPS extended address map 
9eDSHV Monitor10log compress 16 to 10 bitsFIPS extended address map 
9fDSHV Step10log compress 16 to 10 bitsStep at which HV values were 
sampledFIPS BH BOT Temp10log compress 16 to 10 bitsI2C read of LVPS ADC on
mux channel 6FIPS BH TOP Temp10log compress 16 to 10 bitsI2C read of LVPS 
ADC on mux channel 8Spare8zero-filledPads to next word boundaryI2C read 
command address16I2C read addressSoftware (upper byte = 0xe for extended 
read address, 8 LSBs = read address)I2C read command result1616-bit 
integerSoftware



Appendix 4.  EPS High-Priority Data

Table 1. N1 Science Data Telemetry Packet Format (after fast decompression)
NameLength
(bits)ValueDescriptionVersion3000Designates a source 
packetType10Designates a telemetry packetSecondary?11Secondary header 
is presentApp ID11010 0001 0001EPPSGrouping211 = NoneGrouping 
flagsSequence Count14Unsigned integerContinuous sequence count for each 
Application Process IDLength163 – 4099Packet length (bytes) - 1Secondary 
Header32Unsigned integerTime tag (MET)High Energy Ions for detector s 
0-56*352=
211210-bit log compressed values within 16-bit words22 words (16 bits each) 
per detector. Each word contains a 10-bit log compressed count accumulated 
over N1 seconds
(132 words total).  In Diagnostic Mode.  First 8 words are Ion energy 
spectrum.  The rest are all zeros.Electron  (coarse) for detectors 
0-56*128=76810-bit log compressed values within 16-bit wordsEight 16-bit 
words per detector. Each word contains a 10-bit log compressed count 
accumulated over N1 seconds 
(48 words total)Electron (fine)
for 
detectors 0-5
(see note below)6*32*10= 192010-bit log compressed values within 16-bit 
wordsTwo 16-bit words per detector corresponding to a 12-word set. A separate
12-word set is reported for each of the ten  N1/10 second subintervals. Each 
16-bit word contains a 10-bit log compressed value.  (120 words total)Low 
Energy Ion (TOF only)
for 
detectors 0-56*256 = 153610-bit log compressed values within 16-bit 
wordsSixteen 16-bit words per detector. Each word contains a 10-bit log 
compressed count accumulated over N1 seconds. 
(96 words total)
All zeros in Diagnostic Mode.Rate Channel counter & Software 
Rates42*16=67210-bit log compressed values within 16-bit words35 Hardware 
rate counters and 7 software rate counters accumulated of N1 seconds and each 
counter is log compressed to 10 bits and stored within a 16-bit word
(42 words total)Housekeeping52816 bit values33 words of housekeeping data 
recorded over N1 seconds (see table 2)PHA768 to 1344Up to ten sets of 
values (64 to 112 bits each)Up to 10 PHA events (varying in size from 4 to 7 
words) acquired over N1 seconds



Table 2. EPPS housekeeping values contained in High Priority Telemetry Packet
1HVPS setting2Bias setting3TOF anode 0 disc setting4TOF anode 1 disc 
setting5TOF anode 2 disc setting6TOF anode 3 disc setting7TOF anode 4 
disc setting8TOF anode 5 disc setting9TOF Start CFD setting10TOF Stop 
CFD setting11Heavy-0 Disc setting12Heavy-1 Disc setting13Ion Fast 
channels 1, 3, 514Ion Shaped channels 1, 3, 515Ion Fast channels 7, 9, 
1116Ion Shaped channels 7, 9, 1117Electron Fast Channels 0, 2, 
418Electron Shaped Channels 0, 2, 419Electron Fast Channels 6, 8, 
1020Electron Shaped Channels 6, 8, 1021Command word A mirror22Command 
word B mirror23Event parameter A mirror24Event parameter B mirror25Event
parameter C mirror26HVPS clock adjust value27Bias clock adjust 
value28LVPS Control word29Invalid channel ID count30EPS-fifo-reset 
count31I2C Bus error count32Value provided by most recent bus read 
command33Spare

For the “Fine” electron data, super bin 0 is sum of bin 0 to bin 3, while 
Super bin 1 is the sumfrom bin 4 to bin 7.  The subset is fixed.




