Burr, Steven R
THE DESIGN AND IMPLEMENTATION OF THE DYNAMIC IONOSPHERE
CUBESAT EXPERIMENT (DICE) SCIENCE INSTRUMENTS
by
Steven R. Burr
A thesis submitted in partial fulfillment
of the requirements for the degree
of
MASTER OF SCIENCE
in
Electrical Engineering
Approved:
Dr. Charles M. Swenson
Major Professor
Dr. Chris Winstead
Committee Member
Dr. Edmund A. Spencer
Committee Member
Dr. Mark R. McLellan
Vice President for Research and
Dean of the School of Graduate Studies
UTAH STATE UNIVERSITY
Logan, Utah
2013
ii
Copyright
c Steven R. Burr 2013
All Rights Reserved
iii
Abstract
The Design and Implementation of the Dynamic Ionosphere Cubesat Experiment (DICE)
Science Instruments
by
Steven R. Burr, Master of Science
Utah State University, 2013
Major Professor: Dr. Charles M. Swenson
Department: Electrical and Computer Engineering
Dynamic Ionosphere Cubesat Experiment (DICE) is a satellite project funded by the
National Science Foundation (NSF) to study the ionosphere, more particularly Storm Enhanced Densities (SED) with a payload consisting of plasma diagnostic instrumentation.
Three instruments onboard DICE include an Electric Field Probe (EFP), Ion Langmuir
Probe (ILP), and Three Axis Magnetometer (TAM). The EFP measures electric fields from
±8V and consists of three channels a DC to 40Hz channel, a Floating Potential Probe
(FPP), and an spectrographic channel with four bands from 16Hz to 512Hz. The ILP measures plasma densities from 1x104 cm−3 to 2x107 cm−3 . The TAM measures magnetic field
strength with a range ±0.5 Gauss with a sensitivity of 2nT. To achieve desired mission
requirements careful selection of instrument requirements and planning of the instrumentation design to achieve mission success. The analog design of each instrument is described
in addition to the digital framework required to sample the science data at a 70Hz rate and
prepare the data for the Command and Data Handing (C&DH) system. Calibration results
are also presented and show fulfillment of the mission and instrumentation requirements.
(101 pages)
iv
Public Abstract
The Design and Implementation of the Dynamic Ionosphere Cubesat Experiment (DICE)
Science Instruments
by
Steven R. Burr, Master of Science
Utah State University, 2013
Major Professor: Dr. Charles M. Swenson
Department: Electrical and Computer Engineering
Dynamic Ionosphere Cubesat Experiment (DICE) is a cubesat satellite project funded
by the National Science Foundation (NSF) to study the ionosphere. Cubesats are small
satellites in the shape of a cube around 10cm on a side, and allow better access to space.
Three main properties of the ionosphere are measured by the DICE mission, which are
electric field, magnetic field, and plasma density with an instrument for each. The limitation
of power, mass, and volume contributes to the difficulty of cubesat design. Mission and
instrumentation requirements must be carefully planned to ensure mission success. Each
instrument’s requirements and design are described in detail from an electrical engineering
and spacecraft design perspective. In addition, calibration results are provided for each
instrument. DICE is an example of advanced satellite development and also pioneered
mechanism and instrumentation methods due to the number and complexity of instruments
in a small volume.
v
This thesis is dedicated to my parents, James and Michelle Burr,
and also to Andrew C. Christensen whose educational outreach to a student eventually led
the student to write this thesis.
vi
Acknowledgments
There were many people who were involved with the DICE satellite project; it was a
privilege to work with each one. Special thanks to both Chad Fish and Charles Swenson
for tirelessly writing proposals so students can have top tier projects to work on such as
DICE, and also for working with a bunch of students. Tim Nielsen helped write portions of
the digital section and some of his work is featured there. Keith Bradford also contributed
to the Electric Field Boom design and has a thesis featuring the mechanics of the design.
Also, thanks to Dr. Swenson and Aroh Barjatya who helped develop many figures for the
calibration section. None of this work would be possible without the staff of Space Dynamics Laboratory, especially as they allowed students to have access to their resources and
knowledge.
Steven R. Burr
vii
Contents
Page
Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
iii
Public Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
iv
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
vi
List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
x
List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
xi
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . .
1.1 Small Satellite Constellation Missions . .
1.2 DICE Mission Overview . . . . . . . . . .
1.2.1 DICE Science Objectives . . . . .
1.2.2 The DICE Satellite . . . . . . . . .
1.3 DICE Science Board . . . . . . . . . . . .
1.4 Overview of Thesis . . . . . . . . . . . . .
1
1
3
5
7
9
10
2 Ion
2.1
2.2
2.3
2.4
2.5
....
. . .
. . .
. . .
. . .
. . .
. . .
....
. . .
. . .
. . .
. . .
. . .
. . .
Langmuir Direct Current Probe . . . . . . . . . .
Ion Langmuir Probe Concept . . . . . . . . . . .
ILP Mechanical . . . . . . . . . . . . . . . . . . .
Sweeping Langmuir Probe Instrument Design . .
Digital-to-Analog Converter for Sweeping Voltage
Ion Langmuir Probe Sensitivity and Range . . .
3 Electric Field Probe . . . . . . . . . . . . . . . . . . .
3.1 Electric Field Probe Concept . . . . . . . .
3.2 Alternate Configurations . . . . . . . . . . .
3.3 Electric Field Probe Mechanical . . . . . . .
3.4 Guarding and Shielding . . . . . . . . . . .
3.5 Expected Signal . . . . . . . . . . . . . . . .
3.6 Electric Field Direct Current Channels . . .
3.7 Floating Potential Probe . . . . . . . . . . .
3.8 Electric Field Alternating Current Channel
....
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
4 Three Axis Magnetometer . . . . . . . . . . . . . . . . . .
4.1 Magnetometer Concept . . . . . . . . . . . . . .
4.2 Magnetometer Mechanical Layout . . . . . . . .
4.3 Magnetometer Implementation . . . . . . . . . .
4.4 Electrical Design of Magnetometer . . . . . . . .
4.5 Magnetometer Sensitivity and Dynamic Range .
.
.
.
.
.
.
.
.
.
.
.
.
.
....
. . .
. . .
. . .
. . .
. . .
. . .
....
. . .
. . .
. . .
. . .
. . .
....
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
....
. . .
. . .
. . .
. . .
. . .
. . .
....
. . .
. . .
. . .
. . .
. . .
.....
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
.....
. . . .
. . . .
. . . .
. . . .
. . . .
....
. . .
. . .
. . .
. . .
. . .
.....
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
.....
. . . .
. . . .
. . . .
. . . .
. . . .
...
. .
. .
. .
. .
. .
. .
....
. . .
. . .
. . .
. . .
. . .
....
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
....
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
....
. . .
. . .
. . .
. . .
. . .
.
.
.
.
.
.
. . 13
.
13
.
13
.
15
.
16
.
17
.
.
.
.
.
.
.
.
.
. . 21
.
21
.
22
.
23
.
23
.
24
.
25
.
27
.
27
....
. . .
. . .
. . .
. . .
. . .
. . 31
.
31
.
32
.
33
.
37
.
38
viii
5 Payload Digital System . . . . . . . . . . . . . . . . . . . . . . . .
5.1 Data Subsystem Overview . . . . . . . . . . . . . . . .
5.2 Digital Requirements . . . . . . . . . . . . . . . . . . .
5.3 Payload Controller and Data Acquisition Subsystem .
5.3.1 Field-Programmable Gate Array . . . . . . . .
5.3.2 FPGA Firmware . . . . . . . . . . . . . . . . .
5.3.3 Payload Controller Commands . . . . . . . . .
5.3.4 Analog-to-Digital Converters . . . . . . . . . .
5.3.5 Sampling Methods . . . . . . . . . . . . . . . .
5.3.6 EFAC Spectrometer FFT . . . . . . . . . . . .
5.3.7 ILP Sweep Mode . . . . . . . . . . . . . . . . .
5.3.8 Sample Timing . . . . . . . . . . . . . . . . . .
5.4 Data Flow Down . . . . . . . . . . . . . . . . . . . . .
5.4.1 Real-Time Clock . . . . . . . . . . . . . . . . .
5.4.2 Data Packets and Granules . . . . . . . . . . .
5.4.3 Packet Formation and the C&DH . . . . . . . .
5.4.4 Radio-to-Ground Station . . . . . . . . . . . .
....
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
....
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
....
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
...
. .
. .
. .
. .
. .
. .
. .
. .
. .
. .
. .
. .
. .
. .
. .
. .
41
41
42
43
43
44
45
46
47
48
51
51
52
52
53
54
55
6 Calibration and Testing Results . . . . . . . . . . . . . . . . . . . . .
6.1 Instrumentation Testing and Calibration . . . . . . . . . . .
6.2 Ion Langmuir Probe Testing and Results . . . . . . . . . . .
6.2.1 Ion Langmuir Probe Calibration . . . . . . . . . . .
6.2.2 Ion Langmuir Probe Calibration Results . . . . . . .
6.3 Electric Field Probe Testing and Results . . . . . . . . . .
6.3.1 Electric Field Probe Calibration . . . . . . . . . . .
6.3.2 Electric Field Probe Calibration Results . . . . . . .
6.4 Three Axis Magnetometer Testing and Results . . . . . . .
6.4.1 Helmholtz Coil Calibration . . . . . . . . . . . . . .
6.4.2 Three Axis Magnetometer Testing and Calibration .
6.4.3 TAM Calibration data . . . . . . . . . . . . . . . . .
....
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
.
.
.
.
.
.
.
.
.
.
.
.
....
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
...
. .
. .
. .
. .
. .
. .
. .
. .
. .
. .
. .
57
58
58
58
59
61
61
61
62
62
64
66
7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7.1 DICE Comparison . . . . . . . . . . . . . . . . . . . .
7.1.1 Initial Mission Success . . . . . . . . . . . . . .
7.1.2 Instrumentation Performance and Comparison
7.2 Future Work . . . . . . . . . . . . . . . . . . . . . . .
7.2.1 ASSP . . . . . . . . . . . . . . . . . . . . . . .
7.2.2 Satellite Constellations . . . . . . . . . . . . . .
7.3 Lessons Learned . . . . . . . . . . . . . . . . . . . . .
7.3.1 Science Board . . . . . . . . . . . . . . . . . . .
7.3.2 Ion Langmuir Probe . . . . . . . . . . . . . . .
7.3.3 Electric Field Probe . . . . . . . . . . . . . . .
7.3.4 Magnetometer . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
....
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
....
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
.....
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . 69
.
69
.
69
.
70
.
70
.
70
.
71
.
72
.
72
.
73
.
73
.
74
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
ix
Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Appendix A
. . . . . . . . . . . . . . . . . . .
A.1
Science and Mission Instrumentation
A.2
Science Board Schematics . . . . . .
.............
. . . . . . . . . .
Requirements . .
. . . . . . . . . .
....
. . .
. . .
. . .
.....
. . . .
. . . .
. . . .
. . 77
.
78
.
78
.
82
x
List of Tables
Table
Page
1.1
Science and mission functionality requirements traceability matrix. . . . . .
7
2.1
ILP component parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . .
15
2.2
DAC component parameters. . . . . . . . . . . . . . . . . . . . . . . . . . .
17
2.3
DAC system parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
18
2.4
ILP system performance.
. . . . . . . . . . . . . . . . . . . . . . . . . . . .
20
2.5
ILP temperature error. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
20
3.1
EFP system preformance. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
27
4.1
AMR bridge parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
32
5.1
FPGA parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
44
5.2
EFAC spectrometer channels freqency bins. . . . . . . . . . . . . . . . . . .
51
5.3
DAS sample timing table. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
53
6.1
Helmholtz coil calibration results. . . . . . . . . . . . . . . . . . . . . . . . .
64
6.2
Gain matrix for Yhatzee calibration. . . . . . . . . . . . . . . . . . . . . . .
67
7.1
DICE performance and comparison. . . . . . . . . . . . . . . . . . . . . . .
71
A.1 Mission and instrumentation requirements. . . . . . . . . . . . . . . . . . .
78
xi
List of Figures
Figure
Page
1.1
Satellite mechanical diagram. . . . . . . . . . . . . . . . . . . . . . . . . . .
4
1.2
Storm enhanced densities over North American sector. . . . . . . . . . . . .
6
1.3
DICE satellite structure and bus. . . . . . . . . . . . . . . . . . . . . . . . .
11
1.4
DICE science payload. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
11
1.5
DICE functional diagram. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
12
2.1
ILP mechanical and sensor map. . . . . . . . . . . . . . . . . . . . . . . . .
14
2.2
ILP system noise calculations. . . . . . . . . . . . . . . . . . . . . . . . . . .
19
3.1
Electric field system functional diagram. . . . . . . . . . . . . . . . . . . . .
22
3.2
Electric field mechanical and sensor map. . . . . . . . . . . . . . . . . . . .
23
3.3
Electric field system modeled inputs. . . . . . . . . . . . . . . . . . . . . . .
25
3.4
EFDC functional diagram. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
26
3.5
FPP functional diagram. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
28
3.6
EFAC filtering. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
29
3.7
EFAC functional diagram. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
29
3.8
EFAC spice noise simulation. . . . . . . . . . . . . . . . . . . . . . . . . . .
30
4.1
Science payload and instrumentation.
. . . . . . . . . . . . . . . . . . . . .
33
4.2
TAM mechanical and sensor map. . . . . . . . . . . . . . . . . . . . . . . .
34
4.3
Set/reset voltage for AMR bridge. . . . . . . . . . . . . . . . . . . . . . . .
35
4.4
Magnetometer offset. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
36
4.5
Magnetometer functional diagram. . . . . . . . . . . . . . . . . . . . . . . .
38
4.6
TAM system parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
40
xii
5.1
DICE data collection system. . . . . . . . . . . . . . . . . . . . . . . . . . .
42
5.2
FPGA firmware. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
45
5.3
Payload system commands. . . . . . . . . . . . . . . . . . . . . . . . . . . .
46
5.4
DAS sampling scheme. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
49
5.5
DAS functional data plot. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
50
5.6
ILP sweep sampling scheme. . . . . . . . . . . . . . . . . . . . . . . . . . . .
52
5.7
RTC register update. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
54
5.8
DAS granule types and byte sequences.
. . . . . . . . . . . . . . . . . . . .
55
5.9
Science granules in a CCSDS packet. . . . . . . . . . . . . . . . . . . . . . .
56
5.10 Ground station functional diagram. . . . . . . . . . . . . . . . . . . . . . . .
56
6.1
Temperature chamber. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
57
6.2
Yhatzee in temperature chamber ready for calibration. . . . . . . . . . . . .
59
6.3
Linear fits of ILP calibration data. . . . . . . . . . . . . . . . . . . . . . . .
60
6.4
First-order linear fit and ILP calibration data. . . . . . . . . . . . . . . . . .
60
6.5
ILP noise residuals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
61
6.6
EFP linear fit and residuals. . . . . . . . . . . . . . . . . . . . . . . . . . . .
62
6.7
Helmholtz coil calibration and zero-Gauss chamber. . . . . . . . . . . . . . .
63
6.8
Triangle wave input for TAM calibration. . . . . . . . . . . . . . . . . . . .
65
6.9
TAM calibration inputs for Yhatzee testing. . . . . . . . . . . . . . . . . . .
66
6.10 TAM calibration data and residuals. . . . . . . . . . . . . . . . . . . . . . .
67
6.11 TAM gain and offset variations over temperature. . . . . . . . . . . . . . . .
68
7.1
72
Future satellite constellation. . . . . . . . . . . . . . . . . . . . . . . . . . .
1
Chapter 1
Introduction
1.1
Small Satellite Constellation Missions
The most significant advances in space science over the next decade are most likely to
derive from new observational techniques. The connection between advances in scientific
understanding and technology has historically been demonstrated across many disciplines
and time. There are clear ties between advances in our understanding of the processes in
the space environment and the deployment of new sensing techniques, from new vantage
points, which fuel new discoveries. Newer sensing techniques, such as X-ray and UVimaging, Global Positioning System (GPS) based measurements, energetic neutral atom
imagers, and others, along with the ability to place these sensors above the atmosphere and
within regions of interest, have revolutionized space science [1,2]. The study of the Sun and
Earth system requires multi-point observations to develop understanding of the coupling between disparate regions: solar-wind, magnetosphere, ionosphere, thermosphere, mesosphere
on a planetary scale. Changing environmental conditions in the near-Earth region of space
are important to our technological civilization that relies in complex and unexpected ways
on satellites and their proper function. The term “Space Weather” is growing in usage to
describe the study of changes in the ambient plasma, magnetic fields, radiation, and the
thin upper atmosphere that effect the health of satellites. Distributed multi-point measurements are needed to develop understanding of the space weather processes that occur across
temporal and spatial scales. The need for this capability can be seen in the conceptual scientific investigations presented in National Aeronautics and Space Administration (NASA)
roadmaps for science which call for various sizes of satellite constellations [3]. It is clear that
the science and predictive capability of space weather will advance with the development of
2
satellites constellations that can provide multi-point measurements from within the space
environment.
Remote imaging, or cameras, is one source of multi-point measurements of space
weather processes. The field of view of the imager is a multi-point set of observations
of the scene, but not all parameters of interest can be observed through this technique.
Electric field patterns and currents flowing along magnetic field lines, both of which are
important quantities for understanding the coupling of regions, cannot be sensed through
remote imaging and must be sensed in situ by instrumentation. The global and distributed
observations of these quantities require constellations of spacecraft. Even remote imaging
data can be improved by multi-point observations from constellations to aid in understanding global phenomena such as atmospheric tides and auroral storms, or to improve the
combined observations through advanced signal processing methods such as tomography.
The development of satellites which enables both in situ and remote sensing from multiple
points within the space environment is a high priority for the advancement of space weather.
The resources that will be available over the next decades for all areas of space weather
research have limits, and it is therefore important to find ways to leverage the costs of developing new technologies to advance science. The high-cost of access to space, at first review,
is a serious impediment to making multi-point measurements within the space environment or in other words in deploying constellations of traditional satellites. It is therefore
desirable to develop much smaller and lower-cost sensor/satellite systems such that the
largest number of distributed measurements can be economically made in the space environment. The smaller the mass and volume of the sensor/satellite the larger the number
will be that can be deployed from a single launch vehicle. The space engineering community is creating miniaturized sensors and satellite systems by leveraging the enormous
investment of commercial, medical, and defense industries in producing highly capable,
portable and low-power battery-operated consumer electronics, in-situ composition probes,
and novel reconnaissance sensors. The advancement represented by these technologies has
direct application in developing small sensor/satellite systems for space weather research.
3
Within this thesis, we describe the development and testing of a miniaturized set of scientific instruments for space weather research. These instruments have been developed for the
Dynamics Ionospheric CubeSat Experiment (DICE) which consists of two 1.5 unit CubeSats flying as a small constellation to demonstrate multi-point in-situ measurements of the
space environment from a small sensor/satellite system.
1.2
DICE Mission Overview
The “Dynamic Ionosphere CubeSat Experiment” or “DICE” mission was selected and
funded by the National Science Foundation in October 2009 in response to a cooperative
proposal from Utah State University’s Space Dynamics Laboratory (USU/SDL), ASTRA
Inc., and Embry Riddle University. DICE is one of several missions that have been flown
or are currently in development under NSF’s CubeSat-based Science Mission for Space
Weather and Atmospheric Research program. Launch was provided by NASA through the
Educational Launch of Nano-satellites (ELaNa) program. DICE consists of two identical
“CubeSats” deployed on October 27, 2011 as secondary payloads from a Delta II rocket. The
National Polar-Orbiting Operational Environmental Satellite System Preparatory Project
(NPP) was the prime payload of the launch. After NPP was placed into its Sun-synchronous
polar orbit the upper stage of the Delta II was restarted and the two DICE spacecraft were
released into an 809 to 457km at 102◦ inclination with one satellite following the other.
Both spin stabilized satellites are expected to remain on orbit for about 15 years with
the functional life of the spacecraft limited by the number of charge/discharge cycles the
batteries can withstand. The scientific purpose of the mission is to study space weather
phenomena that occur in the Earth’s ionosphere during geomagnetic storms. The payload
of each CubeSat consists of three science instruments, an Ion Langmuir Probe (ILP) to
measure in-situ ionospheric plasma densities, an Electric Field Probe (EFP) to measure
DC and AC electric fields, and a Three Axis Magnetometer (TAM). Figure 1.1 illustrates
the DICE spacecraft configuration and scientific instrumentation location. The two DICE
spacecraft are identical in design and function, and conform to a 1.5U CubeSat form factor
(10 x 10 x 17cm).
4
Ion Langmuir
Probe
Electric Field Probes
extend on booms to 5m
Ion Langmuir
Probe
Three Axis
Magnetometer
Fig. 1.1: Satellite mechanical diagram.
The four EFP wire booms, shown in partial deployment in Figure 1.1, extend 5m
outward from the spacecraft including their probes. Centrifugal force due to the spacecraft
spin is used to both deploy and hold them in proper orientation. The four shorter booms on
the bottom-side of the spacecraft comprise the Ultra-High Freqency (UHF) communications
turnstile antenna and are 0.2m in length. The UHF booms also provide balance for the
controlled spin of the spacecraft about the axis along the Langmuir probes. The ILP
sensor spheres are supported on the top and bottom of the spacecraft by extending scissor
booms that extend 8cm away from the spacecraft. The electronics for the EFP, TAM,
and ILP are housed in the spacecraft on as single, highly integrated, board. On-board
GPS measurements provide daily navigational updates accurate to within 1µs and 10m.
The DICE mission is using government radio bands that are consistent with a government
funded mission. A new half-duplex UHF modem developed by L-3 Communications for
DICE provides a 3 Mbit/s downlink and a 19.2 Kbit/s uplink. Both spacecraft share uplink
5
and downlink frequencies but have unique logical addresses decoded by the modem. The
ground stations are at Wallops Island on the East coast and at SRI on the West coast
where 18m UHF dish antennas are used to track and communicate with both spacecraft.
The DICE spacecraft can be considered a prototype for a small “space buoy” that would
be deployed in large numbers to observe electric fields, electron density, and magnetic fields
in the Earth’s ionosphere.
1.2.1
DICE Science Objectives
The distribution of plasma in the Earth’s ionosphere is dramatically different during
geomagnetic storms times than that found during quiet times. Many storm-time characteristics have only been described recently, including the Storm Enhanced Density (SED)
bulge and plume features. Much of this storm-time morphology has only become apparent
with the recent availability of Two-Dimensional Total Electron Content (TEC) mapping
across the US using GPS receivers. Three-dimensional ionospheric simulations are helping