Appendix 5.  EPS Medium-Priority Data

N2 Science Data Telemetry Packet Format (after fast decompression)
NameLength
(bits)ValueDescriptionVersion3000Designates a source 
packetType10Designates a telemetry packetSecondary?11Secondary header 
is presentApp ID11010 0001 0010EPPSGrouping201 = First
00 = Continuation
10 = Last
11 = NoneGrouping flagsSequence Count14Unsigned integerContinuous 
sequence count for each Application Process IDLength163 - 4099Packet 
length (bytes) – 1Secondary Header32Unsigned integerTime tag (MET)High 
Energy Ion115210-bit log compressed valuesTwelve 16-bit words per detector.
Each word contains a 10-bit log compress matrix box  accumulated over N2 
seconds 
(72 words total). ).  In Diagnostic Mode.  First 8 words are Ion energy 
spectrum.  The rest are all zeros.Electron  (coarse)76810-bit log 
compressed valuesEight 16-bit words per detector. Each word contains a 10-bit
log compress matrix box  accumulated over N2 seconds 
(48 words total)Electron (fine).  (See note Appendix 4)192010-bit log 
compressed valuesTwo 16-bit words per detector. Each word contains a 10-bit 
log compressed super box accumulated over 10 subintervals of N2/10 seconds 
(120 words total)Low Energy Ion (TOF only)76810-bit log compressed 
valuesEight 16-bit words per detector. Each word contains a 10-bit log 
compressed histogram box accumulated over N2 seconds.  (48 words total).
All zeros in Diagnostic Mode.Rate Channel counter & Software Rates38410-bit
log compressed values17 Hardware rate counters and 7software rate counters 
accumulated over N2 seconds. Each counter is log compressed
 to 10-bits and stored within a 16-bit word.
(24 words total)PHA0 to 2240Up to twenty  variable sized PHA eventsUp to 
20 PHA events (varying in size from 4 to 7 words) acquired over N2 seconds (80
to 140 words total)







The specific 17 counters will be command-able through a table upload.  
Currently the default counters are:

    12 Shaped counters
    Electron Event
    Ion Event
    Start
    Stop
    Valid TOF



Appendix 6.  EPS Low-Priority Data

Low Priority Science Data Telemetry Packet Format (after fast decompression)
NameLength
(bits)ValueDescriptionVersion3000Designates a source 
packetType10Designates a telemetry packetSecondary?11Secondary header 
is presentApp ID11010 0001 0011EPPSGrouping211 = NoneGrouping 
flagsSequence Count14Unsigned integerContinuous sequence count for each 
Application Process IDLength163 - 4099Packet length (bytes) - 1Secondary 
Header32Unsigned integerTime tag (MET)N PHA Events0 to 33600N unsigned 
integersUp to 300 PHA Events  (varying in size from 4 to 7 words)  (0 to 2100
words)The actual number of words in each PHA event depends on the event 
type.

Appendix 7.  Pulse Height Analysis Telemetry Data Formats

word 0 (0x03)                EPS Electron PHA Event  Short ID (8 bits)word 
1Energy Peak (8 LSBs)word 2Energy Baseline (8 LSBs)word 3Energy Peak (4 
MSBs)Energy Baseline ( 4 MSBs)word 40000MultipleElectron Channel ( 3 
bits)
word 0 (0x01)                EPS Low Energy Ion PHA Event Short ID (8 
bits)word 1TOF (8 LSB)word 2000Heavy Disc 1Heavy Disc 0TOF ( 3 
MSB)word 300Start Segment  6 bits
Word 0(0x05)   EPS High Energy Ion PHA Event Short ID (8 bits)Word 1Energy 
Peak ( 8 LSBs)Word 2Energy Baseline (8 LSBs)Word 3Energy Peak (4 
MSBs)Energy Baseline (4 MSBs)Word 4TOF ( 8 LSBs)Word 5MulitpleIon 
Channel (3 bits)TOF (3 MSBs)Word 6Heavy Disc-1Heavy Disc-0Start Segment 
(6 bits)
word 0 (0x0B)                EPS Diagnostic Event PHA Event Short ID (8 
bits)word 1Energy Peak (8 LSBs)word 2Energy Baseline (8 LSBs)word 
3Energy Peak (4 MSBs)Energy Baseline ( 4 MSBs)word 40Heavy Disc-1Heavy 
Disc-0MultipleIon firedIon or Electron Channel 
( 3 bits)word 500Start Segment (6 bits)Appendix 8.  Channel Definitions 
(Missing Table entries to be supplied by 3/1/05)

Electron Channels (Nominal Parameters)
ChannelE1 
(elect)E2(elect)E1(in)E2(in)Eff-1Eff-2keVkeVkeVkeVfractionfraction
1025225393396046094594146614622772
273528352547Notes:  “elect” refers to energy deposited in the SSD
and read out by the electronics.  The “in” parameters are the mean input 
energies, taking into account energy losses and scattering.  “Eff” is 
efficiency.  Eff-1 and Eff-2 are efficiencies at the end-point energies.  The 
“in” parameters and the “efficiency” is calculated using GEANT-4


Low Energy Ion Channels (Nominal Parameter) (Exact Organization TBS 3/1/05)
ChannelSpeciesTOF_1TOF_2E1 E2Eff-1Eff-2L, M, HnsnsHy or OHy or 
Ofractionfraction1L01.12L1.17.53L7.540.54L40.521
55L2156M01.17M1.17.58M7.540.59M40.5215
10M21511M01.112H1.17.513H7.540.514H40.5
21515H21516Notes:  “L, M, H” refers to light ions, 
medium ions, and heavy ions, based on the range of pulse-height settings: L = 
noise threshold to H-0 discriminator, M = H-0 to H-1 discriminator, and H = > 
H-1 discriminator.  For calculation of E1 and E1, it is assumed that L = 
Hydrogen and H = Oxygen.  Eff-1 and Eff-2 are efficiencies at the end points 
calculated assuming that a 100 keV proton generates a 0.75 electrons from a 
foil, and then scaling with dE/dX.  