to reveal the corresponding vertical variation of electron density. An example of an SED
plasma plume occurring over North America during solar storms is shown in Figure 1.2.
Several important research questions are still unanswered. First, how exactly the greatly
enhanced plasma is formed over the southern USA (the SED bulge) and what is the source
of the plasma. Second, exactly what physical drivers are involved in the formation and
evolution of the SED plume, and what is their relative importance. Finally, the precise
relationship between the occurrence of penetration electric fields, the subsequent expansion of the Appleton anomaly crests, and the development of SED is still an open research
question, particularly in terms of why there is an apparent preference for the USA geographic sector shown in Figure 1.2 [4,5]. Ultimately, the large redistributions of ionospheric
plasma interfere with radio communications and the SED plume causes GPS navigation
blackouts for users over North America. Since modern society has come to rely upon radio
and more increasingly GPS, the ability to understand and predict space weather effects on
these services is of importance.
The DICE mission will provide insight and measurements for further understanding
6
November 2 0 , 2 0 03 Storm
October 3 0 2 0 0 3 Storm
50
50
Plume
B ulge
Plume
40
B ulge
30
30
20
20
10
10
a) 21:15 UT 10/30/03
0
10
12
14
16
18 LT
40
b) 20:00 UT 11/20/03
12
14
16
18
20 LT
0
Fig. 1.2: Storm enhanced densities over North American sector.
of the formation, evolution, and decay of SED and their related impact on space weather
forecasting. In particular, the mission will provide simultaneous electric field and electron
density measurements in the early afternoon sector where SED events seem to form. The
DICE mission will focus on local times between about 12-17 LT, complementing the current
Defense Meteorological Satellite Program data, and together provide dayside electric field
measurements across a broad swath of local times.
The DICE science team developed a set of requirements to achieve the science goals and
how those requirements trace down to instrument parameters and the mission concepts. A
summary of these requirements and their flow down are presented in Table 1.1 [6]. The subrequirements that pertain to each set of instrumentation derived from the mission concepts
are available in the appendix. The three science objectives for the DICE mission are listed
and correspond to questions about the SED bulge and plumes, their formation and motion.
Table 1.1 also lists the minimum required measurement parameters of electric fields and
density and basic instrument requirements for the investigation including the range and
minimum sensitivity level for each science objective as determined by the DICE science
team. In general, all of these minimum requirements were exceeded during the design
phase of the mission.
7
Table 1.1: Science and mission functionality requirements traceability matrix.
Science Objective
Measurement Req.
Measure
RMS
Fluctuations
in
Electric Field and
Plasma Density:
1. Make co-located
DC electric field
and plasma density
measurements at a
≤ 10 km on-orbit
resolution
2. Make AC electric
field measurements
at a ≤ 10km onorbit resolution
3. Make measurements on a constellation platform of
≥ 2 spacecraft that
are within 300km
1: Investigate formation of the SED bulge over the USA
Instrument Req.
Mission Req.
Electric Field:
1. Constellation size ≥ 2 satellites
1. Max range of ± 0.6 2. Spacecraft spin ≥ 0.8 Hz
V/m
3. Spacecraft spin axis aligned to
2. Min threshold of 0.6 geodetic axis to within 10◦ (1σ)
mV/m
4. Spacecraft spin stabilized to within
3. Min resolution of 0.15 1◦ (1σ) about principal spin axis
mV/m
5. Spacecraft knowledge to within
4. DC sample rate ≥ 4 1◦ (1σ) Constellation time synch
Hz
knowledge ≥ 1s
5. AC sample rate ≥ 4 6.
Orbital insertion inclination
kHz [Telemeter AC FFT between 55 - 98◦ (ideally Sunpower information at ≥ 1 synchronous at 12-16LT)
Hz (3 points)]
7. Orbital altitude between 350 - 630
Plasma (Ion) Density: km
1.
Range of 2x109 - 8. ‘Circular’ orbits (eccentricity of ≥
13
2x10 m−3
0.2)
2. Min resolution of 3 9. Spacecraft ∆V speed of ≥ 50
x108 m−3
km/month
3. Sample rate ≤ 1 Hz
10.
Storage/downlink ≤ 31
Mbits/day.
11. Lifetime ≥ 6 months
Science Objective 2: Investigate formation of the over the
SED plume over the USA
Measurement Req.
Instrument Req.
Mission Req.
Same as Science Ob- Same as Science Objec- Same as Science Objective (downlink
jective
tive
included in Objective 1)
Science Objective 3: Investigate correlation of PPE
with formation and evolution of SED
Measurement Req.
Instrument Req.
Mission Req.
Same as Science Ob- Same as Science Objec- Same as Science Objective (downlink
jective
tive
included in Objective 1)
1.2.2
The DICE Satellite
The configuration of the DICE spacecraft are the result of a number of drivers and requirements including the limited resources of $1.2 M to build, test, launch, and operate two
spacecraft. The most important driver was the need to conform to the Poly-Picosatellite
Orbital Deployer (P-POD) containerized launch system for secondary payloads [7], which
8
both restricted the combined volume (10 x 10 x 17cm) and mass (4 kg) of the mission. The
major science driver was the need to deploy the electric field sensors, some 10 meters, from
the spacecraft and spin them such that observational errors could be identified and removed
from the data. This required a method to store a set of four booms 5m in length from a
compact volume (9.5 x 9.5 x 1.5cm) which could later be released to their full length. A
secondary science objective required the ILP density probes to be placed away from the
spacecraft. This required a spring loaded boom or “scissor” booms to accomplish this task.
The TAM was designed to be deployed on the scissor booms to reduce magnetic noise but
was relocated to the top deck of the spacecraft during final integration. Issues with reliable deployment of the scissor booms made this adjustment to the spacecraft configuration
prudent. A memory shape alloy Frangibolt locking mechanism developed by TiNi locks the
ILP, EFP, and antennas in place during launch. Keeping design complexity to a minimum
was a general rule (although where technical demands warranted this rule was ignored);
off-the-shelf components were purchased to facilitate this philosophy. The GPS, L3 Cadet
Radio, Clyde Space solar cells, Electronic Power System (EPS), and Pumpkin Command
and Data Handling (C&DH) system were purchased. The Attitude Control and Determination System (ACDS), boom deployment system, satellite structure, and science payload
were developed in house. The Attitude Determination and Control System (ADCS) system keeps the satellite in a spin stabilized attitude with torque coils based on information
received from the Sun-sensor and GPS system. The GPS also synchronizes the spacecraft
clocks for time-stamping measurements. A Command and Data Handling is the central
means by which the spacecraft functions and means by which it is controlled. The C&DH
system ensures that the spacecraft is operating correctly, creates housekeeping data, and
organizes and compresses data from the payload. Both the radio and payload can be reset
in the event of an upset. Commands from the ground station are received by the L3 cadet
radio which sends them to the C&DH system for execution. The radio also stores up data
packets that are prepared for downlink by the C&DH system, commands from the ground
station start the downlink of data. The EPS provided by Clyde Space can store energy
9
from the solar cells and generates up to 1.8 Watts of power [8]. The EFP wire booms are
released by the Frangibolt along with the antenna booms which are shown in Figure 1.3.
Temperature is also recorded in housekeeping data by several temperature sensors located
throughout the spacecraft. Primary means of keeping time are done by a Real-Time Clock
(RTC) which is located on the science payload Field-Programmable Gate Array (FPGA)
and synchronizes time with the GPS. The C&DH system then synchronizes time with the
primary real time clock.
The satellite structure was formed from one piece of 7075 aluminum by a wire electrodischarge machining process. Mass was reduced by cutouts in the structure. These cutouts
also to allowed access to the satellite bus during assembly. Four standoffs separated each
board that consists of the satellite bus and are connected to the main structure by spacers.
Aluminum plates with cutouts for mechanisms were installed on the top and bottom to form
the rest of the structure. The ILP booms and antenna mechanisms are spring-loaded and
released by a one-time release bolt. To release the mechanisms a bolt fires which is spring
loaded, this causes the EFP spool to unlock and the ILP to unlock. By requirement only
non-magnetic materials were allowed to avoid interference with the magnetometer. The
satellite initial inertial state was a minor axis spinner of the z-axis. To turn the spacecraft
into a major axis spinner around its z-axis, tungsten weights were added to the tips of each
antennas and extending them up to 20 cm away from the spacecraft.
1.3
DICE Science Board
The electronics for all three science instrument are housed in the spacecraft on a single
Printed Circuit Board (PCB) science boardweighing 75 grams with a dimension of 90.2mm x
95.9mm as shown in Figure 1.4. Due to the compact nature of the board and components,
a 6-layer PCB design was nessacary. The analog and digital sections of the board were
physically and electrically separated to reduce Electro-Magnetic Interference (EMI) near the
instrumentation. A functional diagram is shown in Figure 1.5. While operating the board
consumes <250 mW of electrical power with the majority (100mW) being consumed by the
magneto-resistive devices of the TAM. The PCB included the digitization, time stamping,
10
and data formatting such that packets of data that passed to the on-board computer only
needed to be stored in the telemetry buffer without additional processing.
The DICE EFP, ILP, and TAM data provided as products sampled simultaneously at
either 35 or 75 of Hz each. Internally, the channels are oversampled at a rate of 17kHz and
the EFP data is then subjected to an onboard Fast Fourier Transform (FFT) to produce
four AC spectrometer channels up to 512Hz. The EFP (including AC FFT spectra) and
ILP data is then co-added and decimated for improved Signal-to-Noise Ratio (SNR).
An Actel ALG600 FPGA was chosen for control of the instrumentation in the place of a
microprocessor. This was done to reduce design time through simplification; in addition FPGAs are well-suited to the measurement co-location requirements of the mission. Communication with the instrumentation is done by a Universal Asynchronous Receiver\Transmitter
(UART) running at 56.2 kBaud. Several two byte commands change the three sampling
modes (off, on 35Hz, and on 70Hz) and deployment commands. The FPGA also contains
the control and electronics for the boom deployment system. The analog instrumentation
required ±12V generated switching regulators and a 1.25V and 2.5V reference, which pulled
power from a single 5.8V regulator connected to the 7V-8.2V unregulated battery bus. The
digital electronics drew power from the EPS regulated 3.3V and 5V power busses and also
generated 1.2V locally for the FPGA. The satellite bus contained both UART and power
connections through a PC 104 connector. A PC 104 connector and four standoff mounts
on the corners mechanically connected spacecraft and science payload.
1.4
Overview of Thesis
The rest of this document is organized in the following manner. The design and func-
tionality of each instrument is then described in detail with a chapter for each in the
following order: Chapter 2 details the Ion Langmuir Probe, Chapter 3 with the Electric
Field Probe, and Chapter 4 with the Three Axis Magnetometer. Chapter 5 is about data
management: the data processing of the FPGA and the formation of data packets sent
down to the ground. Chapter 6 includes results on the testing and calibration result for the
three instruments.
11
Fig. 1.3: DICE satellite structure and bus.
Fig. 1.4: DICE science payload.
12
Wire Booms
S1
S2
S3
S4
Electric Field
Instrument
ILP +Z
Probe
Wire Boom
Breaking System
Piezoelectric
Motor
Wire boom
Encoder
Boom Deployment
Controller
Torque Coil
X - axis
Torque Coil
Y-axis
Real
Time Clock
Science
GPS
Attitude Determination
and Control System
Sun Sensor
Command and Data Handling
(C&DH)
Magnetometer
(HMC1043)
Torque Coil
Z-axis
Frangibolt
Mechanism
Deployment
Electronic Power System
Solar Cells
L3 Cadet Radio
Lithium
Polymer
Batteries
Satellite Bus
ILP -Z
Probe
Magnotometer
Antennas
Fig. 1.5: DICE functional diagram.
13
Chapter 2
Ion Langmuir Direct Current Probe
2.1
Ion Langmuir Probe Concept
An ILP measures the plasma density of the ionosphere by observing its electrical
impedance. A voltage is applied to a probe, the change in the impedance can be used
to measure the density and temperature of the ionosphere [9]. Plasma parameters such as
electron density Ne, ion density Ni, and total electron temperature Te can be derived from
the current measured by the probe [10, 11]. USU/SDL has flown many Langmuir probes
on various missions in the past. The DICE ILP will need to use less power and volume
than previous instrumentation. The density of the atmosphere is required to be measured
from a range of 1x104 cm−3 to 2x107 cm−3 with a resolution of 1000cm−3 (for a total list of
requirements see A.1). The ILP dynamic range has been determined by the science team
to be ±50uA to meet the density measurement which also defines a resolution of 1.525nA.
A typical Direct Current Probe fixes the voltage at a single bias point; however, the DICE
ILP is required to sweep its voltage from -4V to 2V to make measurements in the electron
current region. A Digital-to-Analog Converter (DAC), controlled by the FPGA, allows for
sweeping. The ILP probes rest atop two booms on the top and bottom of the spacecraft to
prevent interference from the spacecraft’s plasma density wake.
2.2
ILP Mechanical
The DICE ILP mechanical probe design raises the booms above the plasma density
wake region, as shown in Figure 2.1. Each probe consists of a 1.27cm in diameter gold plated
aluminum probe, the center being located 13cm away from the top of the spacecraft and
21cm away from the center of the spacecraft. Deployable booms were required to stay within
14
Probe
Ion Langmuir
Probe +Z
Sensor is stowed
then springs up
G10 Tube
G d
Guard
Delrin
Delri
in Boom
ez
ey
ex
Spacecraft Body
Coordinates
Ion Langmuir
Probe -Z
Fig. 2.1: ILP mechanical and sensor map.
15
the launch envelope during launch. The probe assembly rests atop a delrin scissor boom
which is stowed away in the top and bottom spacecraft panels. A locking mechanism holds
the booms into place until they are released. Upon release, the ILP booms release, swing
out, then spring forward to their final position. A tube built from G10 insulates the probe
from the voltage guard and also holds the probe assembly together. Exposed potentials
can create a problem with measuring density as it can create additional current into the
probe. To mitigate this problem, no exposed potentials were permitted per requirement on
the spacecraft structure, and a guard was placed at the base of the probe. A bias potential
from the guard shields the plasma from electric fields that may come from potentials on the
spacecraft.
2.3
Sweeping Langmuir Probe Instrument Design
The DICE ILP design had to conform to the sensitivity requirement of 1.525nA but
also needed to be low-power. Only 40mW of power was allocated per requirement, which
restricted component selection to those that were low-power and low-noise. An AD8622
operational amplifier was selected for these reasons; other desirable characteristics are shown
in Table 2.1. Instead of using a two operational amplifier design for the difference amplifier,
an INA129U instrumentation amplifier was selected to reduce usage of circuit board space.
Setting the gain is another advantage because instrumentation amplifiers only require one
gain resistor. The INA129 instrumentation amplifier is also a low-noise, high-precision part.
Table 2.1: ILP component parameters.
Parameter
AD8622
INA129
Operating Temperature -40 to 125 -40 to 125
Input Voltage Vcc ±2.25 Vcc ±2.25
Quiescent Current
0.215
0.7
Temperature Offset Drift
1.2
0.5
Output
Vcc ±1
Vcc ±1
Short Circuit Current
±40
+6/-15
9
Input Impedance
10 ||5.5
1010 ||2
GBW
560
700
Voltage Noise at Input(1kHz)
11
10
CMRR
120
100
Units
◦C
V
mA
uV/◦ C
V
mA
Ohms||pF
kHz
nV/Hz1/2
dB
16
To understand how the system operates, a discussion on the circuit follows. A transimpedance amplifier measures the input current with a gain of Ktrans which is equal to
10000V/A. A voltage bias Vbias is also applied to the probe through the transimpedance
amplifier. The amplifier output equation is (2.1).
(Iin ∗ Ktrans ) + V bias = Vtrans
(2.1)
Biasing the probe and its resultant shift in the signal reduce the dynamic range of the
system. To center the signal around 0V, the INA129U subtracts the voltage bias from the
transimpedance signal. The INA129U is also used to gain the signal by Kdif f shown in
equation (2.2).
(Vtrans − V bias ) ∗ Kdif f = Vout−dif f
(2.2)
Noise is an issue within the system, by requirement the instrument shall limit bandwidth to 40Hz. A low-pass filter located after the instrumentation amplifier accomplishes
this task. Lastly, an amplifier stage is used to center the ±2.5V signal by adding and offset
of 2.5V. This centers the input of the Analog-to-Digital Converter (ADC) around 2.5V with
a range of 0V to 5V. The gain Klpf of the low-pass filter and the amplifier are 1V/V for frequencies below 40Hz. Upon further inspection of the schematics in the appendix, it should
be noted that the low-pass filter was implemented improperly. This happened because a
resistor was added between the low-pass capacitor and the last amplifier. To correct this
mistake the resistors between the instrumentation amplifier and was lowered to a value of
402Ohms, and when added to the filter resistors that make up the low-pass filter is equal
to a gain of 1.
2.4
Digital-to-Analog Converter for Sweeping Voltage
The DAC was added later to enable the probe to sweep its voltage instead of hold it
at a fixed bias of -7V as originally planned. The key to the design was finding a DAC that
would add less than 4mW of power to the system. The DAC7621 is a low-power (2.5mW),
17
12-bit DAC. The main feature is a simple parallel interface, more parameters are shown in
Table 2.2. The DAC7621 has a settling time of 7uSec which is about 0.05% of the system
sampling rate of 14.3mSec. It also has good accuracy and typically falls to within 50% of
the desired voltage value.
With a range of only 0V to 4.096V, the DAC7621 output signal needs further amplification to meet a range of ±7V that is needed to cover any desired sweeping voltage. An
AD8622 was used to convert the DAC signal with a 2.8V/V gain and 1.85V offset. A unity
gain amplifier buffers the signal to provide the needed current for the other amplifiers and
the voltage guard. The DAC system does contribute significant noise to the system mainly
in the form white noise as shown in Table 2.3, this noise is gained up by factor of 2.8V/V
from the amplifier. There is also noise added from the unity gain amplifier. The noise at
the output of the system is 9.74uV-rms. The magnitude of the noise was not calculated
until after the system was built.
2.5
Ion Langmuir Probe Sensitivity and Range
Noise limits the sensitivity in electronic systems. The noise level must be less than
the desired signal sensitivity of the system. Dynamic range is determined by the most
limiting point in the system that restricts the signal range. The ILP system amplifier stage
was designed to make the ADC the most limiting factor of the dynamic range. Excluding
protection diodes for simplicities sake, the ADC has a range of 5V. The dynamic range of
the system is found by taking the ADC range and dividing it by each system gain shown
in (2.3).
Table 2.2: DAC component parameters.
Parameter DAC7621
Units
Settling Time 7
uSec
Operating Temperature -40 to 85
C◦
Step 1
mV/Count
Output Current (Code 800h) ±7
mA
Relative Accuracy ±1/2 (2 Max) LSB
Differential Nonlinearity ±1/2 (1 Max) LSB
18
Table 2.3: DAC system parameters.
DAC7621 System Total Noise
System Bandwidth
DAC White Noise Amplitude
DAC White Noise Amplitude
Total From DAC White Noise
Gain of AD8622 on after output of DAC
Total Noise from AD8622
Gained up noise from DAC
Total Noise output from DAC System
Total Noise output from DAC System
ADC Range
Kdiff ∗Ktrans ∗Klpf
= (5V )
1A
10000V
1V
5V
Value
40
0.55
550
3479
2.8
78.8
9739
9740
9.74
1V
1V
Units
Hz
uA/Hz1/2
nV/Hz1/2
nV-rms
V/V
nV-rms
nV-rms
nV-rms
uV-rms
= 100 uA = ±50 uA
(2.3)
The system sensitivity is found by multiplying the inverse of the system gains and the
ADC range (which is also the dynamic range) divided by the ADC counts as shown in
equation (2.4) and numerically computed in equation (2.5).
Kdiff ∗Ktrans ∗Klpf
ADC Range
(ADC counts) =
216 counts
100 uA
216 counts
100 uA
=
655.36 counts
1.525 nA
∝
1 uA
1 count
(2.4)
655.36 counts
1.525 nA
∝
1 uA
1 count
(2.5)
The total noise of the system was calculated at the input of the probe for comparison
with the input signal. The noise was found by taking the noise at the output of each
amplifier and dividing it by the amplifier gain to find the equivalent noise at the amplifier
input and added by the sum of the squares to the amplifier input referred noise. This
process was repeated until the input noise of the last amplifier was found. The system
SNR can then be found which is the total dynamic range divided by the total noise at the
same point in the system. In the case of the ILP, the noise is a factor of 17 less than the
signal assuming no contribution from the DAC system which was added later. A spice noise
analysis was ran for the circuit and for the system bandwidth and is shown in Figure 2.2.
19
Total system preformance and calculations can be found in Table 2.4.
The frequency cutoff for the system bandwidth is at 40Hz integrating the spice noise
figure from 0.1 to 40Hz yields a value of around 1.56uA for the noise floor level, however,
this result also excludes DAC noise. The actual noise for the system should be somewhere
in between these two values.
The ILP system is affected by temperature offsets. As per requirement the temperature
of any instrument cannot exceed -10◦ C or 30◦ C. This results in a temperature differential
of 40◦ C. The temperature error was calculated at the input of the ADC. The error (shown
in Table 2.5) is calculated by finding the temperature error for each part and multiplying
it by the gain of each successive amplifier until the ADC. All error is collectively summed
to get total projected temperature error which is ±308uV at the input of the ADC and is
±0.006% of the signal.
V(DCP_OUT)
½
1.2µV/Hz
½
1.1µV/Hz
½
1.0µV/Hz
½
0.9µV/Hz
½
0.8µV/Hz
½
0.7µV/Hz
½
0.6µV/Hz
½
0.5µV/Hz
½
0.4µV/Hz
½
0.3µV/Hz
½
0.2µV/Hz
½
0.1µV/Hz
100mHz
1Hz
10Hz
100Hz
1KHz
Fig. 2.2: ILP system noise calculations.
10KHz
100KHz
20
Table 2.4: ILP system performance.
Total Noise
Signal input Range of System
Noise from ADC ADS8343
Noise from ADC ADS8343
Total Noise from AD8622
Total Noise from INA129U
Gain of 3rd amplifier stage AD8622
Gain from 2nd Amplifier Stage INA129
Gain of 1st Amplifer Stage AD8622
Noise at input 3rd amplifer stage AD8622
Noise at input 2rd amplifer stage INA129
Noise at input 1st amplifer stage AD8622
Noise at input 1st amplifer stage AD8622
Noise at input 1st amplifer stage AD8622
SNR
SNR
Value
100
20
20000
78.8
259
1
5
10000
20000
4008
4008
5668
5.67
17.64
12.46
Units
uA
uV-rms
nV-rms
nV-rms
nV-rms
V/V
V/V
V/A
nV-rms
nV-rms
nA-rms
nA
uA
Unitless
dB
Table 2.5: ILP temperature error.
Temperature Error
Upper Temperature
Lower Temperature
Temperature Coeff of AD8622
Temperature Coeff INA129
Gain of INA129
Temperature Differential
Error of Trans. Amplifier AD8622
Error of Trans. Amplifier With Gain
Error of INA129
Error of LPF Amplifier Stage AD8622
Total
Value
30
-10
1.2
0.5
5
40.0
48.0
240.0
20.00
48.0
±308.0
Units
◦C
◦C
uV/◦ C
uV/◦ C
V/V
◦C
uV
uV
uV
uV
uV
21
Chapter 3
Electric Field Probe
3.1
Electric Field Probe Concept
The electric field probe was part of DICE’s primary objective to measure gradients
caused by the electric field in the ionosphere. Instrumentation is required to make power
spectral measurements across 16Hz to 512Hz for high-frequency or Alternating Current
(AC) measurements and 0Hz to 40Hz for low frequency or Direct Current (DC) measurements. The challenge was designing a system within the constraints of the requirements.
Only 40mW of power and 16cm2 of circuit board space for the electronics were allocated
by requirement.
The DICE electric field concept uses the double probe technique to measure the ambient
*
electric field E with a suite of instruments [10]. The Electric Field instrument suite has
four separate instruments attached to the two probe sets consisting of two DC channels, a
floating potential channel, and an Electric Field Alternating Current (EFAC) channel shown
in Figure 3.1. The Electric Field Direct Current (EFDC) channels enable the measurement
of the ambient electric field in two dimensions. The Floating Point Potential Channel was
not initially proposed but later added to the EV12 channel to measure spacecraft charging
potential. The EFAC channel obtains the AC component of the signal with a high-pass
filter to measure the power spectral density of the plasma. Each instrument has an analog
front end which consists of a precision instrumentation amplifier. The desired frequency
range for each instrument is obtained by both passive and active filters, which also mitigate
noise. To match the ADC voltage, the signal is gained and the signal level is offset on each
channel. The FPGA controls the ADCs and simultaneously samples them. The FPGA
then adds headers and sends the data to the C&DH system which process is discussed in
22
Chapter 5. To compute the power spectral density, the EFAC channel also requires digital
signal processing which is also discussed in Section 3.8. The design and performance of
the analog instrumentation sections is discussed in the following order: the EFDC channel,
Floating Potential Probe (FPP) channel, and the EFAC channel.
3.2
Alternate Configurations
Alternate configurations exist for the Electric Field Instrumentation Set. In the past,
two floating point potential probes were used instead of one differencing channel were used.
This configuration would have increased the electric field power usage by 33%, and would
also have required one additional ADC channel. Another option to save power is to digitally
filter the signals, which was not done in this mission because of complexity and lack of FPGA
resources. Passive filtering saved power with the exception of the high-pass filter on the
EFAC channel, which will be discussed later. It should also be noted that volume was a
consideration. Increasing part counts can affect volume since there is a limited amount of
PCB space. A simple way to achieve a low-power, low-volume design is to eliminate active
circuit elements where possible.
E1
DC1 Channel
Low
Pass
Gain
Inst.
- Amplifier
Low
Pass
Gain
-
Low
Pass
Gain
Active
High
Pass
Gain
Analog Filters
Active Gain Stage
-
EV12
Inst.
+Amplifier
E2
FPP Channel
Analog to
Digital
Converter
E3
EV34
Inst.
+Amplifier
DC2 Channel
E4
Probes
Analog Front End
EFAC Channel
Fig. 3.1: Electric field system functional diagram.
To FPGA
+
23
3.3
Electric Field Probe Mechanical
The mechanical layout of the EFP consists of two boom sets with two probes each
mechanically perpendicular to each other labeled Electric Field Set 1 consisting of sensors
1 and 2 and Electric Field Set 2 consisting of sensors 3 and 4 as shown in Figure 3.2. The
Floating Point Potential Probe also shares sensor 1 and the other input is grounded.
The probes consist of gold-plated spheres with a 1cm diameter that are attached to
wires and are initially stowed upon launch on a spool. They are later deployed through
centripetal force and after deployment the booms measure 5m from the center of the spacecraft and a 10m tip to tip distance. More information on the boom deployment system can
be found in this project thesis [12].
3.4
Guarding and Shielding
The current generated by the plasma on a 1 cm sphere is at most 10pA. Proper shielding
and guarding techniques are used to prevent leakage. The 1pA input bias current of the
INA116 must be supplied. This can be done in two ways: by placing a voltage guard
EFProbe 1
ey
Sensor 1
Sensor 4
Sci Mag X
Sun Sensor
Sci Mag Y
Spacecraft Body
Coordinates
ez
ex
Spool
Sci Mag Z
Sensor 3
EFProbe 2
ADC Mag Y
ADC Mag Z
Sensor 2
ADC Mag X
Fig. 3.2: Electric field mechanical and sensor map.
24
around the wire with approximately the same voltage, and/or by using a very high-insulative
material [13]. Shielding the boom wire was impractical because a minimal wire diameter
was needed, so an insulative material was chosen. Teflon was used because it has one of the
highest resistivity’s of 1018 Ohms·cm. It is also durable and able to withstand the spacetime
environment. This was also difficult because a joining connection had to be made on the
top of the deckplate. The impedance issue was solved by insulating the solder joint with
teflon tubing. Another point for leakage is between the signal input and the science board.
The volume resistivity of the material in PCBs varies between 106 Ohms and 108 Ohms.
If the signal is unguarded, most of the current would leak out at this point. The INA116
instrumentation amplifier has guarding pins that near the voltage of the input pins. Because
of this leakage, capacitive effects are significantly reduced if proper shielding techniques are
used.
3.5
Expected Signal
Because the EFAC and EFDC channels have different input amplitudes across their
frequency range, it is necessary to ensure that saturation does not occur. Signal inputs
are driven by physical processes which change with atmospheric variability. A model was
created by the science team to determine the variability of the atmosphere and signal
inputs shown in Figure 3.3. The information for the electric field input model was taken
from several sources: data from previous missions, the fact that a spinning spacecraft will
generate an electric field, and bounding the noise by modeling it as 1/f or pink noise.
Spectrometer data from DE2 was used for the low bound case [14]. The VEFI instrument
from CNOFS was used for the upper bound case. The spin rate of the satellite can also affect
the electric field measurement because of the Lorentz force shown in equation (3.1) [10].
*
*
V × B +E0 = Einput
(3.1)
s/c
At the magnetic field strength of 0.5 Gauss and a tangential velocity of the probes
approximately 30m/s creates an electric field E of about 400mV/m, shown in Figure 3.3.
25
The ambient electric field contributes 200mV/m. To model the minimum 1/f noise, the
VEFI log chart was fitted with a line on its minimum frequencies which corresponds to
quiet time values. The maximum of each model was taken on each frequency bin. The
EFDC and EFAC channels are overlaid on the graph to show their respective frequency
ranges. In this manner, a good model for the input of both the EFDC and EFAC channels
was obtained.
3.6
Electric Field Direct Current Channels
The EFDC probe set measures large scale features in the ionospheres electric fields.
The maximum signal input was given by the science team as ±0.8V/m which comes from
both the spinning of the spacecraft and the ambient electric fields. For more information on
how electric field instruments interact with the plasma see “Design, Test, and Calibration
of the Utah State University Floating Potential Probe” [15]. The main preamplifier which
Amplitude Spectrum and E-field Instrument
1.E+06
DC Channels
V12
V34
1.E+05
Ch-2
Ch-3
Ch-1
Amplitude µV\m
1.E+04
Min Env
Max Env
Ch-4
1.E+03
1.E+02
1.E+01
1.E+00
Spectrometer
Gain =100
1.E-01
1.E-02
0.1
1
10
100
Frequency (Hz)
Fig. 3.3: Electric field system modeled inputs.
1000
26
is essentially a voltmeter is the INA116 instrumentation amplifier as shown in Figure 3.4.
It outputs the difference of the voltage of the probes which are connected to its inputs
provided it has sufficient input bias current. The preamplifier is then followed by two more
amplifiers which limit the system bandwidth at 40Hz with a low-pass filter and match the
range of the ADC. The low-pass filter also attenuates the signal with a gain of 0.312V/V to
bring the ±8V down to a ±2.5V range. The signal is then shifted by +2.5V to match the
ADC range of 0V to 5V. The dynamic range calculation is shown in equation (3.2), and is
calculated for the input of the instrument.
ADC Range
Klpf
= (5V )
1V
0.312 V
= 16 V = ±8 V
(3.2)
Lastly, a unity gain buffer isolates the load from the ADS8343 and the circuit. The
system sensitivity (equation (3.3)) is found by dividing the dynamic range by the total
number of counts.
Dynamic Range
ADC Counts
=
16 V
16
2 counts
=
244 uV
count
(3.3)
It is important that the noise in the system is lower than the system sensitivity noise
or the ADC will convert the noise to counts, and this will lead to a degradation of system
sensitivity by rendering it unobservable. System noise was calculated in a similar manner
as the ILP. The noise in Table 3.1 was found at the system input which makes it easier to
compare to the system sensitivity. Each count is 5V/65536 which is equal to 76.2uV/bit
Electric Field Direct
Current Channel
Low Pass Filter
and Attenuator
Cutoff Frequency
is 40Hz
Gain is 0.31
+
±8V
±8V
Gain is 1
Offset is 2.5V
-
±2.5V
+
INA116
+
-
AD8622
+1.25V
Opamp is in
unity gain
0V-5V
+
AD8622
Fig. 3.4: EFDC functional diagram.
0V-5V
16-bits
A/D Converter
ADS8343
27
the noise is 3764nV so the SNR is 21 or 13dB with additional noise reduction from the
coadding.
3.7
Floating Potential Probe
The FPP measures spacecraft charging which occurs when materials are exposed to
plasma in the spacetime environment such as the spacecraft body. The FPP measures the
potential of the spacecraft ground and body relative to the plasma and the function is
shown in Figure 3.5. The FPP is an aid to both the DC probe and the DC channels of the
spacecraft. It is identical to the DC channels with the exception that one of its inputs is
tied to ground. In this manner, the floating potential is the voltage measured between the
spacecraft body and the plasma. Because the instrument is electrically equivalent to the
EFDC channel, it has the same system sensitivity, dynamic range, and SNR. This reduces
the complexity of the system as a whole.
3.8
Electric Field Alternating Current Channel
The EFAC captures noise or higher frequency information that is associated with waves
or smaller scale features in the plasma and computes the spectral power. On previous mis-
Table 3.1: EFP system preformance.
Total Noise of EFP
Signal input Range of System
System Bandwidth
Noise from ADC ADS8343
Noise from ADC ADS8343
Noise from ADC ADS8343 Not Filtered
Total Noise from AD8622 Filtered
Total Noise from INA116
Unity Gain Amplifier AD8622
low-pass Filter Amplifier
INA116 Gain
Effective Noise at input 3rd amplifier stage AD8622
Effective Noise at input 2rd amplifier stage AD8622
Effective Noise at input 1st amplifier stage INA116
Effective Noise at input 1st amplifier stage INA116
Effective Noise at input 1st amplifier stage INA116
SNR
SNR
Value
16
40
20
20000
332
79
2365
1
0.3217
1
20002
62223
62223
87996
88.0
181824
52.6
Units
Volts
Hz
uV-rms
nV-rms
nV-rms
nV-rms
nV-rms
V/V
V/V
V/V
nV-rms
nV-rms
nV-rms
nV
uV
Unitless
dB
28
Floating Potential
Probe Channel
±4V
Low Pass Filter
and Attenuator
Cutoff Frequency
is 40Hz
Gain is 0.31
+
-
±2.5V
±8V
+
-1V
Gain is 1
Offset is 2.5V
+
-
INA116
AD8622
+1.25V
Opamp is in
unity gain
0V-5V
+
AD8622
0V-5V
16-bits
A/D Converter
ADS8343
The typical input values are shown in
blue, however the dynamic range is ±8V
Fig. 3.5: FPP functional diagram.
sions, this has been done with analog instrumentation by separating the signal in to several
channels with a bandpass filter or by computing an FFT and integrating the frequency
information into spectral bins [10]. The EFAC channel uses the same INA116 instrumentation amplifier and booms as the DC EV34 channel. This sharing contributes to power
and volume reduction. DICE uses an FFT to compute four power spectral density channels in the frequency range of 16Hz to 512Hz, the digital computation will be discussed in
Chapter 5. A high-gain of 100 is necessary for the EFAC channel, a lower frequency signals
with such a gain 100 would lead to saturation of the ADC input and must be filtered out.
Spacecraft spinning up to 2Hz creates an artificial sine wave which can be seen as a spike
of 400uV/m in Figure 3.6. To filter out the large spin signal, a high-pass filter was used. A
butterworth filter was selected because they are known for their flat passband response [16].
It was necessary to try to place the poles as close to 16Hz as possible while maintaining a
flat passband across the range of the FFT and having minimal phase shift. The slope of
the filter is directly related to the SNR of the AC channel. To filter out the spin component
a 60dB/Dec is necessary, a 7-pole butterworth was required. The signal dynamic range is
±50uV which needs a gain of 100 and offset of 2.5V to meet the range of the ADC which is
0 to 5V as shown in Figure 3.7. The signal is then sent to the FPGA where it is separated
into different frequencies and summed up into four spectral bins.
Noise is also an issue with the EFAC system if not eliminated the noise is summed
into the four frequency bins along with the signal. Because of the difficulty of the noise
29
ADC Input Min Envelope Amplitude Spectrum
1.E+07
0th Order
1.E+06
5th Order
1.E+05
6th Order
Amplitude (uV/m)
1.E+04
7th Order
1.E+02
1.E+01
1.E+00
1.E-02
1.E-03
1.E-04
1.E-05
0.1
1
10
100
Frequency (Hz)
Fig. 3.6: EFAC filtering.
Electric Field Alternating
High Pass Filter
Current Channel
Cutoff Frequency @15Hz
Cutoff Frequency @15Hz
G = 100
Offset = 2.5V
This Circuit Represents Three
Simmilar Poles of the Circuit
+
±50mV
RMS
±50mV AC
-
±5mV
-
-
+
+
INA116
+
-
x3
+1.25V
0V-5V
16-bits
A/D Converter
ADS8343
Fig. 3.7: EFAC functional diagram.
calculations, only a spice noise analysis is needed to understand how the noise affects the
signal. The noise was calculated at the input of the ADC; additional input referred noise
was simulated by using a 4MOhm resistor. The instrument’s frequency range is from 16Hz
to 1024Hz. Figure 3.8 shows the noise at the input of the ADC (green line and small
amplitude signal). The noise amplitude of the EFAC is roughly 2000uV which is many
times over the bit range of the ADC which is 76.2uV/count. The red line (large amplitude
signal) shows the noise at the input of the last gain stage and also suggests that the main
noise contributor is from the gain stage in the EFAC system.
30
Voltage Noise at ADC
Voltage Noise Before Last Gain Stage
500µV/√(Hz)
450µV/√(Hz)
400µV/√(Hz)
Noise
350µV/√(Hz)
300µV/√(Hz)
250µV/√(Hz)
200µV/√(Hz)
150µV/√(Hz)
100µV/√(Hz)
50µV/√(Hz)
0µV/√(Hz)
1Hz
10Hz
100Hz
1KHz
Frequency
Fig. 3.8: EFAC spice noise simulation.
10KHz
100KHz
31
Chapter 4
Three Axis Magnetometer
4.1
Magnetometer Concept
Magnetic field measurements are secondary science objectives for the DICE mission.
The science magnetometer implemented was an experimental small volume, low-power sensor. The science team felt that there was great value in demonstrating that a magnetometer
could be included along with the other instruments. The sensing technique selected was
based on the Anisotropic Magneto Resistive (AMR) effect which can be fabricated into
extremely small packages. Honeywell has a line of AMR bridging sensors and integrated
magnetometers that are small in volume, relatively low-power, and sensitivity down to the
nT range. These sensors have been flown on other cube sat missions such as Radio Aurora Explorer [17, 18]. The HMC5363 magnetometer was selected and integrated as the
attitude determination magnetometer for DICE. This is a 12-bit magnetometer providing
sensitivities down to 700nT. The desire for the DICE science magnetometer was to create
a device with 1nT resolution and matching sensitivity. None of the integrated Honeywell
devices approached this level of performance with the closest being the HMC2003 which is
an analog sensor incorporating the most sensitive magneto resistive sensors, the HMC1001
and HMC1002, into a hybrid module as shown in Table 4.1. The concept was to construct a
similar magnetometer using a highly-integrated, precision, 24-bit Analog-to-Digital convert
circuit. The entire TAM would be incorporated into the available space on the science board
of DICE along with the other instruments. The sensor would be deployed externally on
a boom to remove it from the immediate magnetic contamination of the DICE Spacecraft
and was connected via cable to the science board. Because the magnetometer was of value
for secondary science and as a backup attitude magnetometer a best effort was made to
32
achieve the highest sensitivity and lowest noise operation. A summary of the M1 science
requirement (see Appendix A.1) states that the magnetometer shall return 1nT with a SNR
of ≤ 3 over the ±50,000nT range of the Earth’s magnetic field. This requirement formed
the basis of which the TAM was designed around.
4.2
Magnetometer Mechanical Layout
The DICE magnetometer needed to be situated as far away from the spacecraft as
reasonably possible to reduce noise on the sensor from the spacecraft. The magnetometer
was mechanically incorporated into the ILP boom. This, however, created a risk that the
boom might not deploy properly. Different options were provided in the event that the
mechanical system could not be proven. The DICE magnetometer has two options, one
internal on the science board itself and an external option to reduce the effects of electrical
noise from other electronic processes such as the EPS board. Jumpers select between
the internal and external options. A temperature sensor is adjacent to the magnetometer
packages on both options. The internal option shown in Figure 4.1 mounts the HMC1001
and HMC1002 directly to the science board near the middle of the board and occupies
approximately 4cm2 of space. The internal option was intended mostly for prototyping,
and if the measured environment was not too noisy, then it could be used. The external
option was to be used in the event that magnetic noise from other processes reduced the
sensitivity of the magnetometer. The sensors are mounted on a PCB board which is small
in volume being only (1.6 x 1.8 x 0.5cm). There is also an ADS590kf temperature sensor
Table 4.1: AMR bridge parameters.
Parameter
Axis
Board Orientation
Sensitivity
Range
Temperature Range
Noise Density @ 5V
Temperature Stability
Package
Size
HMC1001
One
Perpendicular
3.2
±6
-55 to 150
48
0.3
20-Pin SOIC
13x10x2.6
HMC1002
Two
Parallel
3.2
±6
-55 to 150
48
0.3
8-Pin SIP
1.6x11x7
Units
N/A
N/A
mV/V/Gauss
Gauss
◦C
nV/Hz
◦C
N/A
mm
33
Digital
FPGA
Analog
Digital
Electric Field
Probe
Analog
Ion
Langmuir
Probe
Magnetometer
HMC1002
HMC1001
EExternal Magnetometer Conn.
Fig. 4.1: Science payload and instrumentation.
mounted to calibrate out temperature effects on orbit. A delrin enclosure encases the PCB
board and the parts to help shield them from the environment. In the end the external
option was not placed on the boom due to mechanical concerns with deploying the Langmuir
probes and the bulky 24-wire harness that attached to the external magnetometer science
board. The magnetometer was then attached to the –Z plate which is situated near the
base of the –Z DC Probe boom (shown in Figure 4.2) and located -3.34cm in the X-axial
direction, -1.26cm in the Y-axial direction, and -8.14cm in the Z-axial direction from the
center in the spacecraft body coordinate system. The spacecraft (black) and magnetometer
(blue) coordinate systems are also shown in Figure 4.2.
4.3
Magnetometer Implementation
There are three design issues associated with the magnetometer. Magnetoresistive
sensors have sensitivity reduction that decreases with time or exposure to high magnetic
fields. The material can cause offsets that can decrease or change the dynamic range of the
34
X&Y Axes (HMC1002)
Z+
Y+
Z-Axis (HMC1001)
Temp Sensor
PCB Board
X+
e-z
External (Boom)
Configuraon
Spacecraft Body
Coordinates
S i Ma
Sci
M
ag +Z
Mag
ex
External (flight)
Configuraon
ey
Fig. 4.2: TAM mechanical and sensor map.
instrument which produces measurement error. Lastly, a large potential temperature range
from the thermal environment causes bridge voltage offsets.
AMR technology is sensitive to high magnetic fields. Magnetic field domains line up
incorrectly and oppose the field when exposed to high magnetic field. To reverse this effect,
an S/R strap is provided to realign the domains; a pulse greater than 4V will consistently
reset the device to its ground state. Figure 4.3 shows a consistent degauss pulse that requires
a 4V pulse [19, 20]. The DICE magnetometer had all of the S/R resistors in series which
created a problem because the voltage was divided across each 1.5Ohm resistor. A 12V
signal was needed to generate the required 4V per resistor and 3A per circuit and requires
a retooling of the degauss circuit on the board. A better approach would be to put the
S/R resistors in parallel since it was a challenge to generate a 12V pulse for 20uSec with
the series resistors. The large amount of current also necessitates proper cabling to ensure
35
a low-resistance pathway. The S/R pulse is activated by a command that can be sent from
the ground station to reset the magnetometer on orbit.
Offsets are a common problem with AMR sensors. Manufacturing processes cause
resistor mismatches which lead to a constant voltage imbalance [19]. This can lead to an
offset up to 15mV as shown in Figure 4.4.
The DICE magnetic sensor has an output of ±8mV, with an offset up to 15mV; it
could offset the dynamic range by almost half of its value. This would also cause an offset
in the range of magnetic sensitivity resulting in mostly the positive or negative side of the
values to be detected or railing of the signal. There are two methods used to correct offset
problems, offset straps and external parallel resistors. The offset straps are resistors that
create a magnetic field but also require 2mW of power. A lower-power option is using
resistors in parallel with the larger resistor value of the bridge to match them. For example:
if the bridge resistors are 800Ohms and 810Ohms, this will cause an offset of 12mV. If a
65KOhm resistor is placed in parallel with the 810Ohm resistor, it will reduce the offset in
the micro-Volt range and use 0.15mW of power. DICE had the option to use the parallel
resistor method. Due to project time constraints, this method was not used, instead the
Fig. 4.3: Set/reset voltage for AMR bridge.
36
HMC 1001 Sensor Performance
Magne!c Field nT
-150000 -100000 -50000
0
50000
100000 150000
40.0
20.0
10.0
0.00
Output Voltage mV
30.0
-10.0
-20.0
Bridge Output
Bridge Output + Offset
Fig. 4.4: Magnetometer offset.
dynamic range was increased by a factor of 3.8 to mitigate the offset problem at a cost of
sensitivity.
Another challenge associated with AMR bridges is changes in resistance caused by
temperature. This is caused by resistance changes in the material of the bridges and can
vary from device to device. The HMC1001 and HMC1002 have a change in resistance of
0.25%/◦ C. Temperature changes also cause bridge voltage offsets to vary by ±0.05%/◦ C.
Temperature effects can be removed from the final measurement if the devices are calibrated
across a temperature range and if the operating temperature is known by sensing it. The
temperature sensor chosen for DICE had to be able to monitor temperature both on the
science board itself and the external magnetometer module. The AD590 temperature sensor
used for temperature compensation offers electrical and mechanical advantages as well as a
temperature range of -55◦ C to 150◦ C. The AD590 has a linear current output of 1uA/◦ K
and does not require amplifiers or support circuitry while maintaining a ±0.3◦ C linearity
across its temperature range. The AD590 also has three different package options which
made it ideal for the internal and external options of the magnetometer. The standard
37
TO-52, or can-package, was used on the science board itself while the two-lead CQFP was
small and flat enough to be used for the external magnetometer module.
4.4
Electrical Design of Magnetometer
The electrical design of the magnetometer needed to be low-power and meet the re-
quirement of 1nT sensitivity. To meet the requirement of 1nT with a range of ±50,000nT an