High Energy Ions (Nominal Parameters):  See Figure at end of Appendix 8
ChannelSpeciesE1(el)E2(el)E1(in)E2(in)Effi-1Eff-2keVkeVkeVkeVfra
ctionfraction0BG1P,282P28683P681644P16439
65P95823166P>23167He281648He1649589He>9
5810Heavy<2811Heavy288112Heavy8123313Heav-L233
67314Heav-M23367315Heav-H23367316Heav-L673194117
Heav-M673194118Heav-H>194119Heav-L>194120Heav-H>1941
21BGNotes:  E1(el) and E2(el) is  electronic energy read out 
from SSD.  TOF information is contained here only in the identification of the
species.  The E1(in) and E2(in) is the input energy as calculated with SRIM.  
Efficiency is based on dE/dX scaling from protons.

Ion Diagnostic Channels (Nominal Parameters)
ChannelE1 
(elect)E2(elect)E1(in)E2(in)Eff-1Eff-2keVkeVkeVkeVfractionfraction
1025112253911339601146094115941461161462
2711722735211835254711Notes:  “elect” refers to energy 
deposited in the SSD and read out by the electronics.  The “in” parameters are
the mean input energies, taking into account energy losses and scattering.  


High Energy Ions Notional Channels




Appendix 9.  The JHU/APL Calibration Facility

Calibration of the EPS instrument was conducted at the JHU/APL instrument 
testing facility through a combination of long-duration, radiation-source 
exposures (Am-241 degraded alpha source) and discrete exposures from our 170 
kV particle accelerator system.  The Air Insulated Accelerator, obtained from 
Peabody Scientific, is a versatile system capable of producing a broad range 
of ion species to energies of 170 keV.  The accelerator incorporates a number 
of features that result in reliable, easy operation.

The system includes a Duoplasmatron Ion source, extraction gap, Einzel Lens 
and Wien filter mounted in the insulated terminal structure along with all 
associated power supplies.  A large diameter accelerator tube accelerates the 
beam to it’s final energy.  Within the accelerator tube is an electrostatic 
bi-potential lens that is used to focus the beam to its final size at the 
target position.  

The system will produce beams of H, O, and noble gas ions with intensities 
over the range of 100’s to 1,000,000 particles/cm2/sec at the target position 
in an energy range of 3 keV to 170 keV.

Typical operation of the system is as follows.  Ions are formed in the ion 
source and are extracted to energies of up to 20 keV by the extraction gap.  
The high-brightness, divergent beam is then focused by the terminal Einzel 
lens through the mass analysis system to a focus at the entrance to the 
accelerator tube.  The accelerator tube provides additional acceleration up to
the maximum beam energy, or deceleration to minimum beam energy.  The beam is 
focused by an electro-static lens to an appropriate size at the target 
position.

The Duoplasmatron is a low pressure arc discharge source in which the gas to 
be ionized is electrostatically constricted by a funnel-shaped intermediate 
electrode placed between the filament and the anode.  A strong axial magnetic 
field is developed between the intermediate electrode and the anode, which 
further constricts the discharge to a narrow plasma beam along the axis of the
exit aperture.

The optical quality of the beam depends upon the extraction region of the 
source-lens combination.  The Duoplasmatron source and extraction gap are 
designed to operate as a unit and optimize the efficiency of this 
Pierce-extraction region.  The lens insulators are optically shielded from the
ion beam to prevent surface charging or contamination from the vacuum system. 
This prevents the formation of a spark tracks on the insulators which lead to 
lens voltage breakdown.  The extraction electrode is made from copper for good
heat conduction and is nickel-plated to prevent corrosion.  It is liquid 
cooled by a cooling line external to the vacuum thus eliminating any possible 
leakage in the vacuum.  The divergence of the beam from the ion source is 
related to the extraction geometry and extraction voltage.  The gap geometry 
provides a strong electric field region to extract ions efficiently  from the 
source plasma, resulting in a superior quality beam.