ADC with at least 18-bits was needed. To make a better measurement an ADC with greater
precision was desired. An ADS1248 was selected because it was a low-power, precision, lownoise, and 24-bit ADC. It was also built specifically to differentially measure wheatstone
bridges such as the one found in the magnetometer. The ADS1248 is a Delta-Sigma ADC
and also has other features including a digital filtering and Programmable Gain Amplifier
(PGA). The digital filtering is dependent upon the sampling rate of the ADC and usually
has a -3dB cutoff at half the sampling rate. The PGA operates in powers of two from 1
up to 128. Another analog option was provided in the circuit to differentially measure the
bridge with the use of an instrumentation amplifier so the gain could be adjusted linearly
instead of powers of 2. The INA129 is low-noise precision instrumentation amplifier that
was selected for this option and also because of its dual use in the ILP design. Jumpers were
also provided to select between the selectable gain of the instrumentation amplifier option
or between using the PGA gain and a differential measurement in the ADC. The expected
magnetic field signal input range supplied by the science team is ±50,000nT which is the
planned dynamic range of the instrument. The AMR wheatstone bridge, shown in Figure
4.5, is simplified by viewing its differential output Vbridge , which is equal to the positive
input terminal subtracted by the negative input terminal. The bridge has a sensitivity of
16mV/105 nT. When multiplying this by the signal input of ±50,000nT, the bridge differential output is ±8mV. Measuring the bridge differential voltage is done by subtraction by
the instrumentation amplifier which also gains the signal by 20.5 for an output value of
±164mV. Shifting by an offset of 2.5V gives a value of ±164mV+2.5V at the output of
the instrumentation amplifier. The signal is then digitized by the ADS1248 which also has
the ability to gain, offset, and digitally filter the signal. By setting the gain to 4 with no
38
offset, the dynamic range of the ADC to narrows to ±625mV. The signal is not truncated
by the ADC to a 24-bit and until the offset and gain are implemented. The ADC also has
a +2.5V reference which centers the digital range around that voltage giving an effective
range of ±625mV (1.250V total) and 224 -bits. A 6-bit truncation is then necessary, and the
resultant count range of the system is 218 . An oversight (or lack of attention) to the system
filtering resulted an incorrect system bandwidth. It was later realized the ADC filtering
was fixed to its sampling rate, and the sampling rate had to equal that of 3 times 70Hz
to sample all three channels. The next highest sampling rate is 320Hz, which also had an
undesirable digital filter with -3dB cutoff point at 152Hz.
4.5
Magnetometer Sensitivity and Dynamic Range
Noise limits the sensitivity in electronic systems. The noise level must be less than
the desired signal sensitivity. The magnetometer system was required to have a sensitivity
of 1.5nT. The circuit needed the appropriate gains and filtering to reduce system level
noise and also to maintain the signal at the input voltage range of the ADC. The greatest
determining parameter of gain the system sensitivity is the AMR bridge sensitivity. The
bridge sensitivity is dependent on voltage which increases power demands. The bridge has a
sensitivity of 3.2mV/V/Gauss. To find the sensitivity, the bridge voltage can be multiplied
by the bridge sensitivity to get the system sensitivity in equation (4.1).
Sensitivity
16mV/Gauss @ 5V
B = ±0.5Gauss
+5V +
-
Instrumentation
Amplifier
Gain is 20.5
Offset is +2.5V
+
R
R
ΔV=±8mV R g
ΔV=±160mV+2.5V
R
INA129
Set\Reset strap
Resistor
Degauss
Pulse
Circuitry
ΔV=±640mV+2.5V
+
24-bits
- (18-bits effective )
Ref
R
ADC Range is
Internal PGA has gain of 4
+
-
+2.5V
A/D Converter
ADS1248
Fig. 4.5: Magnetometer functional diagram.
Serial
To FPGA
Magnetometer
39
Bridge
!
Bridge
!
=
5V
Senstivity
V oltage
3.2mV /V
1 Gauss
=
16mV
16mV
= 5
1 Gauss
10 nT
(4.1)
Power is saved at the cost of sensitivity; a lower bridge voltage will yield a lower bridge
sensitivity. The total system sensitivity was found at the input of the bridge. To achieve a
1nT sensitivity, the system noise at the bridge input needed to be less than 160nV. To find
the total system sensitivity equation, the bridge sensitivity was multiplied by the inverse of
the gains and the ADC voltage and counts shown in equation (4.2).
105 nT
0.016 V
1V
20.5 V
1.250 V
218 counts
=
1.454 nT
1 count
(4.2)
The total system noise was also calculated at the bridge output. Three devices are
the main noise contributors in the magnetometer, the resistive bridge, the INA129, and the
ADS1248. Both the INA129 and the resistive bridge have 1/f noise as well as white noise.
Both noise sources need to be modeled accurately for the total system noise; nV can make
a difference in the SNR final calculation. A method to easily find the contribution from 1/f
noise from a chart is to find the amplitude at the 1Hz crossing mark V1Hz
Amplitude ,
then
find the corner frequency FCorner , and the desired frequency of the lower limit of integration
FInt−Limit . The result can be computed in equation (4.3).
V 1 (rms) = (V1Hz
f
Amplitude )
p
log (FCorner /FInt−Limit )
(4.3)
Effects from the filter were ignored since both corner frequencies were well below the
filter cutoff of 150Hz which in this case is the system bandwidth. The white noise source
contribution is found by finding the white noise amplitude (measured in Root-Mean Squared
(RMS) volts) in equation (4.4).
V whitenoise (RM S) = (V noiseamplitude ) ∗ SQRT (SystemBandwidth)
(4.4)
40
The last noise source in the system is the ADS1248, which is also a Gaussian white
noise source as shown by the distribution. However, the datasheet only lists the total noise
in Volts-rms at the output. The PGA gains are set at four while the noise at the input of
the ADC1248 is a factor of four less. The total system SNR is found at the output of the
bridge; this makes accounting the total gain and dynamic range less difficult. For example:
the INA129, with a gain of 20.5, multiplies the noise from the output of the bridge by 20.5.
Similarly, the equivalent noise contribution from the ADC can be found at the input of
the bridge by dividing it by the INA129s gain of 20.5. To find the total system noise, the
contribution of the each of the Gaussian sources from the bridge, INA129, and ADS1248,
is divided by 20.5 plus the 1/f noise sources. The SNR is calculated at the sensor. The
smallest signal is 160nV and the noise spectral power summed is 209nV; dividing the signal
by the noise yields a SNR of 0.762 or slightly over 1nT for the system sensitivity. A excel
chart listing these parameters and other system parameters is found in Table 4.6.
Fig. 4.6: TAM system parameters.
41
Chapter 5
Payload Digital System
5.1
Data Subsystem Overview
The data collection system consists of all the hardware that processes data including
the payload Data Acquisition Subsystem (DAS), C&DH system, the radio, and ground
station as shown in Figure 5.1. The DAS collects the data, processes the data, and returns
the data to the ground. The DICE payload digital subsystem needed to be low-power but
also allow for the control of the payload board and the ability to process and return data to
the C&DH system. The data can then be processed and stored in the radio which sends the
data via downlink to the ground station. First, the data is sampled by the ADCs which are
controlled by the FPGA through multiple serial data links. The data is then time-stamped
by the FPGAs RTC. The RTC is a precision clock synchronized with GPS time, to achieve
time accuracy close to the GPS time system. Each instruments’ data is separated into a
“granule” of data. The granules are collected in a First-In-First-Out (FIFO) buffer which
continually sends packets to the C&DH system when possible. Processing of the granules
is done by the C&DH system which rearranges and condenses the data into radio packets
and sends them to a queue in the radio. When the ground station contacts the satellite,
the data is down-linked and copied to a database that stores Level 0 (or raw, uncorrected)
data.
Additional tasks are also carried out by the DICE digital subsystem, such as the ability
to control the sweeping voltage for the ILP with a DAC. The digital subsystem also had
the option of being augmented by a Digital Signal Processor (DSP). At design time the
method to compute the FFT was unknown, and three options were provided. Other tasks
of the digital subsystem will not be discussed, such as the control of the electric field
42
boom deployment subsystem. Power modes of the instrumentation are also controlled via
command and can switch the instrumentation on and off but the system will not be discussed
here.
5.2
Digital Requirements
A summary of the most important requirements that relate to the data collection
system are listed here. Without meeting these requirements, the data collection system
fails to produce results that are acceptable to the mission. The requirements fall into two
different categories: sampling requirements and the separation of the digital and analog
subsections.
Electro-Magnetic Interference (EMI) is most prevalent in digital systems to prevent this
noise from affecting the analog instrumentation. The boards were separated into different
sections for each. To meet requirement SI4 and also to mitigate noise power, planes, and
ground planes were separated into respective digital or analog areas. Digital lines were not
allowed to be routed through analog areas.
The DAS was required to be able to reliably sample data and time-stamp data. The
sampling rate was determined by the science team. Requirement SI2 states, “The maximum
sampling rate for any science data channel shall be 100Hz.” Requirement S1 further states,
GPS
Command and
Data Handling
PPS
Signal
Serial
(I2C, SPI)
UART
(RS232)
Analog to
Digital
Converters
DSP’s
Cadet
Radio
Ground
Station
Actel Igloo
FPGA
Power
Parallel
DCP Sweep
D/A Converter
Fig. 5.1: DICE data collection system.
Analog
Instrumentation
UART
43
“Measure RMS Fluctuations in Electric Field and Plasma Density: S1. Make co-located
DC electric field and plasma density measurements at a 10km on-orbit resolution.” The
science team also determined that the sampling rate should be more than 0.7Hz to meet
spatial sampling requirements. The telemetry budget was not limited by rates up to 70Hz
and a higher telemetry rate was chosen. Another lower rate of 35Hz was added in the case
that 70Hz consumed too much power or telemetry resources. Requirement SI2 stated that
the “maximum sampling for any channel needed to be 100Hz.” This meant a 10ms time
stamping requirement. However, because the RTC resolution was 1ms, the requirement was
exceeded by a factor of 10. Time-stamping is discussed in Section 5.3.8 of this document.
Requirement E15 states that the “EFAC spectrometer shall have at least three points of
frequency data from 16Hz to 512Hz.” Another channel was added to be consistent with
previous missions [21]. The FFT method was chosen to compute the spectral power.
5.3
Payload Controller and Data Acquisition Subsystem
The several goals of the DAS are: data must be sampled reliably, a method to accept
commands from the C&DH subsystem must be provided, and the subsystem must control
all functions of the science board. The payload digital system consists of all of the digital
hardware on the science board including the FPGA, ADCs, and DAC. These are discussed
in depth in their respective sections.
5.3.1
Field-Programmable Gate Array
The design of the DAS needed a low-power controller to poll the ADCs and control
the payload systems. The ability to do things simultaneously is an advantage of an FPGA
over a microprocessor. The linear nature of microprocessors necessitates the need for much
software design. To ensure the device met real-time constraints at design time, an FPGA
was chosen over this option. An Actel Igloo FPGA series was selected for a controller for
properties of low-power, precise control, and previous use in cubesat designs at SDL. This
limited risk and shortened design time by having a previously implemented hardware design
and VHDL code reuse. Actel FPGAs have advantages and disadvantages. One advantage
44
is that they are very low power operating in the tens of mW range and have a low operating
voltage of 1.2V. They require no boot up time and are ready the instant power is applied
(or a few ns later). Actel Igloo FPGAs are disadvantaged by the fact that they have no
built in DSP hardware such as adders or multipliers, functions have to be implemented on
a gate level and take more FPGA resources. Because of this, less digital filtering can be
utilized by design. Multipliers and other complex digital circuitry have to be avoided. Actel
FPGAs also have memory resources as shown in Table 5.1. An AGL600k gate part was
selected initially with the option of moving to a 1000k gate part since both parts were pin
compatible. Each of these devices are available in a 256 Pin BGA. Both Parts also have a
small footprint of 1.7cm by 1.7cm and have 177 in/out pins. These pins can operate at a
standard digital switching voltage of 3.3V or lower voltages.
5.3.2
FPGA Firmware
The FPGA firmware has many different functions as shown in Figure 5.2. The system
has to be able to receive and respond to commands from the C&DH. The FPGA also
interfaces with external devices. Firmware design is broken down into these separate tasks.
The C&DH UART interface turns incoming serial data into parallel data to the Payload
Controller which then decodes the data and sends out appropriate commands to other
functions in the system. A timekeeper has a free-running clock that counts milliseconds
from the FPGA 12mHz system clock. Time is synchronized from the GPS by a register
which can be updated from an external source. For time-stamping, the GPS time is also
received by the DAS. Sampling and processing is done solely by the DAS which is the
interface to the ADCs. Data is sent to the C&DH system through a UART. The DAS is
discussed in depth in Section 5.1 of this chapter. The DAS can also relay other important
Table 5.1: FPGA parameters.
Parameter
System Gates
RAM
Flash-ROM
Operating Voltage
AGL600
600k
108k
1024
1.2, 1.5
AGL1000
1000k
144k
1024
1.2, 1.5
Units
gates
bits
bits
V
45
information to the C&DH system from the deployment controller such as motor and encoder
data recorded by the boom deployment system. Lastly, there are some auxiliary circuits
such as the degauss pulse for the TAM instrumentation and the ability to switch the analog
power from the instrumentation on and off to conserve power in the event of a low battery
event. These two systems will not be discussed in this thesis.
5.3.3
Payload Controller Commands
The commands received through the UART are processed by the Payload Controller.
Most missions support many different telemetry modes, or the rate in which data is collected.
In the case of DICE, a simplistic approach was used with only four planned modes in the
Analog Data In/Out
Magnetometer
X,Y,Z
ADS8343
DAC7621
ADS8343
ADS1248
Parallel Data
Motor and
Encoder
SPI Bμs
Electric Field
DC, AC
Serial Data
ILP Sweep
Signal
Serial Data
ILP Probe
FPP
GPS
Milliseconds
Commands
DAS Data Out
Payload Controller
Encoder and
Motor Position
TX FPGA to C&DH
TX C&DH to FPGA
Deployment
Controller
Degaμss
Pulse
Analog
Power
C&DH SYSTEM
Fig. 5.2: FPGA firmware.
External
Circuits
C&DH UART Interface
/w 512 Byte FIFO
Time
Keeper
Encoder and Motor Commands and Data
GPS
Milliseconds
Encoder and Motor Position Data
DAS
Commands
DAS Data In
Data Acquisition System (DAS) Controller
46
beginning. These commands, as shown in Figure 5.3, originally were to update the time and
set the DAS sampling frequency. The DAS frequency command also controls the analog
power. Power is enabled by sending a 35Hz or 70Hz command. The complexity of the
commands grew as commands were added later as needed such as the ability to control
the motor and encoder power, control the motor position, degauss the magnetometer, and
set the sweep mode for and ILP sweeping voltage. All commands are one byte sent with
Most Significant Bit (MSB) first; two commands have additional bits. The 0x41 and 0x54
commands have two additional bytes and one additional byte, respectively, that are sent
after the command to set registers for their particular function.
5.3.4
Analog-to-Digital Converters
The ADC selection process was important because the quantization error can be the
main factor in determining system performance. The ADCs need to meet the resolution
requirement, but minimize use of system resources, especially power. The ADC selection
process is difficult due to the amount of parameters involved including: number of bits,
Fig. 5.3: Payload system commands.
47
sample speed, sampling scheme, power, number of channels, and digital interface. The instrumentation required different sampling rates, the EFDC, ILP, and Magnetometer needed
a low rate of 70Hz, but the EFAC needed a rate over 1024Hz. A method to sample multiple ADC channels while making a lower rate conversion is to oversample and interleave.
The magnetometer, however, also needed 18-bit precision. Measurements made by the ILP,
EFDC, and EFAC channels are 16-bit. There are two main types of low-power ADCs to
choose from—both have advantages and disadvantages. The two main types of ADCs are:
Successive Approximation Register (SAR) and Delta-Sigma converters. SAR converters
typically have higher-power, respectively, but have a higher sampling rate. Delta-Sigma
converters tend to filter out high freqencies and take longer to convert, but have better
accuracy and have built in noise reduction. Oversampling and interleaving also saves power
by reducing the number of ADCs that are used by combining channels into the same ADC.
This method works best on SAR ADCs because of their fast conversion times. A Texas
Instruments ADS8343 16-bit ADC was selected for low-power properties because it was a
SAR. The ADS8343 can sample at 100kHz which is well beyond the highest needed 1024Hz
sampling rate of the system. These ADS8343 properties helped fulfill requirements of the
EFP and ILP instruments. A differential reference scheme is used, meaning that the output
in counts is in a two’s complement binary format. The reference was set 2.5V to center
the ADCs zero-point around the midpoint ADCs counting system and to match the full
scale of the 5V analog signal. For the magnetometer, the ADS1248 Delta-Sigma ADC was
used, because it supported the required 18-bit precision of the TAM instrument. Sampling
of this ADC has a 2kHz limit. The multiplexer on the front end can be configured as two
differential pair inputs or seven single-ended inputs. The TAM ADC also has a reference of
2.5V and a 5V input with a 24-bit also in a two’s complement binary format; however, the
values are later truncated to 18-bits.
5.3.5
Sampling Methods
The sampling methods for the DAS are different depending on the sampling scheme
and the ADC. The interleaved sample conversion allows for simultaneous spatial sampling
48
in a 70Hz window by the instrumentation. Differences in the timing and the amount of
co-adding vary by instrument. The ADS8343 both make many 256 conversions and then
sleep for a time to save on power. Interleaving is still preformed by the ADS1248 channels,
but the limited sampling rate prohibits co-adding. The conversion system, shown in Figure
5.4, is driven by a 70Hz clock which marks the start of each sample. Each instrumentation
channel is sampled at different rates, but aligns on the same 70Hz boundary. Channels that
consist of the ILP Z+, ILP Z-, and FPP form a set from the same ADS8343 ADC, which
also performs conversions of these channels in that order. Each instrumentation channel is
oversampled and co-added 256 times to reduce noise. Once 256 conversions have taken place
on all three channels, a sample is completed, and the ADC enters a sleep mode until the next
70Hz clock. The electric field ADS8343 consists of the EF12, EF34, and EFAC channels.
Sampling of the first two channels is done in the same manner as the ILP channels with
256 conversions per sample; however, the EFAC co-adds every eight conversions to increase
its sampling frequency. The electric field sample has a slower conversion interval which is
14.5us for the ILP versus 18us per conversion of the electric field. Three magnetometer
channels from the ADS1248 ADC rely on a completely different scheme because of the
nature of Delta-Sigma ADCs. The samples start on the same 70Hz boundary as the other
channels. Because samples cannot be interleaved, they are preformed sequentially on each
axis in this order X, Y, and then Z. Sampling by the ADS1248 ADC which takes 3.125ms
at 320Hz to complete. This TAM sample set takes 9.375ms to complete. The ADC then
sleeps until another 70Hz clock is detected. Each 70Hz measurement from the EFDC, ILP,
FPP, and Magnetometer are grouped into granules, and the values are stored as explained
in the Data Flow Down (Section 5.4).
5.3.6
EFAC Spectrometer FFT
At design time, it was not known how the FFT would be computed; different options
were designed into the hardware for testing purposes. Initially, the prototype had three
options which were available for the FFT: two digital co-processors and a firmware FFT.
The micro floating point unit (uM-FPU) by Micromega is a floating point emulated mi-
49
Conversion Set #1
Ion Lang. Probe Conversion #1
ILP Z+ (Ch0)
ADS8343
Conversion #2
ILP Z- (Ch1)
Conversion Set #2
Conversion #3
FPP (Ch2)
Conversion #4
ILP Z+ (Ch0)
14.5 μS
ADS1248
Samples
(Magnotometer )
Conversion #6
FPP (Ch2)
Conv. #766
ILP Z+ (Ch0)
Conversion #767 Conversion #768
ILP Z- (Ch1)
FPP (Ch2)
43.5 μS
43.5 μS
256 Conversions in a
Sample for each channel
11.136 ms
ILP
Samples
Conversion #5
ILP Z- (Ch1)
Conversion Set #256
Sample #1
ILP Z-, ILP Z+, FPP
Sample #2
ILP Z-, ILP Z+, FPP
Sleep Mode
Sample #3
ILP Z-, ILP Z+, FPP
Sleep Mode
Sleep Mode
14.285 ms or 70Hz
Conversion
Mag X
Conversion
Mag Y
Conversion
Mag Z
Sleep
Mode
Conversion
Mag X
Conversion
Mag Y
Conversion
Mag Z
Sleep
Mode
Conversion
Mag X
Conversion
Mag Y
Conversion
Mag Z
Sleep
Mode
3.125 ms
9.375 ms
Conversion Set #1
Electric Field
ADS8343
Conversion #1
EF 12 (Ch0)
Conversion #2
EF 34 (Ch1)
Conversion Set #2
Conversion #3
EFAC (Ch2)
Conversion #4
EF 12 (Ch0)
18 μS
Conversion #6
EFAC (Ch2)
Conversion #766 Conversion #767 Conversion #768
EF 12 (Ch0)
EF 34 (Ch1)
EFAC (Ch2)
54 μS
Sleep
Mode
54 μS
Sleep
Mode
256 Conversions in a Sample
(for EF12 and EF32 only)
13.824 ms
Electric Field
Samples
Conversion #5
EF 34 (Ch1)
Conversion Set #256
Sample #1
EF12 and EF34
Sample #1
EF12 and EF34
Sleep
Mode
Sample #3
EF12 and EF34
14.285 ms or 70Hz
EFAC
Conversions
ADS8343
2 Conversions Conversion #3 2 Conversions Conversion #6 2 Conversions Conversion #9
(EF12,EF34)
EFAC (Ch2)
(EF12,EF34)
EFAC (Ch2)
(EF12,EF34)
EFAC (Ch2)
54 μS
2 Conversions
(EF12,EF34)
Conv. #21
EFAC (Ch2)
2 Conversions
(EF12,EF34)
Conv. #24
EFAC
54 μS
8 Conversions in a
Sample
EFAC Sampling
Sample #1
Sample #2
Sample #3
Sample #31
Sample #32
Sleep
Mode
432 μS
14.285 ms or 70Hz
Sleep
Mode
Sleep
Mode
Sleep
Mode
13.824 ms
FFT Group
32 Samples
32 Samples
(64 total)
32 Samples
(992 total)
32 Samples
(1024 total)
14.285 ms
32 Sample Sets for 1024 total samples
Fig. 5.4: DAS sampling scheme.
croprocessor that was also used on the ACDS board of the satellite for processing floating
point values and also floating point other mathematical functions. The um-FPU could only
compute a 64-point FFT and consumed 30mW (60mW at 5V) by power but was included
because of knowledge of its functionality. The second co-processor was a C5000 series chip
built by Texas Instruments which had a built in hardware capable of performing a 1024point FFT. The C5500 had a low-power usage 30mW and even lower if duty cycled, but
was expensive in time and complexity due to needed software and hardware development.
A firmware FFT from opencores.org, comprised the last option, but was not available at
design time. Later in the project, it was available for use and was chosen as the final method
to compute the FFT. Utilization of the FPGA by the firmware FFT is approximately 30-
50
40% of the total gates, and the larger 1000k gate device was needed to accommodate the
firmware. Values from the EFAC channel are oversampled at 17920Hz by the firmware from
the ADS8343. The data is then co-added by a factor of eight, each sample is loaded into a
FIFO as shown in Figure 5.5. Once 1024 values are stored, the FFT computes and returns
the data one sample out at a time which equates to 1Hz amplitude of information. As each
ADS8343
Analog to Digital Converter
ADS8343
Analog to Digital Converter
Magnetometer Z-Axis
Magnetometer Y-Axis
Magnetometer
Channels Sampled
every 70Hz x3 or
~210Hz
Magnetometer X-Axis
Electric Field FPP Ch.
DC Probe Ch. 2
Each of the
ADC8343
Channels are
sampled at
16xxxHz
DC Probe Ch. 1
Electric Field AC Ch.
Electric Field DC Ch. 2
Electric Field DC Ch. 1
sample is computed, it is summed together in each of the frequency bins shown in Table
ADS8321
Analog to Digital Converter
Switch: DCPS Mode
(Only EFAC Ch.)
Co-adder (x256)
Buffer
(512 samples
each channel)
Co-adder (x8)
Buffer
(1024 samples)
FFT
(1024pt)
Granule
Sampled at:
35Hz mode 1
70Hz mode 2
Science Granule
Granule
Sampled every
2Hz
Spectral Binning
Bin1: 16-32Hz
Bin2: 33-64Hz
Bin3: 65-128Hz
Bin4: 128-512Hz
EFAC Granule
Granule
Sampled every
2 minutes
Science FIFO
EFAC FIFO
Traffic
Controller
FPGA UART
RS232
Fig. 5.5: DAS functional data plot.
Sweep Granule
Sweep FIFO
Analog
Digital
51
5.2. Each of the four values then form a granule and are released to the C&DH system at
1Hz.
5.3.7
ILP Sweep Mode
The ILP is required to sweep its voltage; this voltage sweep occurs from -4V to 2V
every 120 seconds and is digitally controlled with a DAC and interrupts the normal mode
0 sampling scheme of the ILP and FPP instruments. The main difference between mode
1 and mode 0 is co-adding is the reduction from 256 to 32 times per sample. A reduction
of timing from 3.712ms to 464us per sample enables the sweep to complete faster. Both
ILP channels and the FPP channel are interleaved to produce a sample every 1.5ms. The
DAC also changes voltage at the beginning of each sweep step. Starting at -4V, the DAC
increments 11.2mV per step the sweep mode samples until 512 samples of each channel have
been completed; the DAC also completes 512 steps during this time. Figure 5.6 illustrates
the digital timing of the sweep.
5.3.8
Sample Timing
Additional clock information is included for further understanding of the DAS sampling.
The section describes how timing calculations for the DAS are and results are shown in Table
5.3. Each ADC has a 2MHz or 500ns Clock. The time taken to complete one conversion
is listed as clock per conversion. The conversion time is determined by the base clock
multiplied by the clocks per conversion. The sample time is the time taken to complete
1 sample. However, for a set this is multiplied by the number of instruments in the set,
which is 3, for each interleaved channel set. The DAS has clocks that determine the start
Table 5.2: EFAC spectrometer channels freqency bins.
Spectrometer Channels
EFDC
Ch-1
Ch-2
Ch-3
Ch-4
Start
0
16
32
64
128
Stop
40
32
64
128
512
Units
Hz
Hz
Hz
Hz
Hz
52
Ion Lang. Probe
Sweep Mode
Conversion #1
ILP Z- Axis
ADS8343
Conversion #2
ILP Z+ Axis
Conversion #3
FPP
Conversion #4
ILP Z- Axis
14.5 μS
START
MODE 1
ILP Sweep
Regular ILP Mode
(Mode 0)
Mode (Mode 1)
Conversion #6
FPP
Conv. #94
ILP Z- Axis
43.5 μS
Sleep
Mode
Sample #1
ILP Z -, ILP Z +, FPP
Conversion #95 Conversion #96
ILP Z+ Axis
FPP
43.5 μS
START
MODE 0
32 Conversions in a
Sample
1.392 ms
Sample #1
ILP Z -, ILP Z +, FPP
Conversion #5
ILP Z+ Axis
Sleep
Mode
Sample #512
ILP Z -, ILP Z +, FPP
SM.
Regular ILP Mode
(Mode 0)
1.5 ms or 667 Hz
768 ms or 667Hz
DAC7621
Voltage
(Counts)
-4 Volts
(F64h)
Settling
Time
7μs
-4 Volts
(F64h)
32 Conversions in a
Sample
-3.??? Volts
(F60h)
Settling
Time
7μs
+2 Volts
(764h)
Settling
Time
7μs
-4 Volts
(F64h)
Settling
Time
7μs
Fig. 5.6: ILP sweep sampling scheme.
of each conversion and start the sampling until the appropriate time. After this period of
sampling the ADCs are put in low-power sleep mode. This waiting period allows slightly
different co-adding schemes. The ADC sleep time is defined numerically as the difference
between the clock time and sample time. The EFAC contributes to the FFT timing and
is a singular instrument that is not part of a set, taken 1024 measurements which happen
every 1Hz. The ILP sweep has the same conversion time but co-adds 32 times instead of
256; this decreases the timing by 1.39ms and has a 1.5ms or 667Hz clock until to completion
512 measurements of the ILP Z+, ILP Z-, and FPP channels which takes 768ms. Table 5.3
shows the coadding and differences between the channels.
5.4
Data Flow Down
5.4.1
Real-Time Clock
Clock accuracy and the ability to associate instrumentation data with time is essential
for the science measurements. The time-stamp enables the ability to associate the data in
a spatial frame and cross referencing it with data from other missions. An incorrect timestamp can affect analysis as it is used for spatial reference; therefore, it is important that
the clock be as accurate as possible. A method was needed to sync the spacecraft clock with
another clock and time-stamp the data with 1ms resolution as shown in Figure 5.7. The
GPS system has a clock that can offer resolutions down to 1ns depending on the receiver
and the GPS signal. Once the GPS clock is synchronized with GPS time, DICE has a GPS
53
Table 5.3: DAS sample timing table.
Instrument
EFDC ch. 1
EFDC ch. 2
ILP ZILP Z+
FPP
EFAC
ILP Z+ Sweep
(Mode 1)
ILP Z- Sweep
(Mode 1)
FPP (Mode 1)
TAM X Axis
TAM Y Axis
TAM Z Axis
Base Clocks Conv.
Clock Per
Time
Conv.
500ns 36
18us
500ns 36
18us
500ns 29
14.5us
500ns 29
14.5us
500ns 29
14.5us
500ns 36
18us
500ns 29
14.5us
Co- Sample
adds Time
(set)
256 4.6ms (13.8ms)
256 4.6ms (13.8ms)
256 3.7ms (11.1ms)
256 3.7ms (11.1ms)
256 3.7ms (11.1ms)
8
4.6ms
32
0.464 (1.39ms)
500ns
29
14.5us
32
500ns
29
14.5us
32
3.12ms 1
3.12ms 1
3.12ms 1
3.12ms 0
3.12ms 0
3.12ms 0
Clock
Time
ADC
Sleep
Time
461us
461us
3.15ms
3.15ms
3.15ms
461us
108us
Meas.
Num.
(set)
14.28ms
N/A
14.28ms
N/A
14.28ms
N/A
14.28ms
N/A
14.28ms
N/A
N/A
1024
0.5ms
512
(1536)
0.464 (1.39ms) 0.5ms
108us 512
(1536)
0.464 (1.39ms) 0.5ms
108us 512
(1536)
3.12ms (9.37ms) 14.28ms 4.91ms N/A
3.12ms (9.37ms) 14.28ms 4.91ms N/A
3.12ms (9.37ms) 14.28ms 4.91ms N/A
Meas.
Time
(set)
N/A
N/A
N/A
N/A
N/A
N/A
256ms
(768ms)
256ms
(768ms)
256ms
(768ms)
N/A
N/A
N/A
system that is used for attitude and time keeping purposes. The GPS receiver sends out a
digital pulse called the Pulse Per Second (PPS) signal. After the signal is sent, the GPS
system updates the GPS week and millisecond registers in the C&DH system. The C&DH
updates a pre-clock register in the timekeeper module of the FPGA with the value of the
next PPS time. Upon the detection of the PPS, the register synchronizes its time with the
GPS and is the master clock for the satellite due to the poor accuracy of the C&DH’s clock.
All DICE data is time-stamped from the FPGA clock at the beginning of a sample.
5.4.2
Data Packets and Granules
The data is grouped by sampling rate before it is sent down to the ground. By packetizing the data, it enables a scheme by which the data can be managed and processed
through the satellite systems as it flows through several before it reaches the ground station. Granules are groups of data, the three types are: science, sweep, and EFAC granules
as shown in Figure 5.8. Each granule has a 1-byte header to delineate the granule type
to the C&DH system. The header is then followed by a 4-byte time-stamp, and is timestamped with the current time from the FPGA RTC with a time resolution of 1ms at the
moment the sampling started. Science granules are sent at a rate of 70Hz or 35Hz. In the
54
Steps:
1 GPS Gets a lock
and updates
C&DH registers
Command and
Data Handling
Actel Igloo
FPGA
UART
Pulse Per
Second
Signal
GPS
Real Time Clock
GPS Week
Pre mSec
GPS mSec
GPS mSec
2
C&DH Updates
FPGA’s
Pre-mSec
Register
Command and
Data Handling
Actel Igloo
FPGA
Real Time Clock
Pulse Per
Second
Signal
GPS
GPS Week
Pre mSec
GPS mSec
GPS mSec
2
The GPS pulse
Per second
Updates GPS
mSec register
Command and
Data Handling
Actel Igloo
FPGA
Pulse Per
Second
Signal
GPS
Real Time Clock
GPS Week
GPS mSec
Pre mSec
GPS mSec
Fig. 5.7: RTC register update.
event that the 35Hz mode is selected, the data is still collected but only every other packet
is sent. Processes such as the EFAC and sweep functions take a few seconds to complete.
Each time-stamp is then followed by data which is in Little-Endean format with the MSB
of the data being the MSB of the byte. Granules are sent out in 1-byte boundaries but are
generally aligned on a 16-byte boundary. The first packet in a sweep contains time stamp
information, whereas the other 64 packets in a sweep have a sequence number for order.
This 1-byte sequence number indicates the packets set number. Each granule channel has a
FIFO buffer that stores the granules and sends them through the UART when allowed by
the traffic controller which ensures that only one granule at one time is sent to the serial
data stream.
5.4.3
Packet Formation and the C&DH
The C&DH system receives the granules from the FPGA and converts them to a
telemetry specification called Consultative Committee for Space Data Systems (CCSDS),
packets that can be sent to the radio [22]. The CCSDS packets add some overhead to
55
Science Granule
35Hz or 70Hz
First Sweep Granule
120 seconds
Byte #
0 HEADER = A5
1
GPS ms when sampling began
3
DC Probe Z+
5
DC Probe Z7
9 Electric Field Probe V12 Axis
11 Electric Field Probe V34 Axis
Floa!ng Poten!al Probe
13
15 "00"MX(17:12)
MX(11:4)
17 MX(3:0)Y(17:14)
MY(13:6)
19 MY(5:0)Z(17:16)
MZ(15:8)
21
MZ(7:0)
Byte #
0 HEADER = B2
Sequence #0
2
GPS ms when sweep began
4
DC Probe+ Sweep Point 1
6
DC Probe- Sweep Point 1
8
Floa!ng Poten!al Probe 1
10
DC Probe+ Sweep Point 2
12
DC Probe- Sweep Point 2
14
Floa!ng Poten!al Probe 2
16
48
50
52
DC Probe+ Sweep Point 8
DC Probe- Sweep Point 8
Floa!ng Poten!al Probe 8
Subsequent Sweep Granules (63)
120 seconds
Byte #
0 HEADER = B2
Sequence #1
DC Probe+ Sweep Point 9
2
DC Probe- Sweep Point 9
4
Floa!ng Poten!al Probe 9
6
DC Probe+ Sweep Point 10
8
DC Probe- Sweep Point 10
10
12 Floa!ng Poten!al Probe 10
44
46
48
EFAC Granule
1Hz
Byte #
0 HEADER = F7
1 GPS ms when data was
sampled (before FFT)
3
5 Electric Field Probe AC CH1
7 Electric Field Probe AC CH2
9 Electric Field Probe AC CH3
11 Electric Field Probe AC CH4
DC Probe+ Sweep Point 16
DC Probe- Sweep Point 16
Floa!ng Poten!al Probe 16
Fig. 5.8: DAS granule types and byte sequences.
the telemetry, but also have important functions such as a check sum for error correction
and serve as a wrapper for the science granules. The granules form the main body of the
CCSDS packet but redundant data is compressed. The granule header is not included to
save space as shown in Figure 5.9. Because of the linear monotonic nature of the GPS time
signal, only the delta time is sent for most granules. The first packet header contains the
full time-stamp, and each subsequent CCSDS packet is referenced from the time header of
the first. This method saves telemetry because for all but one of the packets only 1-byte
out of the three bytes is sent via the telemetry stream. The number of granules in a packet
is 70, which number corresponds to 1 second of data, and the total packet size is 542 bytes.
There are also 16 ILP Sweep granules, the packets of which are 542 bytes in length.
5.4.4
Radio-to-Ground Station
DICE uses a novel software radio receiver for the ground station as shown in Figure
5.10. The antenna for the ground station consists of an 18.3m dish at Wallops, VA. The
radio is only half duplex, to switch between Tx or Rx modes, an amplifier\switch by single
sideband carrier ensures only one channel is talking at a time. The radio uplink consists
of a CC1101 made by Texas Instruments. The downlink is a WBX demodulator which
then feeds the data into a USRP2 software defined radio—this radio converts the data to
a digital stream. It is then processed by the GS computer and stored in a data base in its
raw, Level 0 bit form.
56
Address
0x0000
0x0002
0x0004
0x0006
15
14
13
12
11
10 9 8 7 6 5 4 3 2 1 0
Sec.
Packet Version Pac.
Applica!on Process Iden!fier (See
Hdr.
Number
APID sheet)
Type
Flag
Sequence
Packet Sequence Count or Packet Name
Flags
Packet Data Length
GPS Week Number (16 bits)
0x0008
0x000A
0x000C
GPS Milliseconds (32 bits)
0x0002
0
x
Granule 1
0x
0x0004
0x
0x0006
0x000C
(1* granule
size)
0x0008
Granules 2 through N-1
0x000A
0x000C
0x000C
(N - 1 *
granule
size)
0x000C +
(N * granule
size)
0x000E
0x
0x
0x0010
0x
0x0012
Granule N
DC Probe+ (int16)
DC Probe- (int16)
EF Probe 1-2 (int16)
EF Probe 3-4 (int16)
Floa!ng Poten!al Probe (int16)
(reserved)
MagX[17:4] (int18)
MagX[3:0]
MagY[17:6] (int18)
MagY[6:0]
MagZ[17:8] (int18)
MagZ[7:0]
deltaTime (uint8)
CheckSum
Fig. 5.9: Science granules in a CCSDS packet.
DICE
L3 Cadet
Radio
Connection to
Antenna Feed
18.3m Dish
Wallops, VA
Radio
Uplink
CC1101
SSB
Amplifier\
Filter
Power
Amplifier
D
Downlink
WBX
USRP2
Software Defined
Radio
Ground Station
Computer
Fig. 5.10: Ground station functional diagram.
Database
Server
57
Chapter 6
Calibration and Testing Results
The ideal instrumentation explained in the previous chapters varies because imperfections in analog instrumentation. Analog components are composed of smaller circuit
elements that are susceptible to temperature changes and nonlinearities. Each device is
also different due to manufacturing, no two components function in exactly the same way.
Because of these differences, error is introduced which affect performance in two ways: 1)
Offsets are created which shift the center range of the signal away from zero; 2) Changes
in dynamic range result from gain differences. Both parameters can also vary over temperature which introduces a third source of error. Calibration is necessary to form a model
of these differences by measuring error in the output with respect to a reference input.
Functional testing and verification were simultaneously conducted to save time and ensure
that the instrumentation meets the limits specified by the requirements. Figure 6.1 shows
the functional temperature testing of DICE.
Fig. 6.1: Temperature chamber.
58
6.1
Instrumentation Testing and Calibration
The DICE instrumentation was tested by applying a linear signal with a known ampli-
tude and frequency to the input of the instrumentation. This was either done by stepping
the parameter by hand or by generating a linear waveform. The output was recorded in
digital counts and a linear fit applied to the waveform. System error can be measured by
finding the residual created by subtracting the linear fit from the output. On orbit data
can later be corrected with the characterized error.
Testing was done with the backplane of the satellite built up inside of the chassis.
Control was provided by ground station software which sent commands to control the
instrumentation and received the testing data. The data was automatically recorded in
a database, which also simultaneously tested the end-to-end performance and functionality
of the satellite, software, and ground station. Each preformed test name and statistics were
recorded in an excel spreadsheet for documentation purposes. The tests were conducted
both inside and outside of the temperature chamber, shown in Figure 6.2. The details
of the testing will not be described in this thesis, the interested reader should obtain the
calibration documents for further information. However, a short description of testing done
for each instrument will be reviewed. Calibration was conducted on two different timeframes, 14-15 of Mar., 2011 and 11-24 of Sept., 2011. Because the ILP and EFP September
calibration runs were not processed at the time of writing, only the data from the March
runs will be included.
6.2
Ion Langmuir Probe Testing and Results
6.2.1
Ion Langmuir Probe Calibration
The ILP instrument essentially records current and outputs the quantized value from
the ADC. Resistors were used to calibrate the current into the ILP. The resistor values used
for this test were 92k, 110k, 130k, and 920k. Each resistor was placed in an aluminum block
to keep a constant temperature and calibrated so the resistance value was known at that
temperature. The test was conducted by starting a voltage sweep on the ILP which sweeps
59
Fig. 6.2: Yhatzee in temperature chamber ready for calibration.
from -4V to 2V. The change in current was then observed by recording ILP counts from
the ADC. This test was conducted at three different temperature points -12◦ C, 22◦ C, and
35◦ C.
6.2.2
Ion Langmuir Probe Calibration Results
The simplest way to calibrate the data is to find the first-order linear fit for voltage
versus counts. The data for the sweeps was obtained and a linear fit was applied to each.
By fitting the output data two main factors of the system are shown, the dynamic range
and gain. The dynamic range can be found in the graph where the full scale points of 0
and 65535 are on the graph. Gain is found by the slope of the line, error in the form on
nonlinearities can be found in the residual of the data as shown in equation (6.1). Another
system performance measure is offset which is the value of the output of the system at the
y-intercept of the zero-point of the x-axis. This can also be found in the residual at the
zero-point of the input signal.
Vcounts−out = (Kgain ) (Iin ) +Voffset
(6.1)
The results are shown in Figure 6.3 from six runs, two each for each temperature
point. The gain on each channel is related to the slope of the line, the higher the slope
the higher the gain. The residuals of the linear fits to the data, shown in Figure 6.4, also
60
give information about the linearity of the system and its accuracy. The channels do have
inconsistencies on a fine scale that change with temperature, most likely due to offsets in the
instrumentation. An unexpected result was the highly nonlinear nature of the ILP number
two channel as shown in Figure 6.4. A probable cause of this problem results from a zero
crossing on the ADC or the last offset amplifier in the system.
The linear fits of Figure 6.5 show temperature differences in the ILP data with about
100 counts of error in some cases. This results in a one sigma uncertainty of 30nA and was
a factor of 20 worse than anticipated. The resolution of the ILP will be lower than expected
at 104 cm−3 . More than a simple linear fit will be needed for the ILP data calibration.
For a more exact picture of the noise, a 10th order fit was applied to the data to subtract
any polynomial effects, and the worst case error at 22◦ C shows about 10 counts of error.
This equates to an amplitude of 381uV noise which suggests that DAC system noise is of
Langmuir Probes Linearity testing and Calibration
6.5
x 104
ILP1
ILP2
6
PCM counts
5.5
5
4.5
4
3.5
3
2.5
2
−4
−3
−2
−1
0
1
2
Voltage applied (V)
Fig. 6.3: Linear fits of ILP calibration data.
5
-12°C
Error in ILP applied voltage after calibrations
23°C
ILP1
ILP2
5
4
4
3
3
3
2
2
2
1
0
−1
Error (mV)
4
Error (mV)
Error (mV)
5
1
0
−1
1
0
−1
−2
−2
−2
−3
−3
−3
−4
−4
−5
−4
−3
−2
−1
0
Voltage applied (V)
1
2
−5
−4
35°C
−4
−3
−2
−1
0
1
2
−5
−4
−3
−2
Voltage applied (V)
Fig. 6.4: First-order linear fit and ILP calibration data.
−1
0
Voltage applied (V)
1
2
61
Error in ILP linearity
60
−12°C
ILP1
60
ILP2
40
Error (counts)
Error (counts)
40
23°C
20
0
20
0
−20
−20
−40
−40
−60
0
100
200
300
400
−60
0
500
Sweep Step Number
35°C
40
Error (counts)
60
20
0
−20
−40
100
200
300
400
500
Sweep Step Number
−60
0
100
200
300
400
500
Sweep Step Number
Fig. 6.5: ILP noise residuals.
concern and noise mitigation should have been employed for that system.
6.3
Electric Field Probe Testing and Results
6.3.1
Electric Field Probe Calibration
The EFP is much simpler to test since the system is converting voltage to counts.
The testing consisted of sweeping a voltage on each channel from -7V to 7V and recording
the counts out. This testing was done in conjunction with the ILP testing in March with
two runs at each -13◦ C, 22◦ C, and 35◦ C temperature points. The testing followed in this
manner:
1. The voltage was swept across a ±6.5V range across input channels V12 and V34 to
avoid the voltage limits of the instrumentation;
2. The results were recorded in a database by the ground station software;
3. The testing occurred at the same time as the ILP testing and was done for three
different temperature points -12◦ C, 22◦ C, and 35◦ C.
6.3.2
Electric Field Probe Calibration Results
The same methods used to analyze the ILP data was applied to the EFP data, a
linear fit and residual were found for each channel. The equation is similar to the previous
equation, also as shown in (6.2).
62
Vcounts−out = (Kgain ) (Vin ) +Voffset
(6.2)
The EFP was more well behaved and had less nonlinearities across temperature. The
gain and offsets were more consistent than the ILP stayed close to a ±2mV tolerance. The
same nonlinear crossover effect that shown in the ILP residuals are also present in the EFP
V34 channel residuals shown in Figure 6.6. This error is either caused by the second channel
of the ADS8343 or by the AD8622 amplifier which are present in both instruments. The
limits of the dynamic range of the instrumentation by extrapolating the voltage and hit full
scale at ±7V. The horizontal lines represent one sigma standard deviation after calibration.
For V12, Standard Deviation (STD) is 1.72mV, and for V34, STD it is 2.43mV.
6.4
Three Axis Magnetometer Testing and Results
6.4.1
Helmholtz Coil Calibration
To calibrate the TAM system, a DICE used Helmholtz method to drive the input of
the instrument. The DICE Helmholtz coil was built to test the TAM, shown in Figure
6.7. Because of the near straight magnetic field lines Helmholtz coils are ideal for testing
magnetometers. The coil frame was fabricated with a 3D rapid prototype fabrication printer.
Each copper coil was formed by wrapping wire into six coils, two for each axis. Each pair
of coils is wired in series with 10 turns of 26 American Wire Gauge (AWG) wire for each
7
x 10
4
Error in EFP measurement after calibrations with temperature
Electric Field Probes - Six Linearity Tests
V12
V34
6
6
Error (mV)
PCM Counts
4
5
4
3
2
0
−2
2
−4
1
−6
0
−6
−4
−2
0
2
Voltage applied (V)
4
6
−6
−4
−2
0
2
Voltage applied (V)
Fig. 6.6: EFP linear fit and residuals.
4
6
63
Zero Gauss
Chamber
DICE
Helmholtz Coil
FVM-400
Magnetometer
Fig. 6.7: Helmholtz coil calibration and zero-Gauss chamber.
coil. The radii for the coils are 4.49cm, 5.15cm, and 5.81cm from the inner coil to the outer
coil. To avoid confusion each axis was labeled to correspond to each axis. Helmholtz coil
equation (6.3) shows the predicted magnetic field for a given current.
3
4 2 µ0 nI
B=
5
R
(6.3)
The calibration for the Helmholtz coil was done in the following manner. Each channel
was calibrated individually with a FVM-400 test magnetometer that had an accuracy of
1nT. To reduce magnetic environmental noise the Helmholtz coil was calibrate in a zeroGauss chamber which reduces external magnetic fields to near zero. Care was taken to
align the test magnetometer and place it close to the center of the coil. A current was
driven into each channel and the magnetic strength recorded. The current and magnetic
field strength were measured at 15 points and compared with its theoretical performance.
For brevity only the results at 250mA are shown in Table 6.1, but the errors were similar for
each current measurement. These results were factored into the magnetometer calibration.
The Helmholtz coil was also calibrated for rotation in a similar manner described below.
64
Axis
Inner
Middle
Outer
6.4.2
X
Z
Y
Table 6.1: Helmholtz coil calibration results.
Diameter
Magnetic
Calculated
Field(nT)
Drive
@250mA
Field(nT)
4.32cm
52060
51100
4.95cm
48758
45440
5.59cm
45691
40180
Error
-1.90%
6.81%
12.13%
Three Axis Magnetometer Testing and Calibration
The TAM testing and calibration needed to calibrate both the gains and offsets, but
position error from misalignment from both construction and placement of the system which
resulted in sources of error. The system can be thought of as three independent linear
systems when corresponding axes are perfectly aligned and orthogonal to one another. Real
world situations prevent this and cross-terms must be used to characterize misalignments
between axes by substituting a coordinate transfer matrix A for the gain term as shown in
equation (6.4).
*
Vaxis =Again Baxis +Of f set
(6.4)
Misalignment of the axes causes them to be rotated which can be represented by matrix
A as shown in equation (6.5):