The Einzel lens is a three-element electrostatic cylindrical tube lens with 
the center electrode elevated near to source potential.  Operated in this 
manner, the lens is a classical saddle point Einzel lens which decelerates the
divergent beam from the source and extraction gap and then reaccelerates it 
back to the extraction energy.  The result is a strong focusing action.  The 
focal length is determined by the voltage applied to the center electrode.  

The terminal structure is a strong, rigid welded frame that supports the ion 
source components at high voltage.  The terminal has a smooth, four-inch 
radius outer shell that provides equipotential control of the high voltage.  
This design prevents corona discharge and results in highly stable voltage 
operation.


Appendix 10.  Calibration Runs Listing 

Table A10.1 groups several accelerator calibration runs when the EPS 
instrument was in a flight configuration, including the collimmator.  A total 
of 26 discreet runs are reported here along with the date and EST of each run.
 The Accelerator Parameter – Energy lists the energy of the ion incident on 
the EPS instrument in keV (20 keV extraction voltage + post acceleration 
voltage = Energy).  The Accelerator Parameter – ExB is the actual accelerator 
setting of the Wien Filter used to select the desired species, the intended 
species is also listed.  Which of the 6 EPS detectors and associated time of 
flight (TOF) Anode is listed.  The Discriminator setting for both the Ion Fast
and Ion Shaped channels is listed.  The setting for the High Voltage Power 
Supply (HVPS) and Solid State Detector (SSD) bias is listed, as well as the 
settings for the TOF Anode, TOF Start constant fraction discriminator (CDF), 
TOF Stop CDF, and both Heavy Ion Discriminators.  The next group represents 
the counts observed during the run as a function of instrument cycle.  The 
count rate of the Starts, Stops, Valid TOF, and Ion Events (triples) is listed
as well as the count rate observed on SSD from the Shaped, and Fast systems.  
The count rate on the associated anode is also listed.

The group of values that begin with the word “analyzed” represent analyses of 
the reported energy and TOF peak and full width at half of the maximum (FWHM) 
given in bin number.  Comments about the analyses are also shown.  The group 
labeled “Main Spectra” is to indicate, for those runs with multiple peaks, 
which of those were used in the overall calibration spectra.



Table 10.1

Table 10.1 Continued
Appendix 11.  Data ApIDs and Data Rates

Here is a list of the ApIDs, and other interesting info related to them, 
organized by FILE priority:

High-Priority Data
(1) ApID 211 (hex) - EPS High-Priority Science, typical integration interval =
300 seconds, typical size (compressed) = 373 bytes
(2) ApID 221 or 225 or 229 or 22D (hex),  - FIPS High-Priority Science, 
typical integration interval = 600 seconds, typical size (compressed) = 872 
bytes. Note: there are four ApIDs listed, but only one type is output at any 
given time. The different ApIDs reflect different operating modes that FIPS 
will be in.
(3) ApID 205 (hex) - EPPS status packet, 10-second integration interval, 
typical size = 80 bytes. Note: This packet is output every 10 seconds, but we 
want to put only one out of every 120 in the high priority file. The rest 
should go into contingency.

Medium-Priority Data
(1) ApID 212 (hex) - EPS Medium-Priority Science, typical integration interval
= 30 seconds, typical size (compressed) = 304 bytes
(2) ApID 222 or 226 or 22A or 22E (hex),  - FIPS Medium Priority Science, 
typical integration interval = 60 seconds, typical size (compressed) = 550 
bytes. Note: there are four apIDs listed, but only one type is output at any 
given time. The different ApIDs reflect different operating modes that FIPS 
will be in.

Low-Priority Data
(1) ApID 213 (hex) - EPS Low Priority Science, typical integration interval = 
300 seconds, typical size (compressed) = 600 bytes
(2) ApID 223 or 227 or 22B or 22F (hex),  - FIPS Low Priority Science, typical
integration interval = 600 60 seconds, typical size (compressed) = 430 bytes. 
Note: there are four ApIDs listed, but only one type is output at any given 
time. The different ApIDs reflect different operating modes that FIPS will be 
in.

Contingency Data
(1) ApID 205 (hex) - EPPS status packet, 10-second integration interval, 
typical size = 80 bytes. Note: This packet is output every 10 seconds, but we 
want to put 119 out of every 120 in the contingency file. Every 120th packet 
should go in the high priority file.

Assumptions for the above numbers: FAST compression gives 3 to 1 compression, 
PHA slots mostly filled, FIPS burst mode not used.

This gives the following BYTE rates for the various priorities:
High = 2.76 bytes per second
Medium = 19.3 bytes per second
Low = 2.71 bytes per second
Contingency = 8 bytes per second


??

??

??

??









REV.BY & DATEDESCRIPTIONCHECKAPPROVED & DATE
Prepared by Stefano Livi, 
Bruce Andrews, Dennis Haggerty, George Ho, and Barry MaukJuly 29, 
2004On-ground calibration report for the EPPS/EPS 
instrument.