A11
A12
A13
*


V =  A21 A22 A23

A31
A32
A33


Bx

O1

 


 

  By  +  O2 

 

Bz
(6.5)
O3
Perfect alignment results in zero values for the terms which results in equation (6.6).
Each channel is multiplied by its own gain. Only one set of calibration equipment existed
only which meant only one axial magnetic source could be active at one time. However, by
calibrating each channel independently simplifies the linear cross gain problem.


A
A
A
12
13
 11

V =  A21 A22 A23

*
A31
A32
A33

B


O


V


A B
  x   1   x   11 x

 
 
 
  0  +  O2  ⇒  Vy  =  A21 Bx

 
 
 
0
O3
Vz
A31 Bx


O
  1 
 

 +  O2 
 

O3
(6.6)
65
After calibration linear fits are then applied to each channel three of the cross-gains may
be found because they are separate linear systems. It should be noted that the Helmholtz
coil also had a coordinate transfer matrix associated with its own calibration; these differences were removed from the data at the time of DICE TAM calibration.
TAM testing was conducted in the following manner with two separate input signals.
The Helmholtz coil was used to input a known magnetic field around the magnetometer, a
singular axis of the Helmholtz coil was energized at any moment in the test. The magnetometer and Helmholtz coil were placed inside of the zero-Gauss chamber to minimize EMI
from outside sources. A stepped input was applied to the magnetometer first stepping from
-125mA of current to 125mA of current. This tested the offsets, gains and alignment of the
TAM. A second wave form used to drive the input with a square, sine, or triangle wave into
the Helmholtz coil with a waveform generator with a known frequency and amplitude as
shown in Figure 6.8.
This simplified testing because only the amplitude and frequency were recorded as
opposed to the difficult task of aligning to different data sets and their time series. Matlab
was used to fit the wave form and produce a model that could reconstruct the original
waveform. The two were then subtracted and a residual obtained. Figure 6.9 shows the
TAM output and both sets of waveforms; in this figure misalignment can be where a driven
channels signal can be seen in the un-driven channels. Each of these tests were done over
three temperature points near -12◦ C, 21◦ C, and 35◦ C.
Fig. 6.8: Triangle wave input for TAM calibration.
66
4
8 x 10
Science Magnetometer Calibration 9/7/11a (All Data)
Bx
By
Step Input
6
ADC Counts (int18)
4
Bz
Wave Input
X
2
Y
0
−2
−4
Z
−6
−8
1
1.5
2
2.5
Index of Array
3
3.5
4 5
x 10
Fig. 6.9: TAM calibration inputs for Yhatzee testing.
6.4.3
TAM Calibration data
Figure 6.10 shows data and fits for the TAM, as well as the residuals of those fits, which
shows data from the September Yhatzee sine wave calibration. This example is shown for
driving the x-axis of the Helmholtz coil. The response of the magnetometer is shown mostly
from counts on the x-axis of the TAM. This slope of is also the gain factor A11. When cross
terms have low values it suggests better alignment.
Offsets are found by locating the magnetic field zero-point and recording the number
of counts at that point. Table 6.2 is shown for all of the Yhatzee cross terms and offsets
while Figure 6.11 shows the variation in temperature.
A considerable amount of noise was found in the residuals of the waveform fits subtracted from the data shown in the left of Figure 6.10. At 20 counts of noise, this was 10
times more than expected, however, on orbit and testing suggest that this is due to noisy
test equipment or fitting methodology. This supports the predicted value of three counts
and a sensitivity of 1.5nT characterizes the magnetometer and that the requirements were
achieved.
The offsets and gains over temperature are shown for the linear fits done over testing.
The calibration is linear over temperature for Farkel. Yhatzee, however, seems to reach its
67
Fit Residuals Helmholtz X−Axis
30
Response to Helmholtz X−Axis
−3000
−4000
20
Bx
By
Bz
0
ADC Counts
ADC Counts
−5000
10
−6000
−7000
−8000
−10
−9000
−20
−30
−5
−10000
−4
−3
−2
−1
0
1
2
magnetic field (nT)
3
4
5
4
x 10
−11000
−5
−4
−3
−2
−1
0
1
magnetic field (nT)
2
3
4
5
x 10 4
Fig. 6.10: TAM calibration data and residuals.
X-axis
A11
0.5796
Offset1
7929.6
A12
-0.0139
Offset2
32547
Table 6.2: Gain matrix for Yhatzee calibration.
Y-axis
Z-axis
A13
A21
A22
A23
A31
0.02712 -0.0551 0.6494 -0.0251 -0.01065
Offset3 Offset1 Offset2 Offset3 Offset1
-33150
8060.8
32539 -33169
8098.7
A32
0.01056
Offset2
32560
A33
0.62187
Offset3
-33104
lowest point and then turns up again. This may prove difficult since the temperature span
of the magnetometer could be a bit higher than 35◦ C due to its external location.
68
Gain Varia!on: Abs(Gain)
Gain (nT/Count)
Yhatzee A11
Farkel A11
Yhatzee A22
Farkel A22
Yhatzee A33
Farkel A33
0.2000
0.1800
0.1600
0.1400
0.1200
0.1000
0.0800
0.0600
0.0400
0.0200
0.0000
-20
-10
0
10
20
30
40
50
Temperature ( C )
Offset Varaia!on
Yhatzee Offset 1
Farkel Offset 1
Yhatzee Offset 2
Farkel Offset 2
Yhatzee Offset 3
Farkel Offset 3
8000.0
6000.0
Gain (nT/Count)
4000.0
2000.0
0.0
-2000.0
-4000.0
-6000.0
-8000.0
-10000.0
-12000.0
-20
-10
0
10
20
Temperature ( C )
30
40
50
Fig. 6.11: TAM gain and offset variations over temperature.
69
Chapter 7
Conclusion
7.1
DICE Comparison
7.1.1
Initial Mission Success
The DICE mission was launched at 09:48 UTC on October 28th 2011 from Vandenberg,
CA as an auxiliary payload on the NPP mission and reached its orbit successfully. The lead
DICE spacecraft was ejected with close to a 6mm/s delta-V imposed upon it by a separation
spring between the two spacecraft along the velocity vector. Over a period of five months,
the along-track separation between the spacecraft had grown to approximately 6,000km,
most likely due to slight differences in attitude, orbit, and resulting drag. The EFP locking
mechanisms, antennas, and ILP booms were successfully deployed by each spacecraft 50min
after ejection. Communications with both spacecraft was first achieved on October 30th 2011
or a few days after launch when valid two line orbital element sets became available from
the United States Space Surveillance Network. Tracking and ground station interference
problems delayed the commissioning of the spacecraft which consisted of de-tumbling the
vehicles, aligning them with the geomagnetic axis, and spinning them up for stabilization
and deployment of the wire booms of the EFP science instrument. The spacecraft attitude
control system will be switched into a 0.1Hz spin maintenance stabilization mode when the
spacecraft is fully deployed. At the time of writing downloading data proved difficult due
to radio noise in the ground station vicinity which resulted in limited amounts of data for
retrieval. This also resulted in a delay of the EFP booms being deployed fully. Problems
aside, both spacecraft were functioning correctly and housekeeping, ILP, and TAM data
being collected. The DICE mission and instrumentation demonstrates that low-power, low-
70
volume plasma diagnostic instrumentation can be developed for cubesats and stand the
rigors of the space time environment. By meeting the mission requirements and by its
performance on orbit, DICE shows that cubesats are viable means for gathering vital space
weather data.
7.1.2
Instrumentation Performance and Comparison
A comparison of DICE mission and the Communications/Navigation Outage Forecasting System (C/NOFS) mission which have instruments that are functionally similar;
a comparison is shown in Table 7.1. Instrumentation parameters were either similar or
exceeded by DICE, while using much less resources. A schematic for the instrumentation
is shown in Appendix A.2.
Although C/NOFS are not similar missions because C/NOFS has much more instrumentation and differed from DICE in the reliability requirements and construction. An
example of this is C/NOFS, which used more expensive radiation tolerant parts, whereas
DICE used commercial off-the-shelf parts which cost less and are less reliable. However, the
instruments DICE used were comparable in sensitivity and sampling rate while conserving
mass and power. Another factor to consider is cost C/NOFS budget was 153 million dollars, while DICE mission costs were 1.2 million dollars [23]. Many small satellites could be
purchased for the price of a larger mission and provide greater coverage.
7.2
Future Work
7.2.1
ASSP
The DICE mission paves the way for cubesats and other missions by proving new
technologies and methods in the rigors of the spacetime environment. A DICE successor
that will build off of these technologies is a sounding rocket mission called Auroral Spatial
Structures Probe (ASSP) which will spin out ten instrumentation packages functionally
similar to DICE and measure the Electric field, Magnetic field, and density of the auroral
region. The advantage of smaller instrumentation package is more multipoint measurements
71
Table 7.1: DICE performance and comparison.
EFAC
Parameter
DICE
Sensitivity 0.002
Dynamic
50
Range
Frequency
16Range
512
DC Electric Field
Parameter
DICE
Sensitivity 400
Dynamic
700
Range
Mass
<0.1
Power
50
Sampled
70
Rate
FPP
Parameter
DICE
Sensitivity 400
Dynamic
700
Range
Sampled
70
Rate
C/NOFS Units
0.005
mV/m
45
mV/m
16- 512
Hz
C/NOFS Units
300
uV/m
45
mV/m
>1
>50
1
kg
mW
Hz
C/NOFS Units
300
uV/m
45
mV/m
1
Magnetometer
Parameter
DICE
Sensitivity
1.5
Dynamic
±196000
Range
Mass
<0.1
Power
110
Sampled
70
Rate
ILP/PLP
Parameter
DICE
Accuracy
10
Resolution
3000
Dynamic
104
Range Low
Dynamic
108
Range High
Mass
<0.1
Power
50
Sampled
70
Rate
C/NOFS Units
50
nT
45000
nT
>1
>50
1
kg
mW
Hz
C/NOFS
500
3.5
102
Units
%
cm−3
cm−3
108
cm−3
>1
>50
1
kg
mW
Hz
Hz
which increases the resolution of spatial and temporal scales. The differences of ASSP are:
it is smaller because the power and ACDS systems will be reduced, the electric field booms
will be 3m of length, and it will spin faster to release the booms.
7.2.2
Satellite Constellations
DICE is amongst the first cubesat constellations by launching two satellites that make
ionospheric measurements. Launching a single satellite makes it difficult to see small scale
variations or global variations. Constellations overcome these problems and give a better
view of space weather phenomena. Instead of building one single expensive satellite, several
smaller inexpensive satellites may be deployed that have the same functionality but not
necessarily the same level of redundancy and reliability. In the future, many small satellites
could be deployed in the same mission to give global coverage as shown in Figure 7.1.
72
Fig. 7.1: Future satellite constellation.
7.3
Lessons Learned
7.3.1
Science Board
Because of limited space in volume for the payload, the science board layout had a
high part density that made it very complex. Future missions should work to lower part
densities by increasing board space or by lowering part count. The science board PCB
was a 6-layer design which made it expensive, in future designs using the same volume in
two separate 4-layer boards could be used to house the same parts and lower part density.
Analog and digital sections could be separated on different boards as well as isolating the
power electronics from sensitive analog electronics. Space could also be saved by using a
battery bus voltage greater than 12V, this would allow for DC-DC regulators instead of
charge pumps. This approach would also lower design costs by giving more board space for
designers to work with and lowering the cost of board changes. FPGAs make digital design
more flexible, if a pin needs to be changed the FPGA can re-route the signal. Another
problem is the envelope needs to be defined early on for the payload. In the case of DICE
the Sun-sensors cut into the board which trimmed off additional space which resulted in
73
the 6-layer PCB design and increased the cost and design time of the board.
7.3.2
Ion Langmuir Probe
The ILP did not meet its resolution requirement of 1x103 cm−3 because the dynamic
range was adjusted from being a fixed bias probe with a range of 700pA to 7uA to being a
dual sided ±50uA signal for input current. A better approach would be to use a logarithmic
amplifier or to make the channel have a high gain and a low-gain stage to compensate.
Another major problem with the instrument was not performing a system analysis of the
noise for the DAC system which contributed noise amplitudes up to 10uA at the input of
the system. This problem can be solved by putting a low-pass filter before the unity gain
stage of the DAC system. The low-pass filter of the ILP also needed modification. Low-pass
filters typically have an impedance buffer to separate them from the rest of the circuit. In
this case, there was a resistor accidently placed between the output of the low-pass filter
and the input of the next amplifier stage. This turned the low-pass filter in to more of a
band-pass filter. This was corrected by the pole is not exactly at 40Hz. Future designers
should refer to the EFDC low-pass filters for correct schematics of a low-pass filter stage.
7.3.3
Electric Field Probe
The EFP met its goals. However, power could have been saved by downsizing the highpass filter on the EFAC channel. The spin rate range at design time was 400mV/m at 1Hz
to 9Hz, so the high-pass filter was designed to accommodate any signal across that range.
The ending spin rate was around 2.5Hz max and the 60dB/decade filter was overdesigned.
However, future missions should take care in modifying the design if they have a higher spin
rate. An undiscussed problem belonging to the EFDC channel is the shielding of the boom
wires. The capacitance between the wires and the plasma create undesirable signal effects.
An attempt to compensate with a graphite spray to shield the wires. This attempt was not
able to coat the wires evenly enough to provide continuous conduction. Future missions
should use a small gauge wire with an inner or shielding conductor.
74
7.3.4
Magnetometer
The magnetometer met its goal of 2nT, however, there is room for improvement in
future revisions of this instrument. The system bandwidth of the magnetometer had been set
too high which added a considerable amount of noise to the system, reducing the bandwidth
to 40Hz from 150Hz could reduce the noise amplitude by a factor of 2.5. To also improve
sensitivity the bridge offsets could be compensated for as described in Section 4.5. The
dynamic range could then be reduced 2 to 4 times in doing so, this would also improve
system sensitivity. Better shielding on the cable could also reduce noise on the instrument
since the cable runs down the length of the satellite and is susceptible to EMI. Further
analysis of the on orbit data reveled spikes of up to 10nT which are probably caused by
the power electronics. Placing the magnetometer greater than 6cm away from the center of
the spacecraft could reduce the noise to a level lower than the sensitivity of the instrument.
The instrument was not moved on DICE because of potential mechanical issues.
75
References
[1] B. D. Wilson, G. A. Hajj, L. Mandrake, X. Pi, and C. Wang, “Witnessing a revolution
in ionospheric remote sensing,” American Geophysical Union Fall Meeting Abstracts,
p. B519, Dec. 2003.
[2] R. P. McCoy, “Space weather comes of age: new sensors and models for ionospheric
specification and forecast,” in Society of Photo-Optical Instrumentation Engineers
(SPIE) Conference Series, Hung-Lung A. Huang, Hal J. Bloom, Ed., vol. 5548, pp.
341–347, Oct. 2004.
[3] NASA. (2009) Nasa heliophysics roadmap. [Online]. Available: http://sec.gsfc.nasa.
gov/sec roadmap.htm
[4] J. C. Foster and H. B. Vo, “Average characteristics and activity dependence of the
subauroral polarization stream,” Journal of Geophysical Research (Space Physics), vol.
107, p. 1475, Dec. 2002.
[5] J. C. Foster, “Storm time plasma transport at middle and high latitudes,” Journal of
Geophysical Research (Space Physics), vol. 98, pp. 1675–1689, Feb. 1993.
[6] “Cubesat: Dynamic ionosphere cubesat experiment (dice),” Proposal, Space Dynamics
Laboratory, 2008.
[7] S. Lee. (2009) Cubesat design specification. [Online]. Available: http://www.cubesat.
org/images/developers/cds rev12.pdf
[8] R. Burt, “Distributed electrical power system in cubesat applications,” Master’s thesis,
Utah State University, Logan, UT, 2011.
[9] C. Carlson, “Next generation plasma frequency probe instrumentation technique,”
Master’s thesis, Utah State University, Logan, UT, 2004.
[10] R. F. Pfaff, “In-situ measurement techniques for ionospheric research,” in Modern Ionospheric Science, H. Kohl, R. Rster and K. Schlegel, Ed., Katlenburg-Lindau: European
Geophysical Society, pp. 459–551, 1996.
[11] F. Chen, “Lecture notes on langmuir probe diagnostics,” in Proceedings IEEE International Conference on Plasma Science Mini-Course on Plasma Diagnostics, Jun. 2003.
[12] K. Bradford, “Miniature wire boom systems for nano satellites,” Master’s thesis, Utah
State University, Logan, UT, 2013.
[13] R. Morrison, Grounding and Shielding Techniques in Instrumentation.
John Wiley and Sons, 1977.
New York:
[14] J. P. Heppner, M. C. Liebrecht, N. C. Maynard, and R. F. Pfaff, “High-latitude distributions of plasma waves and spatial irregularities from de 2 alternating current electric
field observations,” Journal of Geophysical Research, vol. 98, pp. 1629–1652, 1993.
76
[15] J. Gregory, “Design, test, and calibration of the Utah State University floating potential
probe,” Master’s thesis, Utah State University, Logan, UT, 2009.
[16] D. Lancaster, Don Lancaster’s Active-filter Cookbook.
Indianapolis: Newnes, 1996.
[17] J. Springmann, C. J., and H. Bahcivan, “Magnetic sensor calibration and residual
dipole characterization for application to nanosatellites,” in AIAA/AAS Astrodynamics
Specialist Conference Proceedings, Aug. 2010.
[18] H. Bahcivan, M. C. Kelley, and J. W. Cutler, “Radar and rocket comparison of UHF
radar scattering from auroral electrojet irregularities: Implications for a nanosatellite
radar,” Journal of Geophysical Research (Space Physics), vol. 114, no. 13, p. 6309, Jun.
2009.
[19] Honeywell. (2008) 1- and 2-axis magnetic sensors hmc1001/1002/1021/1022. [Online].
Available: http://www51.honeywell.com/aero/common/do...et.pdf
[20] M. J. Caruso, C. H. Smith, T. Bratland, and R. Schneider, “A new perspective on
magnetic field sensing,” Honeywell, Technical Report, May 1998.
[21] R. Pfaff, NASA/GSFC. (2011) C/nofs, vector electric field instrument (vefi). [Online].
Available: http://www.kirtland.af.mil/shared/media/document/AFD-080721-032.pdf
[22] Consultative Committee for Space Data Systems. (2003) Space packet protocol
ccsds 133.0-b-1 blue book. [Online]. Available: http://public.ccsds.org/publications/
archive/133x0b1c1.pdf
[23] Committee on Assessment of Impediments to Interagency Cooperation on
Space and Earth Science Missions; National Research Council, Assessment
of Impediments to Interagency Collaboration on Space and Earth Science
Missions. The National Academies Press, 2011. [Online]. Available:
http:
//www.nap.edu/openbook.php?record id=13042
77
Appendices
78
Appendix A
A.1
Science and Mission Instrumentation Requirements
This section contains the mission requirements, instrument requirements, and also de-
rived instrument requirements from which the instrumentation was developed.
Table A.1: Mission and instrumentation requirements.
Req.
Description
Mission Objectives
MO1 Investigate the physical processes responsible for formation of the geomagnetic
storm enhanced density bulge in the noon to post-noon sector during magnetic
storms.
MO2 Investigate the physical processes responsible for the formation of the geomagnetic storm enhanced density plume which forms at the base of density bulge
and the transport of the high density plume across the magnetic pole.
MO3 Investigate the relationship between the penetration electric fields and the formation and evolution of the storm enhanced density bulge and plume.
Science Requirements
S01
Return continuous observations of the ionosphere using two spacecraft that are
within the same orbital plane and within 1-min to 6-min of each other.
S02
Return observations of the ionosphere from ≥ 55◦ latitude with a preference of
observations ≥ 80◦ latitude.
S03
Return 90 days of observations of the ionosphere from the 13 to 17 local time
sector using two spacecraft that are within the altitude range of 350km to 800km
with a goal of 180 days. Run
S04
Return observations of co-located electric fields, magnetic field fluctuations, and
plasma density at a ≤ 10km on-orbit spatial sampling with a ≤ 0.1km goal and
that are absolute time located to within ≤ 1ms UT.
S05
Return observations of the presence of electric fields fluctuations in the 10 to
1000Hz range at a ≤ 10km on-orbit spatial sampling.
Instrumentation Requirements
Electric Field Instrument
E1
The spacecraft shall return post-flight, post-analysis, observations of the ambient
electric fields perpendicular to the local magnetic field with and accuracy of ≤
2mV/m and with a goal of 0.1mV/m over a range of ±200mV/m.
Continued on next page...
79
Table A.1 – continued from previous page
Description
The electric field instruments shall provide two double probe observations with
sensors deployed 90◦ ±0.1◦ relative to each other and simultaneously sampled.
E3
Each electric field sensors shall be spherical, ≥ 0.75cm diameter and deployed
≥ 2 meters from the spacecraft with a goal of 10m tip to tip distance.
E4
The location of each of the electric field sensors in the spacecraft body coordinate
shall be known to within 1/1000 of the boom length when the sensors are fully
deployed.
E5
The electric field sensors shall be rotated about the axis perpendicular to the
deployed plane of the sensors with a frequency, f, such that 0.1 ≤ f ≤ 10Hz and
at the geometric center point of the sensors to within 1/1000 of the boom length.
E6
The electric field booms shall be deployed to be perpendicular to the Earths
geodetic axis of rotation to ≤ 30◦ .
E7
The post-flight attitude knowledge of the electric field boom orientation shall be
≤ 0.1◦ with a goal of ≤ 0.01◦ .
E8
The electric field instrument shall provide measurements under the environmental conditions for ambient plasma density of 102 Ohms to 108 Ohms/cm−3 .
E9
The electric field instrument shall provide measurements under the environmental conditions for temperature ranging from -10 to 30◦ C.
E10
The electric field instrument shall operate from power supplies of +5V and the
unregulated spacecraft battery bus (7.4 ±0.8V)). IT shall have a total power
usage of ≤ 50mW .
E11
The electric field instrument shall require no more than 4.0 x 4.0cm of circuit
board space.
E12
The input impedance of the electric field instrument shall be ≥ 1013 Ohms and
the instrument shall be in a short to ground configuration until deployed.
E13
The electric field probe shall be able to accommodate an induction field, VxB,
plus ambient field of ≥ 600mV/m without saturating.
E14
The electric field instrument shall provide two 16-bit two’s complement values
to the telemetry system representing the double probe measurements, V12 and
V34, of the observed fields.
E15
The electric field instrument shall provide at least 3 points of electric field spectral information from the frequency range of 10 to 1000Hz at a ≤ 10km onorbitspatial sampling rate.
Magnetic Field Science Instrument
M1
The science magnetometer shall return post-flight, post-analysis observations of
ambient magnetic fields with an accuracy of 2 nT and a signal to noise ratio of
≥ 3 over a range of ≥ ±50,000nT.
M2
The science magnetometer z-axis shall be aligned with the spin axis of the spacecraft to within 2◦
M3
The science magnetometer x and y-axis shall be aligned to the electric field boom
axis to within 2◦ .
Continued on next page...
Req.
E2
80
Table A.1 – continued from previous page
Description
The spacecraft shall have residual and stray magnetic fields of rms amplitude of
≤ 2nT at the location of the science magnetometer.
M5
The science magnetic field instruments shall operate with a power supply of
+5V with a total power usage of ≤ 100mW .
M6
The science magnetometer shall provide measurements under the environmental
conditions for temperature ranging from -10◦ C to 30◦ C
M7
The science magnetometer shall require no more than 4.0 x 4.0cm of circuit
board space.
M8
The science magnetometer shall provide three 18-bit twos complement values to
the telemetry system representing Bx, By and Bz of the measured field.
Ion Langmuir Probe
L1
The Ion Langmuir Probes shall return post-flight, post-analysis plasma density
observations over the range of 2x103 to 2x107 cm−3 with a resolution of 350cm−3 .
L2
The voltage on the sensor shall be ≤ -5V relative to the spacecraft structure
with a goal of -8V.
L3
The Ion Langmuir Probe shall be aligned with the spacecraft spin axis and
return observations of the ion ram current in wake free regions.
L4
The Ion Langmuir Probe shall provide two observations with sensors deployed
180◦ C ± 0.1◦ C relative to each other and simultaneously sampled.
L5
Each Ion Langmuir Probe sensor shall be spherical, ≥ 1.27cm diameter and
deployed ≥ 5cm from the spacecraft such that one of the sensors shall be outside
of the spacecraft wake at all times.
L6
The Ion Langmuir Probe shall provide measurements under the environmental
conditions for temperature ranging from -10◦ C to 30◦ C.
L7
The Ion Langmuir Probe instrument shall operate from power supplies of +5V
and the unregulated spacecraft battery bus (7.4V ±0.8V). It shall have a total
power usage of ≤ 50mW.
L8
The Ion Langmuir Probe shall require no more than 4.0 x 4.0cm of circuit board
space.
L9
There shall be no exposed potentials on the spacecraft to the space environment aside from the Ion Langmuir Probe. All solar panel interconnects shall
be isolated from the plasma and all connectors shall similarly be covered during
operation of the spacecraft.
L10
The surface of the Ion Langmuir Probe shall be cleaned prior to flight.
L11
The surface of the Ion Langmuir Probe will be goldcoated.
Science Instrument Interface
SI1
The science Instruments will interface to the rest of the spacecraft using a single
SPI interface
SI2
The maximum sampling rate for any science data channel shall be 100Hz
SI3
Science preamplifiers must be located as close as possible to the sensor mechanical interfaces.
Continued on next page...
Req.
M4
81
Table A.1 – continued from previous page
Description
Digital control lines may not be routed through the areas allocated to the Electric
Field Instrument, Science Magnetometer, and Ion Langmuir probe.
Derived Instrumentation Requirements
Electric Field Instrument
ED1
The input voltage range of the Electric field instrument shall be ±8V for a 10
meter tip to tip boom length. This requirement is derived from E1, E3, E13 and
includes a 30% margin.
ED2
The telemetry rate for the V12 and V34 channels shall be 80Hz giving a spatial
resolution of 93m at 350km altitude and 96m at 800km. This requirement is
derived from SO3, SO4, SI2.
ED3
The input bias current of the instrument shall be lower than 1pA and input
impedance more than 101 3Ohms. This requirement is derived from E8 and E12
ED4
The frequency response for the channels V12 and V34 shall be a DC low-pass
response with a near linear phase response within the pass band and with at
least a 20 dB/decade roll-off for out of band signals. The 3dB cutoff frequency
shall be 35Hz. This requirement is derived from ED2.
ED5
The electric field instrument response shall be calibrated over the temperature
range of -10 to 30◦ C using a polynomial. This requirement is derived from E1
and E9
ED6
The temperature of the electric field instrument electronics shall be known to
within 0.2◦ C while collecting science data and telemetered at ≥ 17mHz . This
requirement is derived from ED5
ED7
The electric field spectrometer shall consist of four channels which cover the
spectral ranges of 16-32Hz, 32-64Hz, 64-128Hz and 128-512Hz. The gain for
these channels shall be 500 . This requirement is derived from SO5 and E15
ED8
The electric field spectrometer shall employ a high pass filter to reduce the VxB
spin induced signal. The roll off of the filters shall be at least 60dB per decade
for the out of band signals with a cutoff frequency of no more than 12Hz.
ED9
The insulation on the electric field wire booms shall have a volume resistively of
more than 1017 Ohms/cm.
Ion Langmuir Probe
LD1
The input current range of the Ion Langmuir Probe shall be from 700pA to 7uA
referenced as positive current to the probe surface from the space environment.
This operating range shall be quantized with at least 16-bits. The operating
range is calculated for a 1.9cm diameter spherical sensor. This requirement is
derived from L1 and L5.
LD2
The telemetered rate for the ILP1 and ILP2 data shall be 80Hz. This rate
results in a spatial resolution of 93m at 350km altitude and 96m at 800km. This
requirement is derived from SO3, SO4, SI2.
Continued on next page...
Req.
SI4
82
Table A.1 – continued from previous page
Description
The frequency response for the channels ILP1 and ILP2 shall be a DC low-pass
response with a near linear phase response within the pass band and with at
least a 60 dB/dec roll-off for out of band signals. The 3dB cutoff frequency shall
be 35Hz. This requirement is derived from LD2
LD4
The Ion Langmuir Probe response shall be calibrated over the range of -10◦ C to
30◦ Cusing a polynomial. This requirement is derived from L6
LD5
The Ion Langmuir Probe electronics temperature shall be known to within 0.2◦ C
while collecting science data and telemetered at ≥ 17mHz. This requirement is
derived from LD4
Magnetic Field Science Instrument
MD1 The science magnetometer sensor head shall be kept at least 15cm away from
the power supply conditioning electronics. This requirement is derived from M1
and M3.
MD2 The telemetered rate for the Magnetic Field Instrument shall be 80Hz giving a
spatial resolution of 93m at 350km altitude and 96m at 800km. This requirement
is derived from SO3, SO4, SI2.
MD3 The Magnetic Field Instrument response shall be calibrated over the range of
-10◦ C to 30◦ C using a polynomial. This requirement is derived from M6.
MD4 The Magnetic Field Instrument electronics temperature shall be known to within
0.2◦ C while collecting science data and telemetered at ≥ 17mHz . This requirement is derived from MD3.
Req.
LD3
A.2
Science Board Schematics
The following section contains the Science Board schematics taken from SDL document
141-0007 which is the electrical design of the science instrumentation.
A
B
C
D
7
Form No. QF0412 Rev D
8
7
6
6
DICE Stack Connector
THIS DRAWING CONTAINS INFORMATION THAT IS PROPRIETARY TO SPACE DYNAMICS LABORATORY (SDL). REFERENCE USURF BP 409.1
8
5
5
4
4
3
3
4/20/2010
LAST MODIFIED
T. Neilsen
ENGINEER
NEXT ASSY TITLE
NEXT ASSEMBLY
DICE
PROGRAM
2
OF
SCHEMATIC
INSTANCE
OF
INSTANCE NAME
1
SCH REV
OF
A
B
C
D
@HierPageTotal
TOTAL SCHEMATIC PAGES
@HierPageNumber
01 OF 07
SCH-INST SHEET
DRAWING NUMBER
DICE Science Payload Board
North Logan, Utah 84341
UTAH STATE UNIVERSITY RESEARCH FOUNDATION
141-0007
SCHEMATIC NAME
TITLE
1
SPACE DYNAMICS LABORATORY
Deck Plate Connector
2
83
FPGA with DICE Science HDL Design (141-0004)
84
A
B
C
D
Spec Temp
‐40 to +85 °C
Ib
30 (120) fA
Iout
±5 mA
Slew Rate
0.4 V/us
Guard Drive Iout
+2/‐0.05 mA
Op Temp
‐40 to +125 °C
Iq
1.4 mA
Vout
Vcc/Vee ±1 V
GBW (G = 1)
800 KHz
Guard Drive Voff
50 mV
Guard Output Imp
650 O
Vcc/Vee
±18V
Ioff
10 (40) fA
Cap Stability
1 nF
Settling Time (0.01%)
22 us
7
Vin
Vcc/Vee ±4 V
Voff
5 mV
Input Impedance
> 10 15 O / 7 pf
Vn (RTI)
2 uVp‐p (to 10 Hz);
28 nV/(Hz)1/2 (1kHz)
CMRR
73 dB
Package
SOL ‐ 16
Overvoltage
±40 V
Voff Drift
10 uV/°C
Output Impedance
Not Available
In
0.1 fa/(Hz) 1/2 (1kHz)
Form No. QF0412 Rev D
8
7
THIS DRAWING CONTAINS INFORMATION THAT IS PROPRIETARY TO SPACE DYNAMICS LABORATORY (SDL). REFERENCE USURF BP 409.1
8
6
6
5
5
4
Package
SOIC (N) ‐ 8
PSRR
70 d B @ 1 kHz
Spe c Te m p
‐4 0 to + 8 5 °C
Ioff
3 5 pA (5 0 0 pA )
C ap Stability
1 nF
Se tt ling Tim e (0.01% )
2 0 us
Op Te m p
‐4 0 to + 1 2 5 °C
Ib
3 5 pA (5 0 0 pA )
I ou t
± 4 0 m A
Sle w Rate
0.4 8 V/us
V off
2 0 u V
I npu t Im pe dance
> 1 0 1 2 O / 3 p f
V n (RTI )
0.2 uVp‐p (to 10 Hz);
1 1 nV/(Hz) 1 / 2 (1 kHz)
V cc/V e e
± 18 V
V in
Vcc /Ve e ± 1 .2 V
± 1 0 V Differ ential
Voff Drift
1.2 uV/°C
Output Im pe dance
1 .5 O @ 1 kHz
In
0.1 5 fa/(Hz)1 / 2 (1 kHz)
E-Field Instrumentation
4
3
Vout
Vcc /Ve e ±0 .2 5 V
GBW (G = 1)
8 0 0 KHz
CMRR
1 2 0 dB
Iq
2 4 0 uA
3
4/20/2010
LAST MODIFIED
T. Neilsen
ENGINEER
NEXT ASSY TITLE
NEXT ASSEMBLY
DICE
PROGRAM
Spec Temp
‐40 to +85 °C
2
OF
3 LSB
Digital Level
5.5V
100 kHz
Data Format
Binary 2s Compl
4 Multiplexed
Package
SSOP ‐ 16
1 mV
Channels
Vin
‐0.3 to Vcc+0.3V
Bipolar Error
INSTANCE
OF
INSTANCE NAME
1
SCH REV
OF
A
B
C
D
@HierPageTotal
TOTAL SCHEMATIC PAGES
@HierPageNumber
03 OF 07
SCH-INST SHEET
DRAWING NUMBER
DICE Science Payload Board
North Logan, Utah 84341
UTAH STATE UNIVERSITY RESEARCH FOUNDATION
141-0007
SCHEMATIC
Iq
2 mA
ILE
±6 LSB
Throughput
15 bits
PSSR
Vcc
6V
No Missing Codes
1
SPACE DYNAMICS LABORATORY
20 uVrms
Channel Isolation
100 dB
Noise
Input Capacitance
25 pF
SCHEMATIC NAME
TITLE
0.024%
THD
‐95 dB
Gain Error
5 MO
Op Temp
‐40 to +85 °C
Input Resistance
Peak Reverse Voltage Forward Cont. Current Forward Surge Current
40 V
200 mA
600 mA
Reverse Leakage Current
Total Capacitance
Reverse Recovery Time
20 nA
4 pF
5 ns
Temperature Range
‐55 to +125
2
85
A
B
C
D
Op Temp
‐55to +150 °C
Linearity
2% FS
BW
5 MHz
Max Field
10000 Gauss
7
Spec Temp
‐55to +150 °C
Hysterisis Error
0.1% FS
S/R Strap
1.8 O (3.0 Amps)
Resistance TmpCo
0.25%/°C
Vcc
12 V
Repeatability Error
0.1% FS
Disturbing Field
3 Gauss
Sensitivity TmpCo
0.3%/°C
Form No. QF0412 Rev D
8
7
THIS DRAWING CONTAINS INFORMATION THAT IS PROPRIETARY TO SPACE DYNAMICS LABORATORY (SDL). REFERENCE USURF BP 409.1
8
Resistance
600‐1200 O
S/R Repeatibility
100 uV
Bridge TmpCo
0.001%/°C
Sensitivity
3.1 mV/V/Gauss
6
Field Range
±2 Gauss
Bridge Offset
‐60 mV
CrossAxis Error
0.5% FS
Package
SIP – 8
SOIC – 20
6
4
5
4
Magnetometer Instrumentation
5
3
3
4/20/2010
LAST MODIFIED
T. Neilsen
ENGINEER
NEXT ASSY TITLE
NEXT ASSEMBLY
DICE
PROGRAM
2
OF
SCHEMATIC
SPACE DYNAMICS LABORATORY
1
INSTANCE
OF
INSTANCE NAME
1
SCH REV
OF
A
B
C
D
@HierPageTotal
TOTAL SCHEMATIC PAGES
@HierPageNumber
04 OF 07
SCH-INST SHEET
DRAWING NUMBER
DICE Science Payload Board
North Logan, Utah 84341
UTAH STATE UNIVERSITY RESEARCH FOUNDATION
141-0007
SCHEMATIC NAME
TITLE
2
86
A
B
C
D
7
Form No. QF0412 Rev D
8
7
6
6
r
o
t
a
r
e
n
e
G
p
e
e
w
S
C
A
D
THIS DRAWING CONTAINS INFORMATION THAT IS PROPRIETARY TO SPACE DYNAMICS LABORATORY (SDL). REFERENCE USURF BP 409.1
8
4
5
4
DC Probe Instrumentation
5
3
3
4/20/2010
LAST MODIFIED
T. Neilsen
ENGINEER
NEXT ASSY TITLE
NEXT ASSEMBLY
DICE
PROGRAM
2
OF
SCHEMATIC
SPACE DYNAMICS LABORATORY
1
INSTANCE
OF
INSTANCE NAME
1
SCH REV
OF
A
B
C
D
@HierPageTotal
TOTAL SCHEMATIC PAGES
@HierPageNumber
05 OF 07
SCH-INST SHEET
DRAWING NUMBER
DICE Science Payload Board
North Logan, Utah 84341
UTAH STATE UNIVERSITY RESEARCH FOUNDATION
141-0007
SCHEMATIC NAME
TITLE
2
87
A
B
C
D
7
Form No. QF0412 Rev D
8
7
6
5
6
5
1.05VD Generation
TI Coprocessor (option B)
THIS DRAWING CONTAINS INFORMATION THAT IS PROPRIETARY TO SPACE DYNAMICS LABORATORY (SDL). REFERENCE USURF BP 409.1
8
4
4
3
3
1
4/20/2010
LAST MODIFIED
T. Neilsen
ENGINEER
NEXT ASSY TITLE
NEXT ASSEMBLY
DICE
PROGRAM
2
OF
SCHEMATIC
INSTANCE
OF
INSTANCE NAME
1
SCH REV
OF
A
B
C
D
@HierPageTotal
TOTAL SCHEMATIC PAGES
@HierPageNumber
06 OF 07
SCH-INST SHEET
DRAWING NUMBER
DICE Science Payload Board
141-0007
SCHEMATIC NAME
TITLE
North Logan, Utah 84341
UTAH STATE UNIVERSITY RESEARCH FOUNDATION
SPACE DYNAMICS LABORATORY
FPU Coprocessor (option A)
2
88
Stuff only one side of this page
A
B
C
D
Spe c Te m p
‐4 0 to + 8 5 °C
L oad Re gulat ion
2 50 p pm /V
C ap St ab ilit y
> 10 u F
V in
2 0V
V DO
1.75 V (@ 1.25 Vo ut)
V n (RTI)
1 0 u Vp ‐p (to 10 H z) ;
14 u VR M S (1 0 t o 1 kH z)

  I A D J  R 2 A D J   5 . 7 9 9

1.2VD Generation

R2A D J

VA D J   1 
R1A D J
Form No. QF0412 Rev D
8
7
R 1 A D J   1 . 0 5 1 0
SOIC ‐ 8
20 uA
 9
Iq
.9 mA
Iout
550 mA
Package
Vin
±10V
VDO
110 mV
Reverse Leakage I
I A D J   2 0 1 0
300 uVrms (to 100 kHz)
1 uV/(Hz)1/2 (1kHz)
> 10 uF
VA D J   0 . 2
Spec Temp
‐40 to +85 °C
Load Regulation
2 mV
Vn (RTI)
Op Temp
‐40 to +125 °C
Line Regulation
1.75 mV
Cap Stability
Iq
7 5 uA
I ou t
+ 5 /‐1 mA
V ou t TC
1 0 pp m/°C
+1.25VA Generation
+5.8V Generation
50 ppm /(kHr) 1 /2
O p Te m p
‐40 to + 1 25 °C
L ine Re gu lat io n
22 0 pp m/V
V ou t D rift
7
THIS DRAWING CONTAINS INFORMATION THAT IS PROPRIETARY TO SPACE DYNAMICS LABORATORY (SDL). REFERENCE USURF BP 409.1
8
3
6
R 2 A D J   2 . 9 4 1 0
Vout
±10V ±0.075 V
Ripple Rejection
70
V ou t
1 .25 V ± 0 .18 %
PSSR
60 d B
P ackage
TSO T ‐ 6
6
4
4
5
Op Temp
‐40 to +85 °C
Output Resistance
140 O
Spec Temp
‐40 to +85 °C
Vout Ripple
50 mV
Vin
6.2 V
Iout
20 mA
4
Iq
1.15 mA
Cap Stability
> 10 uF
+-12V Generation
+2.5VA Generation
Vout
12 v
Package
SOIC ‐ 8
120 uVrms (to 10 Hz);
90 nV/(Hz)1/2 (1kHz)
> 10 uF
3
Spec Temp
‐40 to +85 °C
Load Regulation
3 mV (0.04%)
Vn (RTI)
3
Op Temp
‐40 to +125 °C
Line Regulation
12 mV (0.17%)
Cap Stability
Power Generation and Conditioning
5
1 mA
SOT 23 ‐ 5
Iq
180 uA
Iout
220 mA
Package
4/20/2010
LAST MODIFIED
T. Neilsen
ENGINEER
NEXT ASSY TITLE
NEXT ASSEMBLY
DICE
PROGRAM
Vin
±20V
VDO
0.19 V
Reverse Leakage I
2
OF
SCHEMATIC
SPACE DYNAMICS LABORATORY
1
INSTANCE
OF
INSTANCE NAME
1
SCH REV
OF
A
B
C
D
@HierPageTotal
TOTAL SCHEMATIC PAGES
@HierPageNumber
07 OF 07
SCH-INST SHEET
DRAWING NUMBER
DICE Science Payload Board
North Logan, Utah 84341
UTAH STATE UNIVERSITY RESEARCH FOUNDATION
141-0007
SCHEMATIC NAME
TITLE
Vout
±20V ±0.075 V
Ripple Rejection
46 dB
-7V Generation
+5VA Generation
2
89
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertising