Skip to main content

Full text of "Ground segment preparation for NPSAT1"

See other formats


Calhoun 

iniQiuiic^iul Ar{hiv« of tilt Mil vdl Poii^roduiit School 


Calhoun: The NPS Institutional Archive 
□Space Repository 



Theses and Dissertations 


1. Thesis and Dissertation Collection, all items 


2007-09 

Ground segment preparation for NPSATl 


Koerschner, Luke E. 

Monterey, California. Naval Postgraduate School 


http://hdl.handle.net/10945/3261 


Downloaded from NPS Archive: Calhoun 



DUDLEY 

KNOX 

LIBRARY 


htt p://w ww. n ps. e du/l ib ra ry 


Caflwuo is the Naval Postgraduate School's public access digital repository for 
research mate rials and institutiional putilicatiaos created by the NPS community. 
Calhoun is named for Professor of Mathematics Guy K. Caftiouo, NPS's first 
appointed — and putJlished — schoteily author. 

Dudley Knox Library / Naval Postgraduate School 
411 Dyer Road / 1 Univefsity Circle 
Monterey, California USA 93943 







NAVAL 

POSTGRADUATE 

SCHOOL 

MONTEREY, CALIFORNIA 


THESIS 


GROUND SEGMENT PREPARATION FOR NPSATl 

Thesis Advisor: 

by 

Luke Koerschner 

September 2007 

James A. Horning 

Second Reader: 


David Rigmaiden 


Approved for public release; distribution is unlimited 




THIS PAGE INTENTIONALLY LEET BLANK 



REPORT DOCUMENTATION PAGE 


FormApprovedOMBNo^ 0704-018^^^^ 
Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instruction, 
searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send 
comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden, to 
Washington headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 
22202-4302, and to the Office of Management and Budget, Paperwork Reduction Project (0704-0188) Washington DC 20503. 

I. AGENCY USE ONLY (Leave blank) 2. REPORT DATE 3. REPORT TYPE AND DATES COVERED 

September 2007 Master’s Thesis 

ATITLEANDSUBTITLE^GroundSegmentPreparationforNPSAT^^^^^^^^ 5. EUNDING NUMBERS 

6. AUTHOR(S) 

7. PEREORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PEREORMING ORGANIZATION 

Naval Postgraduate School REPORT NUMBER 

9. SPONSORING /MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORING/MONITORING 
N/A AGENCY REPORT NUMBER 

II. SUPPLEMENTARY NOTES The views expressed in this thesis are those of the author and do not reflect the official policy 

or position of the Department of Defense or the U.S. Government. _ 

12a. DISTRIBUTION / AVAII.ABIITTY STATEMENT 12b. DISTRIBUTION CODE 

Approved for public release; distribution is unlimited 

13. ABSTRACT (maximum 200 words) 

Most satellites rely on a ground control station to command their payloads and through which they download 
data from their payloads. The Naval Postgraduate School’s satellite (NPSATl) is no exception. The spacecraft’s 
payloads, which include the Coherent Electromagnetic Radio Tomography (CERTO), Langmuir probe, Configurable 
Eault Tolerant Processor (CETP), as well as the Visible Wavelength Imager (VISIM), all generate data that require 
collection on the ground through a radio frequency downlink. Telemetry from NPSATl’s unique attitude control 
system, which uses only MEMS angular rate sensors, magnetic coils, a magnetometer and a GPS could aid in the 
development of improved or more economical attitude control systems. The goal of this thesis is to ready the ground 
control segment for operation for collection of data from and command of NPSATl immediately after launch. 

Included is a description of the spacecraft to ground calculation, bidirectional, link budget and the operation 
and testing of the ground antenna pointing control system. Euture space systems students and faculty will use the 
ground control segment to harvest the data and reap the knowledge of the experiments that will orbit inside NPSATl. 
What better way to test the pointing of the antenna than to use it to track the Midshipman Space Technology 
Applications Research Program’s first satellite (MidSTARl). 


14. SUBJECT TERMS Ground Segment, NPSATl, MidSTARl 15. NUMBER OE 

PAGES 

_77_ 

16. PRICE CODE 

17. SECURITY 18. SECURITY 19. SECURITY 20. LIMITATION OE 

CLASSIEICATION OE CLASSIEICATION OE THIS CLASSIEICATION OE ABSTRACT 

REPORT PAGE ABSTRACT 

_ Unclassified _ Unclassified _ Unclassified _ UU _ 

NSN 7540-01-280-5500 Standard Form 298 (Rev. 2-89) 

Prescribed by ANSI Std. 239-18 


1 




























THIS PAGE INTENTIONALLY LEET BLANK 


11 



Approved for public release; distribution is unlimited 


GROUND SEGMENT PREPARATION FOR NPSATl 

Luke E. Koerschner 
Major, United States Army 
B.S., North Carolina State University, 1990 


Submitted in partial fulfillment of the 
requirements for the degree of 


MASTER OF SCIENCE IN SPACE SYSTEMS OPERATIONS 


from the 


NAVAL POSTGRADUATE SCHOOL 
September 2007 


Author: Luke Koerschner 


Approved by: James A. Homing 

Thesis Advisor 


David Rigmaiden 
Second Reader 


Professor Rudolf Panholzer 

Chairman, Space Systems Academic Group 



THIS PAGE INTENTIONALLY LEET BLANK 


IV 



ABSTRACT 


Most satellites rely on a ground control station to command their payloads and 
through which they download data from their payloads. The Naval Postgraduate School’s 
satellite (NPSATl) is no exception. The spacecraft’s payloads, which include the 
Coherent Electromagnetic Radio Tomography (CERTO), Eangmuir probe, Configurable 
Eault Tolerant Processor (CETP), as well as the Visible Wavelength Imager (VISIM), all 
generate data that require collection on the ground through a radio frequency downlink. 
Telemetry from NPSATl’s unique attitude control system, which uses only MEMS 
angular rate sensors, magnetic coils, a magnetometer and a GPS could aid in the 
development of improved or more economical attitude control systems. The goal of this 
thesis is to ready the ground control segment for operation for collection of data from and 
command of NPSATl immediately after launch. 

Included is a description of the spacecraft to ground calculation, bidirectional, 
link budget and the operation and testing of the ground antenna pointing control system. 
Euture space systems students and faculty will use the ground control segment to harvest 
the data and reap the knowledge of the experiments that will orbit inside NPSATl. What 
better way to test the pointing of the antenna than to use it to track the Midshipman Space 
Technology Applications Research Program’s first satellite (MidSTARl). 


V 



THIS PAGE INTENTIONALLY LEET BLANK 


VI 



TABLE OF CONTENTS 


I. INTRODUCTION.I 

A. STATEMENT OF THE PROBLEM.I 

B. NPSATI OVERVIEW.I 

II. NPSATI GROUND SEGMENT OVERVIEW.3 

A. GENERAL.3 

1. Frequencies.3 

2. NPSATI Antennas and Pointing.3 

3. NPSATI Passes.5 

B. COMMAND PATH (UPLINK AND DOWNLINK).5 

1. Computer and Software.5 

2. Digital Telemetry Receiver with Tracking.8 

3. Controller.9 

4. Enclosure.12 

III. NPSATI LINK BUDGET.15 

A. COMMUNICATIONS LINK BUDGET.15 

1. Margin.16 

2. Slant Range.16 

3. Bit Error Rate.18 

4. Antenna Gains.18 

a. Ground Antenna . 18 

b. NPSATI Antennas . 18 

5. Pointing Error.19 

6. Efficiency.21 

7. Noise Temperature.21 

8. Wavelength.22 

9. Beam Width.22 

10. Atmospheric and Rain Losses.23 

11. Free Space Path Loss.23 

12. Pointing Error Loss.24 

13. Effective Isotropic Radiated Power.24 

14. Propagation & Polarization Loss.25 

15. Link Budget.25 

B. TEST LINK BUDGET.28 

IV. TESTS, INSTALLATION, & CALIBRATION PROCEDURES.31 

A. FEED HORN.31 

B. TEST EQUIPMENT.36 

C. PROCEDURES.36 

I. Sources of Error.36 

a. Timing Errors . 36 

b. Satellite Orbital Ephemeris . 36 

vii 











































c. Antenna Location . 37 

d. Pointing Calibration . 37 

2. Initial Assembly and Checkout.38 

3. Slewing Initial Checks.38 

4. Aiming Point Tests.39 

a. Close Aiming Point . 39 

b. Medium Aiming Point . 39 

c. Distant Aiming Point Tests . 39 

D. WINDPROOFING.40 

V. COMMUNICATIONS CONTINGENCIES.51 

A. REDUNDANT GROUND STATIONS.51 

B. NPSATI CONTROL.52 

VI. CONCLUSION AND RECOMMENDATIONS.55 

LIST OF REFERENCES.57 

BIGLIOGRAPHY.59 

INITIAL DISTRIBUTION LIST.61 


viii 



















LIST OF FIGURES 


Figure 1. Horizon Fade.4 

Figure 2. Uplink Frequency Mixing.6 

Figure 3. Communications Block Diagram NFS ATI.7 

Figure 4. Connections on back of RC2800 PRK Dual Rack Mount Controller.10 

Figure 5. Antenna Deck Spanagel Hall.10 

Figure 6. Minimum & Maximum Elevations.12 

Figure 7. Controller Pointing Resolution.19 

Figures. Test Link.28 

Figure 9. Feed Horn Placement.32 

Figure 10. Feed Horn Mounted Inside Support Arms.33 

Figure 11. Feed Horn Signal Measurement.35 

Figure 12. Ballast Roof Mount.41 

Figure 13. Antenna Base.42 

Figure 14. Wind Loading Perpendicular to Antenna Aperture.44 

Figure 15. Wind Loading Parallel to Antenna Aperture.45 

Figure 16. NPSATl Communications Contingencies.53 



















THIS PAGE INTENTIONALLY LEET BLANK 


X 



LIST OF TABLES 


Table 1. NPSATl Uplink Budget, Short Form.15 

Table 2. Link Budget.27 

Table 3. Feed horn position final tests.34 

Table 4. 25G BRM Allowable Antenna Areas.49 

Table 5. Communications Parameters Comparison.51 








THIS PAGE INTENTIONALLY LEET BLANK 



ACKNOWLEDGMENTS 


I would like to thank my second reader, David Rigmaiden, for the hands on work 
that he did to make the antenna setup a reality. Thanks also to Professor Billy Smith of 
the U.S. Naval Academy for his invaluable assistance. The morning he spent showing me 
his ground control segment saved me weeks of work with the Nova software. LTC 
Lawrence Halbach directed my initial self directed study of the ground segment. Glenn 
Harrell’s work machining the feed horn mount and creating a measurement tool to check 
that the feed horn was in the center of the parabolic dish was much appreciated. Dr. 
James Newman had the idea of using an anemometer to park the antenna in the safe 
position during periods of high winds allowing us to use a commercial ballast mount. I 
would also like to thank Mr. David G. Brinker P.E., S.E. of the Rohn Products Division 
of Radian Communications Services Inc. for permitting me to publish Rohn figures in 
this thesis. MAJ Steve Moseley mounted the feed horn cover. Professor Rudolph 
Panholzer suggested moving the azimuth motor lower and closer to the ballast mount to 
improve stability following azimuth changes. Einally I would like to thank my thesis 
advisor, Jim Homing, for the software scripts he wrote for my thesis work with the 
controller and for his tireless proofreading of this thesis. 



THIS PAGE INTENTIONALLY LEET BLANK 


XIV 



I. 


INTRODUCTION 


A. STATEMENT OF THE PROBLEM 

Most communications satellites are in geostationary (GEO) orbit allowing 
terrestrial transmitters and reeeivers to point their antennas to fixed elevation and 
azimuths indefinitely. Other military dishes are designed to traek geosynchronous 
satellites by dithering toward the strongest signal strength to follow the minor ehanges in 
azimuth and elevation of their geosynehronous target. Many low earth orbiting (LEO) 
satellites relay data to GEO satellites whieh pass that information down to terrestrial 
receivers. NPSATl is a LEO satellite without the benefit of a relay satellite. Data from 
NPSATl experiments will only be available if telemetry ean be requested and reeeived 
by a ground segment. The ground antenna’s four degree spot beam will require a high 
degree of pointing aeeuraey from the eontroller. Other eonsiderations arise from a student 
satellite with a finite design life. With a limited lifespan it is desirable to establish 
eommunieations with the satellite immediately after launeh; ideally the ground eontrol 
segment should be fully operational prior to launeh. 

B. NPSATl OVERVIEW 

The student and faeulty built NPSATl is a LEO satellite whieh is designed to be a 
seeondary payload on a military or government launch. It incorporates an Evolved 
Expendable launeh vehiele Secondary Payload Adapter (ESPA) for mounting as a 
seeondary payload under the Orbital Express primary payload whieh was to be launehed 
on an Atlas V roeket. That launeh was missed, so arrangements are being made to launeh 
in 2009 on a Minotaur roeket with an ESPA seeondary payload suite. NPSATl uses 
eommereial, off the shelf, lithium ion batteries. The eylindrieal polygon shape of 
NPSATl has solar panels mounted on eaeh of its twelve faees, and will ineorporate on 
orbit testing of a triple junction solar cell design. The telemetry and eommand patch 
antenna design is deseribed in detail by Erel (2002). Testing of these mierostrip antennas 
was doeumented by Gokben (2003). Two sets of transmit and reeeive antennas are found 


1 



on the satellite. The primary transmit/receive (TX/RX) antennas are on the nadir pointing 
side of the satellite and the back up antennas are on the zenith pointing side. The transmit 
antenna is an elliptical patch measuring 5.66 cm across the short axis and 6.6 cm across 
the long axis. The receive antenna is a larger elliptical patch measuring 7.293 cm across 
the short axis and 8.509 cm across the long axis. Naval Research Laboratory (NRL) 
experimental payloads include the coherent electromagnetic radio tomography (CERTO), 
and a Langmuir probe. Naval Postgraduate School (NPS) experiments consist of a three 
axis micro-electromechanical (MEMS) rate sensor combined with magnetic coils to 
implement a magnetic attitude control test and a visible wavelength imager (VISM). The 
Solar cell Measurement System (SMS) experiment will test the new solar cell technology 
that will orbit on the satellite. Additionally, the CETP is a Naval Postgraduate School 
(NPS) designed payload that will orbit on NPSATl. A CETP is currently in use on 
MidSTARl. Results from MidSTARl telemetry show that the CEPT is experiencing 
single event upsets over the South Atlantic Anomaly (SAA) region. The SAA is a region 
in space over Brazil where the magnetosphere has a decrease in strength. The 
magnetosphere protects the Earth and LEO spacecraft from most solar high energy 
radiation particles which are strong enough to change a bit in a processor. More in depth 
reports of the CETP voting circuit operation will be included in NPSATl telemetry. 


2 



II. NPSATI GROUND SEGMENT OVERVIEW 


A. GENERAL 

The ground segment consists of those components on the ground that allow 
control of and communications with the spacecraft. The NPSATI ground segment 
includes a 10 foot parabolic dish antenna which is operated through a general purpose 
computer that sends commands to the controller which steps the azimuth/elevation 
motors. The uplink to NPSATI, and downlink from it are handled with two frequencies 
and those signals are passed through a modulator/demodulator (MODEM) between the 
computer and the antenna. An overview of the ground segments components is depicted 
in Figure 3. This section covers components of the ground segment in more detail. 

1. Frequencies 

A single ground relay antenna is used to transmit to the NPSATI at 
1767.565 MHz L-Band and receive transmissions from NPSATI at 2207.3 MHz S-band. 
Doppler shift is compensated for in the high and low frequency synthesizers. The 
separation of these two frequencies allows full duplex communications without 
interference between the two frequencies. 

2. NPSATI Antennas and Pointing 

Communications with NPSATI is not contingent upon the proper functioning of 
its Attitude Control Subsystem (ACS). Normally the ACS keeps transmit and receive 
antennas pointed toward nadir. The zenith pointing antennas act as a backup to the Nadir 
pointing antennas in the event the spacecraft looses pointing capability and begins to 
tumble. The tolerance of NPSATl’s nadir pointing via its ACS is estimated to be +/- 10 
degrees. NPSATI uses hemispherical patch antennas with half power beam widths 
determined by Erel (2002) to vary between 60.1 and 79.5 degrees at the uplink frequency 
and between 65.6 and 74.3 degrees at the downlink frequency (p. 42, 46). The average 
uplink half power beam width is 69.8 degrees, and the downlink average beam width is 

3 



69.5 degrees. The fact that they are omni directional allows them to transmit and received 
at much wider beam widths if the link is strong enough. For the purpose of calculating 
the link budget the rounded average beam width of 70 degrees was used for both uplink 
and downlink from NPSATl. 

NPSATl’s sister satellite MidSTARl, which was built for Naval Academy 
payloads, also contains a CFTP that was designed at NPS. The same design will be 
employed on NPSATl. MidSTARl does not have an attitude control system so it 
experiences roll fades. A roll fade is a drop in radio frequency signal strength that occurs 
when the satellite rolls from one omni directional antenna to another. Roll fades on 
MidSTARl can cause the temporary loss of communications when combined with 
pointing error losses. This is mentioned because NPSATl will also experience roll fades 
if its attitude control system fails. NPSATl’s attitude control system points it to nadir not 
directly toward the ground antenna. As a result of the nadir pointing antenna on NPSATl 
fades in signal strength will be experienced at low elevation angles even when the 
attitude control system is working. These fades can occur because the antennas on 
NPSATl will not always have the ground antenna in their half power beam width. This 
concept of “horizon fade” is best understood by Figure 1 which is conceptual and 
obviously not drawn to scale, because the four degree spot beam has an arc length of 149 
km at 10 degrees of elevation. 



4 










3. 


NPSATl Passes 


Depending on the inclination of the spacecraft’s orbit there should be at least four 
good opportunities to communicate with NPSATl each day. The original launch 
inclination would have yielded four daily satellite ground passes high enough above the 
horizon to permit time for downlink and uplink. Professor Smith of the Naval Academy 
has had good success with low grazing passes too, and if transmission and reception 
initiates below 10 degrees elevation then the system may have six usable passes daily. 
Presently MidSTARl’s orbit offers six good passes a day. Since MidSTARl has an 
inclination higher than the latitude in Monterey, CA it can pass directly over, or to the 
North of, the antenna at NPS. Those passes may be associated with loss of connectivity 
near zenith as the azimuth is changing faster than the antenna controller can receive 
commands and send status updates to the computer. This topic is discussed in more detail 
in this chapter (Section B3). 

B. COMMAND PATH (UPLINK AND DOWNLINK) 

I. Computer and Software 

One software component of the computer is the orbit propagator. Since orbital 
ephemeris is only down linked once a day, software must predict the satellite’s position 
over time with mathematical algorithms. The propagator that was tested for this thesis 
was embedded in Northern Lights Software’s Nova program. Satellite Toolkit (STK) was 
also used to propagate orbital data in early tests that used software written for an 
operating system shell to send commands to the controller. Both propagators worked well 
but the Nova software communicates directly to the controller while STK requires that 
the pertinent orbital data be exported and requires more programming. 

The computer with propagation software relays to the modulator de-modulator 
(MODEM) which mixes the intermediate frequency with the carrier frequency and feeds 
that communications signal through the low frequency synthesizer. The communications 
signal is sent back through the modem and out the antenna. Similarly signals received 
from NPSATl are sent through the MODEM to the high frequency synthesizer which 

5 



sends the signal back to the modem and on to the computer. Figure 2 illustrates the 
mixing of the intermediate frequency with the local oscillator for modulating the uplink 
frequency. 


Frequency 

Synthesizer 

Antenna ^577 i^hz 



Similarly the downlink frequency will be demodulated in a L3 software radio card 
that is on order. The L3 digital TT&C will eliminate the stand alone high frequency 
synthesizer from the architecture as will be described in this chapter (Section B2). Both 
synthesizers account for the Doppler shift of the moving satellite through programmed 
step routines. Doppler shift, the apparent increase in radio frequency of transmissions 
from an ascending satellite as it approaches the ground antenna and decrease in radio 
frequency of the frequency of the same transmissions from the satellite on its decent, is 
significant given the high velocities of spacecraft in the LEO regime. 

Other inputs to the computer include the weather station and may also include a 
digital camera, and a GPS. The weather station signals will send data to the computer 
through a serial port. The weather station signal of interest is the wind speed which will, 
in high winds, alert the computer to command the controller to elevate the dish antenna to 
a safe position. Digital cameras could be affixed to the antenna to provide visual 

6 











feedback to a remote computer being used to control the antenna over the campus 
network. The UHF antenna that was used for a previous NPS-built satellite had a light- 
sensitive diode mounted on it that allowed the ground controller to bore sight its Yagi- 
Uda antennas with the Sun. A GPS could be connected to the computer to keep the 
computer time synchronized with GPS time and consequently the satellite’s ephemeris 
time. 

The computer is the nerve center of the entire ground communications system. It 
is an Intel® Core™ 2 CPU 6300 @ 1.86GHz 1.86 GHz with 1 GB of RAM. It was 
ordered with multiple PCI card slots to accommodate the L3 communications card on 
order as well as the PCI card that allows it to connect to the Frequency Synthesizer. The 
current setup uses Northern Lights Nova software to communicate through a single serial 
cable to a M RC2800PRK dual rack mount controller. The controller is described in this 
chapter, this section, number 3. 

COMMUNICATIONS BLOCK DIAGRAM NPSAT1 


Full Duplex ^. 




ANTENNAAND COMMUNICATINS & CONTROL ENCLOSURE LOCATED ON 
ROOF OF SPANA3EL HALL36J60 DEG N LATTITUDE 121B8 DEG W 


Figure 3. Communications Block Diagram NPSATl 


7 









































2. Digital Telemetry Receiver with Tracking 


Delivery of the L3 Communications PCI-2070 Digital Telemetry Receiver with 
Tracking is not anticipated until after this thesis is written, but its capabilities will be 
discussed here. The L3 Technical Bulletin (2004) states the following: 

Capable of accepting RF input signals from -10 dBm to -VOdBm, the PCT 
2070 will receive the RF signal, condition and digitally demodulate FM, 

FSK, PM, BPSK, QPSK, OQPSK data. The image frequency bandwidth is 
programmable from 50 kHz to 30 MHz. The AFC (auto frequency control) 
tracking feature compensates for Doppler shift and other transmitter 
anomalies by using DSP algorithms to determine if the input spectrum is 
centered at the programmed center frequency. If the input spectrum is not 
symmetrical, the digital down converter is automatically stepped to track 
the input frequency. 

One of the biggest advantages of this digital telemetry receiver card is its tracking 
capability which allows it to automatically compensate for Doppler shift with its 
automatic frequency control (AFC). This card will eliminate the need for the high 
frequency synthesizer depicted in Figure 3. 

The card uses a Phase Lock Loop (PLL) in conjunction with Digital Signal 
Processing (DSP). The phase lock loop uses one or more traditional analog oscillators in 
combination with DSP. This card does not use Direct Digital Synthesis (DDS) in which 
the oscillator waveforms are generated in a processor. Some advantages of combining 
PLL technology with DSP is that the card is smaller and better at reducing spurious 
signals. Another advantage of this hybrid signal processing card is that its clock speed 
does not have to be multiples faster than the frequency of the generated waveform as is 
required in DDS. With a true DDS card, the clock speed of its processor would have to be 
at least twice the frequency because as described by Reed (2002) “The Nyquist sampling 
theorem limits the theoretical maximum attainable output lowpass frequency to half the 
clock frequency...” (p. 131). It is more likely that the clock speed of a comparable DDS 
card processor would have to be approximately 7 GHz (1.76 GHz (4)) because Reed 
(2002) states “it is customary to limit Ar to Fcik/4 to accommodate non-ideal analog 
filters.” (p. 135). Ar represents a frequency word. Essentially a DDS card of equal 


8 



capability would have to have a much larger processor that would consume more power, 
and radiate more heat, than the computer’s two 1.86 GHz CPUs. The interface of the card 
to the PC is via a PCI slot using a 32 bit PCI form factor. 

3. Controller 

The controller sends signals to two electric motors one for azimuth and the other 
for elevation. The controller pans across the heavens based on an open loop control 
scheme for elevation and azimuth of the dish antenna. In other words, once the elevation 
and azimuth are set off of a known point or celestial object the antenna may drift from 
those settings. The Naval Academy used the sun as the reference point for their antenna 
and they reset their azimuth and elevation calibration before every pass when possible. 
The motors send feedback to the controller for a closed loop control scheme. The 
controller has the antenna follow the predicted path of NPSATl during an overhead pass. 
One drawback of the Dual Rack mount controller is that it has a single 9600 baud serial 
port connection which has to receive separate commands for azimuth and elevation 
changes. The fastest update rate that can be used between the Nova software running on 
the computer and the controller is one second. Setting the update rate faster than that 
could result in the dropping of commands by the controller. Dropping commands occurs 
when the controller receives commands faster than it can execute them and subsequent 
commands are sent before the previous command has been executed, so commands are 
“dropped” by the controller. The RC2800PX/AZ and the RC2800PX/EL controllers were 
also purchased as spares. They allow the option of switching to separate elevation and 
azimuth controllers with individual serial port connections. Although the computer only 
has one 9-pin serial port, a USB port to serial cable adapter was tested with 
HyperTerminal to demonstrate that separate azimuth and elevation serial connections 
could be used. If separate controllers are used the CPU will have to send commands to 
both of them simultaneously through multiple RS-232 serial connection achieving more 
responsive antenna control. The connections to the dual rack mount controller are shown 


9 



below in Figure 4. The black and white wires are connections to the azimuth and 
elevation motors and the orange and blue wires connect to the pulse switches which give 
motion feedback to the controller. 



Figure 4. Connections on back of RC2800 PRK Dual Rack Mount Controller 


4. Ground Antenna 

A mesh parabolic ground antenna is located on the roof of Spanagel Hall (8th 
floor) at the Naval Postgraduate School in Monterey, CA. 36.595 degree North Latitude 
by 121.875 degree West Latitude. Figure 5 is a sketch of the location of the antenna in 
relation to other equipment on that deck. 


1 Side 

8th Deck BLDG 232 Spanagel Hall 



Figure 5. Antenna Deck Spanagel Hall 


10 

















































The uplink beam width of the ground antenna is approximately four degrees and 
is a function of the frequency and antenna diameter. The ground antenna’s downlink 
beam width is approximately three degrees. The pointing accuracy of the ground antenna 
must be less than or equal to two and a half degrees to maintain the down link as will be 
discussed in Chapter III. A two and a half degree error translates into an error arc length 
of 24 kilometers while pointing at NFS ATI 560 km directly overhead. The maximum 
path loss is 162.7 dB on the uplink and 164.6 dB on the downlink as will be calculated in 
Chapter III. The antenna is a 3.048 meter (10 feet) parabolic dish type reflector. The 
antenna reflects signals transmitted from NFS ATI onto the feed horn. The feed horn also 
radiates the parabolic reflector with signals transmitted to NFSATl. A minimum transmit 
elevation over land of 10 degrees may be used to mitigate the chance of interfering with 
ground receivers. Over the Monterey Bay it should be safe to transmit and receive at zero 
degrees elevation because there are fewer ship borne transmitters and receivers that are at 
risk of interference on the bay than on land. 

An antenna limitation is that it cannot slew through more than 374 degrees of 
azimuth (14 degrees of overlap) or more than 90 degrees of elevation. Because of these 
limitations the antenna will not be able to continuously follow a satellite that passes 
directly overhead. Once the elevation of the antenna reaches 90 degrees the antenna 
would have to rotate through 180 degrees of azimuth before following the satellite as it 
descended on the through the eastern horizon. The time required to rotate would result in 
a temporary loss of connectivity. Antennas that have to be slewed at their maximum 
elevation to follow the satellite on its descending pass are said to have a “keyhole” in Air 
Force jargon because one has to turn the antenna just like a key. Figure 6 is helpful in 
visualizing this keyhole where the antenna azimuth has to be rotated once the maximum 
dish elevation is reached. 


11 



MINIMUM DISH ANTENNA ELEVATION OVER LAND 10 DEGREES 



MAXIMUM DISH ANTENNA ELEVATION 90 DEGREES 

Figure 6. Minimum & Maximum Elevations 

4. Enclosure 

The outdoor enclosure has a single door with two lockable handles, which both 
latch. The enclosure is 24” wide, 20” deep, and 30” high. DDE is the manufacturer and 
model PSOD-302429FT was purchased. The purpose of the outdoor enclosure is to 
protect the computer, controller, frequency synthesizer, uninterruptible power supply 
(UPS), and transceiver card from the elements. It is located as close as possible to the 

feed horn, on the antenna base, to minimize the line losses between the feed horn and the 

12 









transceiver card. The UPS depicted in Figure 3 will need to power all of the equipment in 
the enclosure and the azimuth and elevation motors for twenty minutes. None of the 
satellite passes will be longer than twenty minutes, so if the AC power is lost at the 
beginning of a satellite pass the system will still have enough battery power to track and 
communicate through the entire pass. 


13 



THIS PAGE INTENTIONALLY LEET BLANK 


14 



III. NPSATl LINK BUDGET 


A. COMMUNICATIONS LINK BUDGET 

This chapter seeks to clarify the calculations used for the generation of the link 
budget. The link budget is a cumulative calculation of transmitter to receiver gains and 
losses which determines if the link is strong enough for reliable communications. 
Bidirectional communications mean that this link budget must be calculated from the 
satellite to the ground receiver and from the ground receiver to the satellite. 

The short form of the uplink budget to NPSATl is depicted in Table 1, and a 
carrier to thermal noise ratio is calculated using the format in Gordon and Morgan’s 
Table 2.5 (1993) (p. 44). 


Receiving earth station location: Monterey, CA 

Uplink frequency/u: 1.76757 GHz 

Transmit earth station antenna diameter: 3.048 m 

Satellite: NPSATl _ Uplink beam: 4 degee spot beam 

Parameter _ Sign Value Units Section 

Earth Station 

Power at the antenna for 6.99 dBW 

P* = 5_ W/carrier 

Transmit antenna gain G + 32.43 dBi 4.a. 


39.42 dBW 13. 


162.69 dB 11. 


-21.6 dBi/K 


Earth station EIRP 
Earth to Satellite 

Eree space path loss L for 5u = 1840 km 
Satellite 

Satellite G/Ts,u -i- 


Carrier/thermal noise Cu/Tu 
l/k {k = Boltzmann’s constant) 


-144.87 dBW/K 
-I- 228.6 dB(W/Hz K)'‘ 


Cu/kTu 


83.731 dBHz 


Table 1. NPSATl Uplink Budget, Short Eorm 


15 










This short format has section numbers which correspond to the calculations that 
follow in this chapter. The drawback of this short format is that it does not include losses 
for the pointing errors of both the ground and spacecraft antennas. The short form is 
useful though because the high carrier to thermal noise value of 87.73 dB Hz indicates 
that the link should have adequate strength. This value will be compared with the carrier 
to thermal noise from the long uplink budget. The long link budget is a more detailed 
spreadsheet that is developed with information from the calculations that follow in this 
chapter. 

1. Margin 

How much margin is sufficient for reliable communications? The guidance given 
by Space Mission Analysis and Design (SMAD) edited by Larson & Wertz (1999) is to 
“Adjust the input parameters until the margin is at least 3 dB greater than the estimate 
value for rain degradation, depending on confidence in the parameter estimates.” (p. 568). 
Rainfall is sparse in Monterey and outages during the handful of days annually with 
heavy precipitation are acceptable. Since Gaussian Minimum Shift Keying is being used 
a value of 9.6 dB is extracted from Larson and Wertz’s Table 13-11 as the minimum 
received energy per bit over noise-density (Eb/No) (p 562). 

2. Slant Range 

The slant range is calculated by knowing the maximum altitude of NPSATl and 
the minimum elevation of the ground antenna. Presently the launch parameters of 
NPSATl are unknown so // = 560 km will be used because it was the maximum altitude 
of the Orbital Express Eaunch. The 10 degree minimum elevation that is imposed on the 
antenna to reduce interference from and to ground stations is also used. Work began with 
equation (5-24) from Earson & Wertz (1999) (p.l 13). 

Re 

sin p = cos A,o = - ; Or [Equation 3-1] 

Re + H 


sin p =- ; So 

Re^H 


16 



6371.0003fcm . 

Sin p = -.-.sin 0 = 0.919204 

6371 .0003km + 560km 


p = 66.8099° 


Using Equation (5-26b) from Larson and Wertz (p. 113) 


sin t] = cos a sin p 


sin 7] = cos(10°)0.919204 


sin;; = 0.905239 


[Equation 3-2] 


.-.;; = 64.8555° 

Using equation (5-27) from Larson and Wertz (p. 113)... 
X + a = 00° 

64.8555°+;L +10° =90° 

.•.;L = 15.1445° 


[Equation 3-3] 


Einally slant range, D, is solved with Larson and Wertz’s equation (5-28) (p. 113). 


D = i?£(sin X / sin rf) 

D = 6371.0003Msinl5.1445° 70.905239) = 1838.69/tm 


[Equation 3-4] 


In the interest of simplicity, this is rounded up to 1840 km. Since this study began 
NFS ATI missed the Orbital Express launch. Euture launch opportunities include a 
Minotaur with an orbital altitude of between 600 and 700 km. Eor H=700 km the above 
calculations are performed to obtain D= 2155 km. 


17 



3. 


Bit Error Rate 


The bit error rate (BER) is the probability of a single bit being erroneous. A 
probability of a bit error of 10'^ was chosen because that is a typical BER that is tolerable 
for telemetry and command signals. Using figure 13-9 of Earson & Wertz, with this 
probability of error, it is found that Gaussian Minimum Shift Keying (OMSK) yields a 
required energy per bit over noise ratio (Eb/No) of 9.6 dB (p. 561). With OMSK the 
spectrum utilization of 1 represents good use of spectrum. The bit rate for both uplink to 
and downlink from NFS ATI is 115 kbps. 

4. Antenna Gains 

a. Ground Antenna 

The aperture of the dish is 10 feet which is multiplied by 0.03048 to 
convert to 3.048 meters. The uplink gain is calculated using Gordon & Morgan’s 
equation (6.5) (p. 140). 

G = 20 log loD -r 20 log lo/ -i- lOlog lorj -i- 20 A{dBi) 

G = 20logio(3.048) +20log 10(1.76757)+ 101ogio(0.55) + 20.4(dB/) [Equation 3-5] 

G = 32A{dBi) 


Similarly, a downlink gain is calculated with the above equation using the 
2.207 GHz downlink frequency and the result is 31.0 dB. 

b. NPSATl Antennas 

The gain of the patch antennas on NPSATl can be calculated using the 

same formula. 

G = 20 log loD -r 20 log lo/ -r lOlog lorj -i- 20 A{dBi) 

G = 20logio(0.0612) + 20logio(2.2073) + lOlogio(0.90) + 20A{dBi) [Equation 3-5] 

G = 2.6{dBi) 


18 



The receive elliptical patch antennas on NFS ATI are slightly larger with 
an average diameter of 0.0764 meters. The receive frequency of 1.76757 GHz must also 
be used in the above equation to calculate a receive antenna gain of 0.4 dB. 

These values were checked with a modified version of the antenna gain 
equation from Larson and Wertz (13-18b) (p. 555), and yielded identical values. It should 
be noted that dBi refers to isotropic decibels. 


5. Pointing Error 


0.5 de^ee 
Elevation 


% -► 

0.25 degree 
Azimuth 


Figure 7. Controller Pointing Resolution 


The pointing error of the ground antenna is more difficult to estimate. Controller 
tests revealed that the elevation drive only makes changes of one degree or more and that 
the azimuth rotor makes changes in half degree increments. The best pointing accuracy 
that can be hoped for is half of the hypotenuse of the pointing resolution, because the 
controller must wait for a 0.5 degree increase or decrease in elevation to change the 


19 



elevation of the dish and it waits for a 0.25 degree change in azimuth to bump the 
azimuth to the next closest azimuth increment. Consequently, the best possible pointing 
accuracy is the hypotenuse of the two values depicted in Figure 7, or 0.559017 degrees. 
That is the resolution of the M controller but the software being tested does not 
command the controller to adjust the antenna unless there is a change in elevation or 
azimuth of a degree. The Nova software defaulted to 1.8 degrees of azimuth or elevation 
difference before commanding a change, but this was lowered to one degree. The 
hypotenuse of 1 degree of both azimuth and elevation is the square root of two or 1.41 
degrees. This does not mean that the best pointing accuracy is 1.41 degrees because the 
Nova software can be set to lead the satellite. The Nova software allows setting of the 
rotator to lead the satellite in either time or degrees. By leading the satellite the 
hypotenuse of 1.41 is split which gives the best theoretical pointing accuracy of 0.7 
degrees. Professor Smith of the Naval Academy uses the degree settings to lead an 
ascending portion by +1 degree and then changes the settings at zenith so that the 
elevation controller leads the satellite on the descending pass by -1 degree. The satellite is 
not really being led by the antenna. Instead the goal is to move the antenna in concert 
with the satellite passage. By setting a lead time of a few seconds the ground antenna 
adjusts while the satellite is moving so that it will not constantly be 1.41 degrees behind 
the satellite. Timing inaccuracies and direction errors reduce the 0.7 degree theoretical 
pointing accuracy but it is estimated that the total pointing error will be at least one 
degree. At elevations closer to zenith, above 50 degrees, the azimuth changes very 
quickly and the pointing accuracy decreases, because of the one second update rate of the 
single serial port connection. Because of this a 2 degree ground antenna pointing 
accuracy will be used in the link budget calculations. Final implementation of the 
controller may eliminate the use of the Nova software, and instead use a custom 
programmed antenna tracking routine. Still, the Nova software is an excellent program 
for testing of the ground antenna while programmers at NPS are focused on completing 
the NPSATl software. The pointing error of NPSATl towards nadir is estimated to be 10 
degrees. 


20 



6 . 


Efficiency 


The ground antenna transmit feed efficiency of 55% is garnered from the range of 
typical values. Gordon and Morgan (1993) state “The typical range of antenna efficiency 
is 0.4 to 0.8 and a common approximation is 0.55” (p. 36). The same value is used for 
the receive efficiency of the antenna. NFS ATI transmit and receive efficiencies of 90% 
were extracted from tests by Erel (2002) who depicts his results in his Figures 25 and 32 
(p. 40, 44). 

7. Noise Temperature 

The ground receiver noise temperature is the combination of cosmic, galactic and 
troposphere noise as the antenna is pointed skyward. As elevation increases the 
troposheric sources of noise decrease so the minimum elevation of 10 degrees points the 
ground antenna above most of the noise radiation from the Earth. Gordon & Morgan’s 
FIGURE 9.8 is entered with the 1.76 GHz transmit frequency and the minimum 
inclination of 10 degrees and yields a receiver noise temperature of 20 Kelvin (p. 206). 
This agrees with the summation of the maximum galactic noise temperature of 6 Kelvin 
from FIGURE 9.6 and the tropospheric noise temperature of 12 Kelvin from Figure 9.7 
(20) Kelvin « (12 + 6) Kelvin). This value assumes that the Sun and Moon are not in the 
side lobes or main lobe and that there are no terrestrial sources of interference in the back 
lobe or side lobes (p. 204,205). 

The noise temperature that is used in the link budget is a total system noise 
temperature. It includes transmitter noise, noise from both antennas, and the receiver 
noise. Values are taken from Larson & Wertz’s (1999) Table 13-10, and a brief 
description of their table is given (p. 558). 

Table 13-10 shows typical noise temperatures for satellite systems using 
uncooled receivers. When a narrow satellite-antenna beam looks at Earth, 
the uplink antenna noise temperature is the temperature of the Earth, about 
290K. In the future improvements in design of low-noise amplifiers will 
reduce the receiver noise figures, especially at higher frequencies. 


21 



The system noise temperatures from the table are 135 Kelvin for the downlink 
and 614 Kelvin for the uplink. 

8. Wavelength 

The uplink frequency of 1.76757 GHz and the downlink frequency of 2.2073 GHz 
are easily converted to wavelengths by dividing the speed of light, 299,792,458 m/s, by 
them. 

A = c/f 

Xu = (2999,792,458m / s) / (1767570000//z) 

[Equation 3-6] 

Xd = (299,792,458m / 2201300000Hz) 

Xu = 0.1696m, /Id = 0.1358m 

9. Beam Width 

The beam width of the ground antenna represents the cross section in degrees of 
the strongest part of the signal radiating to or from the antenna. It is the angle of the beam 
in degrees on the edge of which the signal experiences a 3 dB, or 50%, loss. Using 
equation 6.7 from Gordon & Morgan (1993) where/and D representing the frequency in 
GHz and diameter in meters of the antenna respectively 3.9 degrees is calculated for the 
uplink frequency (p. 143). 

^3 = 21/yD(deg) 

03 = 21/(1.76757G//z)(3.048m) [Equation 3-7] 

^3 = 3.89787° 

The receive beam width at the downlink frequency, 3.12178 degrees, is also 
calculated by inserting 2.207 GHz in the above equation. 

Empirical data was used to determine the 70 degree half power beam width of the 
NFS ATI hemispherical patch antennas. This value was averaged from Erel’s (2002) 
Eigures 28, 29, 35 & 36 (p.42, 46). 


22 



10. Atmospheric and Rain Losses 


Atmospheric and rain losses are difficult to determine at these relatively low 
frequencies, Gordon and Morgan (1993) state the following: 

The 1- to 10-GHz range is already used extensively by both terrestrial 
microwave and satellite services. Although the noise level and attenuation 
are lower than those at the higher frequencies, the potential for 
interference from terrestrial point-to-point services has limited the location 
of earth stations, (p. 179). 

Interference is a greater concern at these frequencies because atmospheric and 
rain losses are negligible at 1.76 to 2.0 GHz. Half a decibel could be subtracted from the 
margin of the link budget spreadsheet to account for losses during periods of rain. 
Fortunately the school’s proximity to the Monterey Bay allows an antenna site that can 
acquire the satellite over the ocean mitigating terrestrial interference. Placing the antenna 
on the tallest building on campus combined with the proposed minimum elevation over 
land of 10 degrees mitigates terrestrial interference on the descending half of satellite 
passes. 


11. Free Space Path Loss 

Free space path loss is the loss due to the slant range or distance between the 
transmitter and receiver. Slant range in kilometers and frequency in GHz are used for 
equation (2.30) from Gordon and Morgan (1993) to calculate a loss of 162.7 dB as is 
shown below (p. 39): 

L = 20 log loS + 20 log 10 / + 92.45( JB) 

L = 201ogiol840km + 201ogiol.76757G//z + 92.45(JB) . . on 

[Eqnatron 3-8] 

L = (65.2964 + 4.94753 + 92.45)( JB) 

L = 162.694 JB 


23 



The downlink path loss is almost identical to the uplink path loss because the only 
value that changes is the frequency resulting in a loss of 164.62 dB. The path loss will 
decrease as the slant range decreases and be at a minimum at the highest elevation during 
a pass. 


12. Pointing Error Loss 

Pointing error loss is related to both the pointing accuracy, e, and the half power 
beam width, 83 . Larson & Wertz (1999) use 0 (p. 556) where Gordon & Morgan (1993) 
use 03 . 

Le = -\l{eie)^dB{\l>-2\) 

Le = -12(2/3.89787)^dB [Equation 3-9] 

L9 =-3.15926% dB 

This calculation is also performed for the downlink to the ground antenna which 
has a narrower receive beam width due to the higher frequency and 4.809148 dB of loss 
is the result. 

NPSATl antennas are more forgiving of pointing errors due to the omni 
directional properties of the patch antennas. 

Le = -12(10 / 72.0) dB [Equation 3-9] 

Le = -0.23 dB 


13. Effective Isotropic Radiated Power 

Effective Isotropic Radiated Power (EIRP) combines the gain of the antenna with 
its power. Gordon & Morgan (1993) define it as the sum of the antenna gain in dB and 
the transmitter power in dB (p. 36). 

EIRP = I01ogioP -t Gt (dBW) (2.21) [Equation 3-10] 

EIRP = 10 log io(5 Watts) + (32.4 Gain -4.16 Losses pointing & line) (dBW) 

EIRP = 35.27 (dBW) 

Notice that the antenna pointing loss and line loss of one dB is subtracted from 

the antenna gain. An effective EIRP of 35.27 dB is obtained. 

24 



The same equation is used for NPSATl’s EIRP and because it has much less 
transmitting power (1 W) and antenna gain (2.56-0.23 dB) the result is 2.33 dB 
(remembering that 1 W = 0 dB). From 2.33 the line loss of 1 dB is subtracted leaving an 
effective EIRP of 1.33 dB 

14. Propagation & Polarization Loss 

Propagation loss is taken from Earson and Wertz’s figure (13-10) from which a 
0.3 dB loss is extracted (p. 563). It may include losses from transmitting through the 
plastic feed horn cover which is about the same thickness of a radome. In an example of a 
satellite using almost identical frequencies Earson and Wertz state “I would also add a 
loss of 0.3 dB to account for polarization mismatch for large ground antennas. Using a 
radome adds another 1 dB loss.” (p. 568). For now, the loss of the plastic cover will be 
neglected because it may be removed during operation. 

Polarization loss is attributed to the circular polarization of the signal being 
imperfectly matched with the polarization of the feed horn on the ground antenna or 
receive antenna on NPSATl. Two feed horns are available, one for right hand circular 
polarization (RHCP) and the other for left hand circular polarization (EHCP). The feed 
horn used can be chosen based on the orientation or polarization of the satellite 
transmission to minimize polarization losses. 

15. Link Budget 

The long form of the link budget equation is given by Larson and Wertz at 
equation (13-13) in decibels as shown (p. 554). 

Eh/ No = P + Li + Gt + Lpr + L^ + La + Gr + 228.6-10logTs-lQ\ogR [Equation 3-11] 

P is the transmitter’s effective power in dB. L/ is the line loss. Gt is transmit 
antenna gain less its pointing loss. Lpr is the pointing loss of the receive antenna. Ls is the 
free space path loss. La is the propagation and polarization loss. Gr is the receive antenna 
gain. Ts is the system noise temperature. R is the data rate. Table 2 summarizes the 
calculations in the link budget. 


25 



Analysis of the link budget in Table 2 shows that pointing accuracy of the ground 
antenna is critical. The downlink is lost when the pointing error is greater than 2.53 
degrees, and the uplink is lost when the pointing error exceeds 4.19 degrees. The 
downlink is more sensitive to pointing error because of the smaller half power beam 
widths at the higher frequency of 2.207 GHz. This is corroborated by the operational 
experience of the sister ground antenna at the Naval Academy. The downlink from 
MidSTARl is lost before the uplink is lost. 


26 



Transmitter 

Transmit Frequency (f) 

Power Budget Allocation in watts (Pt) 
Transmitter Efficiency {qdc) 
Available Transmit Power (Pta) 
Transmitter Power in Decibels (Pt) 
Transmitter Line Loss (LI) 
Transmit Antenna Beamwidth (9bt) 
Transmit Antenna Pointing Error (Get) 
Assumed Antenna Efficiency (q) 
Transmit Antenna Diameter (Dt) 

PeakTransmit Antenna Gain (Gpt) 

Transmit Antenna Pointing Loss (LGt) 
Transmit Antenna Gain (Gt) 
Equiv. Isotropic Rad. Pwr. (EIRP) 
Spatial Geometry 

Sat Xmt Ant Max Cvg Ang (q”) 
Earrth Central Angle (A) 
EGA (A) in degrees 
Slant Range (S) 
Coverage footprint Diameter 
Coverage footprint in NM 

Power Flux Density (PFD) 

PFD/4kHz band 

Space (path) Loss (Ls) 
Propagation & Polarization Loss (La) 


TM Down 

CMD to NPSAT1 


Section, Reference and or Equation 

2.207 

1.76757 

Ghz 

II. A. 1. 

12.05 

56.18 

•watts 

25% and 75% of RF Pwr Bdgt, from "Power 
Budget" page 

0.083 

0.089 



1.00015 

5.00002 

watts 


0.000651393 

6.989717415 

dBw 


-1 

-1 

dB 


70 

3.89787 

deg 

III. A. 9. 

10 

2 

deg 

III. A. 5. 

0.9 

0.55 


III. A. 6. SMAD Figure of Merit p 553 = 0.55 

0.0612 

3.048002523 

m 

1. B. ;ll. B. 4. 

2.563500201 

32.44146578 

dB 

III. A. 4. a. & b. G = - 

159.59+20*LOG(Dt)+20*LOG(f~GHz)+10*LOG(q) 

0.244897959 

-3.159268491 

dB 

III. A. 14. L0 = -12*(0et/0bt)''2 

2.318602241 

29.28219729 

dB 

Gpt+Lpt 

1.319253634 

35.2719147 

dBw 

Pt+LI+Gt 


0.610865238 

0.034015333 

rad 

q° = O.5*0bt 

0.079120784 

0.003734866 

rad 

A = 180-{q-acos[sin(q)/(Re/Ro)]+90} 

4.533287012 

0.213992082 

degrees 

1840 

1840 

km 

III. A. 2. S = [(Ro-Re*cos(A))''2 + (Re*sin(A))''2]''.5 

2110.761286 

125.1522865 

km 

plane geometry estimate 

1139.720332 

67.57685316 

NM 


134.9692015 

-101.0165404 

dB 

PFD = EIRP/(4pS''2) 

170.9898014 

-137.0371403 

dB 

PFD/4000 

164.6224031 

-162.6938889 

dB 

III. A. 11. Ls = 147.55-20log(S~m)-20log(f~Hz) 

-0.3 

-0.3 

dB 

III. A. 15. SMAD Table 13-13 


Assumed Antenna Efficiency (q) = 

0.55 

0.9 

Receiver Antenna Diameter (Dr) = 

3.048 

0.0764 n 

Peak Receiver Antenna Gain (Gpr) = 

34.36997281 

0.423026505 c 

Receiver Antenna Beamwidth (0br) = 

3.121777879 

70 c 

Receiver Antenna Pointing Error (0er) = 

2 

10 

Receiver Antenna Pointing Loss (L0r) = 
Receiver Antenna Gain (Gr) = 

4.925351613 

29.4446212 

-0.244897959 c 

0.178128545 c 


III. A. 6. SMAD Figure of Merit p 553 

III. A. 4. G = - 

159.59+20*LOG(Dt)+20*LOG(f~GHz)+10*LOG(q) 

III. A. 9. 0 = 21/(D*f) 

III. A. 5. 

III. A. 12. Le = -12*(0et/ebt)''2 
Gpr+Lpr 


System Noise Temperature (Ts) 
Data Rate (R) 

Eb/No(1) 

Carrier-to-Noise Density Ratio (C/No) 
Bit Error Rate (BER) 
Required Eb/No (2) 
Implementation Loss (3) 


135 

614 

K 

1.15E+05 

1.15E+05 

bps 

17.60580401 

22.32259427 

dB 

68.21278241 

72.92957268 

dB-Hz 

1.00E-05 

1.00E-05 


9.6 

9.6 

dB 

-2 

-2 

dB 


III. A. 7. SMAD Table 13-10 

SMAD pg. 385, Table 11-19 

Eb/No = EIRP+Lpr+Ls+La+Gr+228.6-10LogTs- 

lOLogR 

C/No = Eb/No+ 1G*logR 

III. A. 3. 

III. A. 1. SMAD Table 13-11 
Estimate 


Margin = 6.005804007 10.72259427 dB III. A. 1. (1)-(2)+(3) 

Table 2. Link Budget 


27 











B. TEST LINK BUDGET 


The test link budget is described here because it is the link that is used in the next 
chapter to find the optimal position of the feed horn and ensure that component gains and 
losses correspond to their expected values. 

The measured line loss in the cable connecting the signal generator to the feed 
horn was -1.28 dB. The transmitting test antenna gain was calculated in an earlier test by 
pointing two test antennas at each other in the lab on a short five meter range. The test 
antenna gain was calculated as -1-6.93 dB with this empirical test. The initial test slant 
range, which is the distance between the transmit antenna and the aperture of the 
parabolic antenna, is 10 meters. Gordon & Morgan (1993) give us equation (2.30) (p. 39). 

L = 20 log io5 + 20 log 10 / + 92.45(dB) 

L = 201ogio(10/1000)km + 201ogiol.76757G//z + 92.45(dB) 

[Equation 3-11] 

L = (-40 + 4.94753 + 92.45)(dB) 

L = 57.398 dB 



cable loss 


Figure 8. Test Link. 


28 







Since the measured value of 4.28 dB is 4.01 dB less than that shown in Figure 8 
as the expected or “perfect efficiency” output of 8.29 dB, attenuation is probably causing 
a loss of 4.01 dB. Signal blockage caused by the clutter in the short range may be 
attenuating the signal resulting in a 4.01 dB loss. The test setup used had the parabolic 
antenna mounted on a plinth in the lab facing out the window toward the sidewalk. The 
test antenna horn was sited on a tripod on the sidewalk outside and pointed through the 
window at the parabolic antenna. Clutter consisted of the glass that the signal was sent 
through as well as the window frames and pillar that blocked the edges of the dish. 
Another possible source of the loss may be attributed to a near field test. The radio 
waves may not be parallel when they reach the parabolic reflector if this was a near field. 
The fact that the radio waves may not have been parallel means that their reflections were 
not as focused as they would be in a far field test. The far field distance for the same test 
horn antenna was calculated by Gokben (1996) as greater than or equal to 3.652 m (p. 18) 
using the downlink wavelength and the uplink wavelength for these tests. The longer 
wavelength of 0.1696 meters yields a lower calculated far field distance using equation 
(9-51) from Stutzman and Thiele (1998) (p. 413). 


IjPf _ 2(0.4984m)" 
/I 0.1696m 


[Equation 3-12] 


The 10 meter range is above the far field range so losses must be attributed to blockage 
and hence attenuation of the signal. A test outside should be performed to check this 
theory. Using the same equation the for the 10 foot (3.048m) parabolic reflector a far 
field range of 110 meters is calculated. 


2(3.048m)" 

0.1696m 


109.6m 


[Equation 3-12] 


So, a much longer range will be required to test transmissions from the ground antenna. 


29 



THIS PAGE INTENTIONALLY LEET BLANK 


30 



IV. TESTS, INSTALLATION, & CALIBRATION PROCEDURES 


This chapter describes the calculations and test procedures for the setup of the 
ground antenna. It begins with the calculation and test of the feed horn placement and is 
followed by procedures for bore sighting the antenna. This chapter concludes with the 
description of the wind loading calculations. Some of the calibration procedures were not 
implemented yet and should be taken as recommendations. 

A. FEED HORN 

Using equation (6.1) from Gordon and Morgan (1993) gives the relationship used 
for calculating the focus of a parabola (p. 138). The calculated focus gives a starting point 
for the placement of the feed horn. 

F = D^— [Equation 3-13] 

16d 

Where D = the diameter of the dish and d = the depth of the dish. A 10 foot 
diameter was measured and a 21.25 inch depth, obtaining a 42.35 inches calculation for 
the focal point. The distance calculated is depicted in Figure 9 as the arrow drawn from 
the center of the reflector to the feed horn. 


31 




Figure 9. Feed Horn Placement 

Other ways to check the focal point include placing a reflective mirror on the 
surface of the dish and using a laser pointer to confirm that light is reflected to the feed 
horn. The diameter of the feed horn mounting ring is six inches. 

Tests were conducted to confirm the location of the feed horn by placing a 
transmitter on the sidewalk between the buildings to radiate to the dish inside the satellite 
lab. The feed horn was moved to find the spot with the most gain. 1767.56 MHz was the 
frequency used to transmit from the sidewalk through the window into the satellite 
laboratory. First the distance between the transmitting test horn and the dish was 
measured. The measurement of 10 meters was used to calculate the expected gain, as 
shown in Figure 8, of 8.29 dB. 

Initial checks of the feed horn gain were made to determine whether the feed horn 
needed to be moved closer to the parabolic reflector or further away from it. Checks were 
made to ensure that the feed horn was perpendicular to the reflected signal from the dish 
antenna by shimming the feed horn. When it was shimmed to the left side gain improved 


32 




to -0.6 dBm. Where dBm are decibels of the power divided by one mW (1 x 10'^ Watts). 
Shimming the feed horn not only changes the angle of its aperture but also moves it 
closer to the reflector, so both sides were shimmed and the gain improved to -0.16 dBm. 
Realizing that the gain was improving, the feed horn was moved closer to the dish, and 
received signal strength improved to 0.46 dBm. When the feed horn was moved all the 
way in a signal of 1.12 dBm was measured. The original feed horn mounting ring did not 
permit the feed horn to be moved closer to the dish, but it appeared that the gain would 
improve if it could be further adjusted. At this point, it was apparent that the feed horn 
mounting plate would have to be re-machined so that the feed horn could be mounted 
closer to the parabolic reflector. 


Tests began the next day with the feed horn mount bolted to the inside of the 
support arms allowing feed horn adjustment closer to the reflector as shown in Figure 10. 



Spacer 

Washers 


Figure 10. Feed Horn Mounted Inside Support Arms 


33 






The washers shown in Figure 10 were abandoned and replaced with springs for faster 
adjustment. The first test with new feed horn position immediately yielded 4.2 dBm. 
When the feed horn was moved out to the edge of the mount a signal of 3.8 dBm was 
observed. The feed horn was then moved half way back to the starting point with shims 
and tape and an output of 3.99 dBm was obtained. This indicated that the signal gain 
increased toward the starting point. When the feed horn was moved in further with 
smaller nuts and tape, a 4.01 dBm signal was observed. At that point it was decided to 
again modify the mounting bracket to allow further adjustment toward the reflector. 

On May 8, 2007 the feed horn mounting plate was redesigned by Glenn Harrell. 
Excess aluminum was machined away from the new plate to minimize its blockage of the 
reflector aperture. When the new machined mounting ring for feed horn was installed the 
first reading was 4.2 dBm. Table 3 depicts the sequence of the tests on the installed feed 
horn mounting plate. 


Step 

Direction Moved 

Signal 

1 


4.2 dBm 

2 

In 

4.1 dBm 

3 

All the way out 

4.09 dBm 

4 

In to 5/16 inches from plate 

3.89 dBm 

5 

In to 3/8 inches from plate 

3.93 dBm 

6 

In to 5/8 inches from plate 

4.08 dBm 

7 

In to % inches from plate 

4.2 dBm 

8 

In to 7/8 inches from plate 

4.22 dBm 

9 

In to 1 inch from plate 

4.25 dBm 

10 

In to 1 & 1/16 inch from plate 

4.19 dBm 

11 

Back to 1 inch from plate 

4.25 dBm 


Table 3. Feed horn position final tests 


34 







Conveniently, the rubber gasket on the feed horn was aligned flush with the back 
of the mount at this optimal position. The distance between the center of the parabolic 
reflector and the feed horn was measured as 43 & 7/8” +/- 1/32”. A graphical depiction of 
the output signal at the best measured signal position is shown in Figure 11. 



Figure 11. Feed Horn Signal Measurement 


35 











B. TEST EQUIPMENT 


Two standard gain signal horns 

Engineering Development Units using the same transmitters as those in NPSATl 

Cellular telephones 

Vehicle 

EDU Battery Supply 
Binoculars 
Handheld GPS 
Antenna 

Spectrum Analyzer 
Surveyor’s Tripod 
Maps USGS 
Google Earth 

C. PROCEDURES 

1. Sources of Error 

Sources of pointing error for the link between ground station and satellite, and 
methods of mitigating them are discussed in this section because the bore sighting 
procedures that are described later in this chapter minimize the largest source of error. 

a. Timing Errors 

If the computer time is wrong then it will track the satellite either early or 
late. This is eliminated at the Naval Academy because their computer time synchronizes 
with GPS time over the internet. By using either the internet or by using a dedicated GPS 
receiver to bring a time value directly into the computer through a serial port connection 
timing errors will be minimized. 

b. Satellite Orbital Ephemeris 

The satellite orbital parameters and the time associated with its location in 

the orbit is defined in the orbital ephemeris. Ephemeris is automatically downloaded with 

36 



the Nova software on a daily basis from http://www.space-track.org/perl/login.pl . 
MidSTARl was loaded in “My Favorites” and the NOVA software was set to download 
updates from “My Favorites” on a daily basis. When NFS ATI is launched it will have to 
be added to “My Favorites” on this website. The “Navy Fence” is a line of VHF 
transmitters and receivers that is now operated by the USAF to collect and update orbital 
ephemeris data on satellites as they traverse the 33“^ parallel of the US. 

c. Antenna Location 

The Nova software has a Monterey, CA observer location as 36.60 
degrees North and 121.88 degrees West. Two GPS readings of the antenna location were 
taken on 20 June 07 with a circular probable error of -I-/-10 meters. The following two 
positions were received and recorded: 

Northing I = 36°35'42.03" 

Westing. = \lV5Tl%.6Q" 

Northing! = 36°35'41.96" 

Westing! = 12r52'2SJ2'' 

Averaging these results in a location of 36.59499 degrees North and 121.87460 degrees 
West. Rounding and truncating these to two decimal places results in 36.59 degrees 
North and 121.97 degrees West. The least significant digit of both the latitude and 
longitude was different than the default for Monterey in the Nova database. A separate 
observer called Tower should be created to minimize the antenna location error. Nova 
does not include the elevation of the ground antenna but STK does, so STK should 
generate slightly more precise antenna pointing data by accounting for the height of the 
roof of Spanagel Hall above Monterey. 

d. Pointing Calibration 

With a spot beam smaller than four degrees any error in the calibration of 
the antenna will be added to the inherent pointing limitations, so the procedures listed in 
the sections below address how the antenna will be aligned to true azimuths and 
elevations. The Naval Academy uses the Sun to bore sight their antenna’s azimuth and 

37 



elevation before each pass but there are limitations to this approach. They experienced 
degraded azimuth calibration when the Sun is near zenith because the noise signal from 
the Sun is less sensitive to changes in azimuth near zenith and more sensitive to changes 
in elevation. Conversely, when it is on the horizon the Sun is an excellent azimuth 
reference. In the Academy’s use of the Sun a bullet camera was placed on the feed horn 
and pointed toward the parabolic reflector so the shadow of the feed horn could be 
viewed to check their alignment with the Sun. They also could site on the sun through 
cloud cover by dithering the antenna to the highest noise signal on a receiver. Another 
drawback of the Sun is that it is a moving target. The Naval Academy placed a 
transmitter at a known location on their roof to improve their calibration of azimuth while 
the Sun is high in the sky. A transmitter should also be placed on the roof of Spanagel 
Hall at a know direction and as far from the antenna as possible so that there is a known 
directional calibration point regardless of the sun’s position. 

2. Initial Assembly and Checkout 

A standard gain antenna horn will be used to transmit through the glass to the 
antenna. The parabolic dish assembly area offered only a range though the glass window 
in the laboratory. At each phase of the test the expected loss will be calculated and 
compared to the actual loss. The antenna will then be dithered to check for pointing 
accuracy. 


3. Slewing Initial Checks 

The remaining procedures in this section are proposed and were not completed at 
the time this thesis was published. The antenna will then be moved to the roof of 
Spanagel Hall where it will be slewed to a known point on the roof of Spanagel Hall. 
Again, a standard gain horn will be used to transmit to the antenna. 


38 



4. 


Aiming Point Tests 


a. Close Aiming Point 

The standard gain antenna horn will be used to transmit to the antenna 
across the roof of Spanagel Hall. At this time the weatherproof enclosure will be 
integrated and testing of remote access of the computer controller in the rack over the 
Internet will begin. 

b. Medium Aiming Point 

The antenna will be slewed to a know point on Hilltop Field at the 
Presidio of Monterey. Although there is not a visual line of sight to this field there is a 
sufficient radio frequency signal through the forest between the two points. The standard 
gain horn will be used to transmit to the antenna and the spectrum analyzer will be used 
to check pointing accuracy. Voice communication between the aiming point and the 
Naval Postgraduate School’s ground control can be established with cellular telephones. 

c. Distant Aiming Point Tests 

The antenna will be slewed to at least two distant know aiming points. 
These can include Mt. Toro, the lighthouse in Santa Cruz, and Freemont’s Peak. Visible 
aiming points can also be used to align the antenna visually. The direction to the nearest 
smokestack at Moss Landing was measured as 20 degrees east of north and the antenna 
calibration can be checked by slewing it to that direction and visually sighting it. The 
Santa Cruz Mountains provide locations that are several miles away from the antenna, but 
high enough to permit line of sight radio signal reception. The ground antenna is 
calibrated by placing an L-band directional transmitter at two distant know points. The 
azimuths to the distant aiming points will be determined by using United States 
Geological Survey Maps and Global Positioning Receivers. Hiking may be required to 
the distant aiming points to orient and activate the directional transmitter. The directional 
transmitter will be visually sighted toward Naval Postgraduate School and a clear day 
after frontal passage should offer the best visibility for sighting the calibration antenna. 


39 



Optical enhancement, like binoculars and scopes, can be used to improve the sighting 
accuracy. The direction should be checked with a calculated direction to a known point. 
When the calibration antenna is sighted and powered the ground control team will point 
the antenna to the distant aiming point by loading the calculated direction and elevation. 
Reception signal strength will be measured and characterized to find the azimuth and 
elevation corresponding to the best signal reception. It is important to find a second 
location that has a different elevation than the first location to check the elevation and 
deflection slewing accuracy of the controller. This could be difficult because local terrain 
may present few options 10 degrees or more above Spanagel Hall. One approach may be 
to suppress the elevation of the elevation motor so that it can be tested through a range of 
motion on local terrain, and then reset it to an operational elevation after testing. 

D. WINDPROOFING 

Wind concerns were voiced by attendees of the 24 April,2007 briefing on the 
antenna. One constraint is that drilling into the roof of Spanagel Hall is not allowed, 
because its roof was recently weather sealed and is under warranty. Initially, the plan was 
to weld a custom ballast mount out of existing components but it was decided that buying 
commercially available Rhon antenna base was an easier solution. The Rohn ballast roof 
mount and short antenna base are depicted in Figures 12 and 13 respectively from 
Antenna Solutions and Control Inc. (1999). The azimuth rotator was mounted on the 
accessory shelf depicted in Figure 13 for an August 2007 demonstration of the assembled 
components. Attendees observed that the torque from azimuth motor was twisting the 
short base and causing the antenna to momentarily shake after a change in azimuth. 
Professor Panholzer suggested moving the accessory shelf to the bottom of the short base 
so that it attaches to the three mounts that are bolted to the ballast roof mount. This would 
mount the accessory shelf on a more rigid portion of the assembly and reduce the twisting 
of the antenna base. A much longer piece of pipe connecting the azimuth motor through 
the thrust bearing to the antenna will be required for this change. 


40 




Figure 12. Ballast Roof Mount 


41 




Thrust bearing 



Bearing 

Plate 


0’11 


Short 

Base 



Figure 13. Antenna Base 


42 































Professor Panholzer asked for the weight of the antenna, because he was 
concerned about the load placed on the roof. The roof of Spanagel Hall is rated for 200 
lbs per square foot. The disassembled antenna was weighed piecemeal. The elevation 
motor mount and dish ring mount assembly weighed 134 lbs. The antenna base assembly 
consisting of the components depicted in Figure 13 and the azimuth rotator weighs 
79 lbs. The four antenna quadrants weighed 16.5 lbs and the middle plate weighed 2.95 
lbs without mounting hardware so their total weight was rounded up to 20 lbs. The four 
feed horn arms weight five lbs with hardware. The feed horn and mounting ring four lbs. 
So the total weight of the antenna from the antenna base up is 237 lbs. The antenna 
weight is insignificant when compared to the weight of the ballast which will be 
discussed later in this section. 

Assuming that the weight is evenly distributed across the 40 cinder blocks in the 
ballast roof mount which cover 38.4 square feet a 7,680 lbs load can be imposed on the 
ballast frame. This does not mean that that much ballast can be placed on the roof 
because the wind will add to the load of the ballast frame opposite the wind. An 
assumption is made that to account for wind loading the weight of the antenna and ballast 
should be half of the roof limit or 3,840 lbs. This is because at the instant before the wind 
topples the antenna the downward force on the roof opposite the wind, depicted in Figure 
14, will equal approximately half the weight of the ballast. 

Doctor Newman suggested that an anemometer be placed on the roof that is 
connected to the computer. A signal from the anemometer will elevate the dish in a safe 
of position of 90 degrees when the winds exceed a speed that is dangerous to the dish or 
the mount. This concept is depicted in Figures 14 and 15. 


43 



Wind 


Ballast 



Figure 14. Wind Loading Perpendicular to Antenna Aperture 


44 

























Figure 15. Wind Loading Parallel to Antenna Aperture 


One concern of a longer mast is that the wind pushing on the dish has a longer 
moment arm so the platform may be less stable in severe winds. An advantage of the 
longer mast is that it would keep the dish above the PSOD-302420 outdoor enclosure. 
This outdoor enclosure is 30” high, 24” wide and 20” deep and can be pipe mounted on 
the side of the truss for the mast. The advantage of placing the transmitter and receiver 
enclosure close to the antenna is the reduction of line losses. It may be useful to estimate 
the force exerted on the mast since it is the weakest component in the assembly. 
Additionally, the force applied to the dish can be converted to a downward force on the 
opposing ballast frame to verify that roof loading tolerances are not exceeded. The final 
consideration is that the upwind ballast is not lifted by the force of the wind on the 
antenna. 


45 























Fortunately the Rohn mount that was purchased had data sheets available for wind 
loading. Bob Broadston said that a 150 mile per hour (130 knot) wind survival 
requirement was used for the roof of Spanagel Hall, because the fastest winds recorded in 
Monterey were 100 miles per hour. Assuming that the 10 foot dish is solid then it has an 
area of or 78.54 square feet. Ricardo (2001) states, “Mesh dishes act as solid dishes 
at about fifty miles per hour though will still experience approximately 40% less force 
than a solid dish.” Even if that area is reduced by 40%, because it is a mesh dish, the area 
of 47.12 square feet is off the table that Rhon provides for their mount. 18 square feet is 
the largest antenna area shown on Table 4 and that is only 38% of the calculated 
maximum area of the antenna. Wind survival is calculated based on attaching an 
anemometer to the box which elevates the dish to 90 degrees when the wind achieves 
30 nautical miles per hour (knots) or more. The depth of the dish is 21.25 inches, and that 
is used for this calculation. The area of the dish when elevated to 90 degrees is the area of 
the crescent shape exposed to the wind. Adding % of an inch to the depth to take into 
account the depth of the ribbing that gives the dish its strength increases the dish depth to 
22 inches. Using an equation from Beer & Johnston for the area of a parabola where 
h = 22 inches and a = 60 inches. The area of the parabola = 4ah/3 = 12.2 ft (p. 175). 

Calculating the weight of the ballast is done by multiplying the weight of each 
cinder block by the number of blocks. A high density cinder block from the existing 
mounts on the roof was weighed. The empty weight was 34.5 lbs. Since each cage holds 
ten cinder blocks on each of it, four sides could be loaded with 345 lbs per side or 
1380 lbs with a single layer of these blocks. The dimensions of the blocks are 
8” by 8” by 16” and cement was poured in the cinder block holes to fabricate heavier 
solid blocks. This increased the individual weight of each cinder block to 60 lbs. Forty of 
these cinder blocks weigh 2400 lbs. The sum of the calculated edge of dish area of 12.2 
ft plus the area of the box steel section, which acts as the middle connector between the 
elevation mechanism and the antenna mounting ring, gives the total area. The box section 
“sail” area of up to 1.63 ft^ plus the area of the stowed dish gives a total area of 13.83 ft^. 
That is rounded up to 14 ft to account for the areas of the elevation motor and ring. 
Table 4, from Antenna Systems and Solutions Inc. (1999) is entered in the effective 


46 



2 

projected area row of 14 ft and the entry weight of 2400 lbs is interpolated between the 
ballast column values of 2250 lbs and 2500 lbs (p. MS-4). Following that to the wind 
velocity, Vs, for one section at both 2250 and 2500 lbs wind speeds of 158 and 166 mph 
are extracted. Simple linear interpolation is used to calculate a wind velocity. Vs, of 
162.8 mph, which is rounded down to 162 mph. Table 4 is the manufacturer’s table for 
loading the antenna base. This table is being used conservatively because the short base 
that NFS procured is shorter than the section of 12.4’ that is shown in the table. The 
measured height is 8.5’ and even with the antenna fully elevated the top lip of the antenna 
will only be 10.3’ high. The total weight of the 2400 lbs of ballast, the 237 lbs antenna 
assembly, and the 200 lbs ballast mount is 2,837 allowing approximately 1000 lbs of 
margin before the 3,840 lbs roof limit. The enclosure assembly weight is expected to be 
less than 100 lbs. 


47 




29G BftM ALLOWtAetC ANTENNA AREAS 


FFFECTIWE 

BALLAST 

ZERO 

Vi 

Vs 

Vma* AT CENTROlQ OF 

PROJECTED 

(LBS) 

VELOCiTY 

ONE 

TWO 

PROJECTED AREA (MPH) 

AREA 


LOAD 

SECTION 

SECTIONS 

CENTROID OF ANTENMA EPA 

(EW) 


(PSF) 

(MPM) 

(MPH) 



(FT*) 



11*12 4 FT 

h •224FT 

I»■t24#(1s«a) 

h ■224«(2 Wd) 


SCO 

SO 

91 

78 

72 

48 


750 

75 

112 

94 

08 

58 


1000 

10O 

129 

108 

101 

67 


1250 

12 S 

144 

121 

113 

75 

8 

1500 

150 

158 

132 

122 

81 


1750 

175 

171 

143 

129 

86 


3000 

200 

103 

153 

138 

90 


2250 

225 

104 

162 

142 

95 


2500 

250 

204 

171 

149 

99 


2750 

27 5 

214 

179 

154 

103 


3000 

909 


167 

157 

104 


500 

50 

a4 

72 

66 

44 t 


750 

75 

103 

89 

80 

54 


1000 

10 0 

119 

102 

93 

63 


1250 

12 5 

133 

114 

104 

70 

10 

1500 

ISO 

146 

125 

112 

76 


1750 

175 

150 

135 

118 

80 


2000 

200 

160 

145 

124 

84 


2250 

225 

179 

153 

1» 

88 • 


2500 

250 

180 

162 

136 

92 


2750 

27 5 

198 

169 

141 

95 


3000 

300 

207 

177 

144 

97 


500 

SO 

79 

69 

61 

42 


750 

75 

97 

84 

74 

51 


1000 

10O 

112 

97 

66 

59 


1250 

12 5 

125 

100 

96 

66 

12 

1500 

ISO 

137 

119 

104 

71 


1750 

175 

148 

126 

110 

75 


2000 

200 

158 

137 

115 

79 


2250 

225 

167 

146 

121 

83 


2500 

250 

176 

154 

128 

86 


2750 

27.5 

105 

161 

131 

90 


3000 

300 

193 

168 

133 

91 

1 

1 1 

NO 

.1>RFV. DESCkiPTiUN 

tCtATE 

♦ ftfv&y 

<9CHKO BY 

♦APPD BY 


THIS DRAWING IS THE PROf ERTY OE ROHNINOUSTRCS. INC ITIS NOT TO BE REPROOOCEO. 

COPIED OR TRACED IN WHOLE OR IN PART WITHOUT OUR WRITTEN CONSENT. 


DRAWN BY: MSJ DATE 8/ 78/99 ROHN InduStriW. Inc. 

C Hfcct^DBY, W A_yjoy? 

AggPgNGLjgj_OATE^ a/3l/y9 25G BRM ALLOWABLE 

APPO SALES DATE ^ 

FILE NUMBER 4t83S0B 

UKAWINO NUMBER AP9?094-2OF 


9 


ANTENNA AREAS 













3SG BRM ALLOWABLE ANTENNA AREAS 


; EFFECTIVE 
PROJECTED 
AREA 

(FT') 

ballast 

ILBS) 

ZERO 

VELOCITY 

LOAD 

(PSF) 

Vs 

ONE 

SECTION 
(MPH) 
11-12 4 FT 

Vs 

TWO 

SECTIONS 

(MPH) 

h«224FT 

VrwATCENTROlOOf' I 

PROJECTED AREA (MPH) 

CENtROlO OF antennaFPa 

^ h ■ 1 ? 4“fi (1 tact) I ti ■ 2^4 II (2 teeip 

14 

! 

KW 

750 

1000 

1?S0 

1500 

1750 

7000 

2250 

2500 

2750 

3000 

sO 

75 

lOO 

125 

150 

17 5 

200 

225 

250 

27 5 

300 

t4 

SI 

105 

117 

120 

139 

149 

158 

186 

174 

182 

68 

60 

93 

104 

114 

123 

131 

139 

147 

154 

161 

57 

70 

80 

90 

97 

103 

loe 

113 

118 

123 

125 

_ . 

35^ 

48 

58 

62 

87 

71 

75 

78 

81 

65 

66 1 

1 

i 

16 

500 

7S0 

1000 

12S0 

1500 

1760 

2000 

2250 

2500 

2750 

3000 

50 

75 

100 

US 

150 

175 

200 

225 

250 

27 5 

300 

70 

S6 

100 

111 

122 

132 

141 

149 

157 

165 

172 

63 

77 

69 

99 

109 

118 

126 

133 

141 

147 

154 


37 

46 

53 

59 

84 

67 

71 

74 

77 

60 

82 


500 

50 

67 

60 

51 

98 


750 

75 

62 

T4 

62 

44 


lOOO 

10 0 

95 

88 

72 

50 


1250 

12 5 

106 

96 

SI 

56 

18 

1500 

150 

118 

106 

S7 

61 


1750 

17 5 

126 

113 

92 

64 


2000 

200 

134 

121 

97 

68 


2250 

225 

142 

126 

101 

71 


2500 

25 0 

150 

135 

106 

74 


2750 

27 5 

157 

142 


77 


9000 

300 

164 

148 

112 

78 


NO ' fftEsrWSCRifTKSi fDATE tCHkt) ffr fAPPO 

BY 


THIS DRAWMG IS THE PROPERTY OF ROHN INDUSTRIES. INC IT IS MOT TO BE REPRODUCED, 
COPIED OR TRACED IN WHOLE OR IN PART WITHOUT OUR WRITTEN CONSENT 


DRAWN BY MSJ 

DATE 0/20/99 

ROHN Industrios, Inc. 

Checked By ns 

0AT£ aO(V99 



DATE 0/31/99 

25G BRM ALLOWABLE 




ANTENNA AREAS 

DRAWING NUMBER 

A9920S4-3 0F9 


M$-4 


Table 4. 25G BRM Allowable Antenna Areas. 



































THIS PAGE INTENTIONALLY LEFT BLANK 


50 



V. COMMUNICATIONS CONTINGENCIES 


A. REDUNDANT GROUND STATIONS 

The design of similar ground stations at both the Naval Academy and NFS 
enables either to act as a back up ground station if the ground station owning the satellite 
is inoperable. A comparison of the communication parameters of MidSTARl and 
NFS ATI is shown below in Table 5. 


MidSTARl 

• Inclination = 46.02 ° 
overhead passes 

• Frequencies 1.767GHz 
up/ 2.2022 GHz down 

• Phase Modulation 

• GMSK Gaussian 
Minimum Shift Keying 

• 64 kbps up/100 kbps 
down 

• Beacon with some 
Telemetry every 4 
minutes 


NPSAT1 

• i = 57 ° ? minotaur 2009 

• 1.767 GHz up/2.027 GHz 
down 

• Frequency Modulation 

• GMSK 

• 115.2 kbps 

• Beacon only in fail mode 


Table 5. Communications Farameters Comparison 


51 



B. NPSATl CONTROL 


What happens if communications with NPSATl are lost? NPSATl uses a 
software controlled radio, so one possibility is for the satellite to step down its bandwidth 
if communications are spotty. NPSATl’s default bandwidth is 115.2 kbps as shown in 
Table 5 and this number drives the speed at which data is transferred to and from the 
spacecraft. Dynamically lowering the bandwidth from 115.2 kbps increases the margin in 
the link budget, but lowers the data rate. This is analogous to speaking more slowly on a 
cellular telephone if the person on the other end cannot understand what is being said. 
Because the coding of the field programmable gate array (FPGA) that controls 
communications has not been finalized the exact band width that will be used, if there are 
communications problems, has yet to be determined. 

Another point of failure for communications with NPSATl is the on/off routine 
for the antennas which is based on calculations of when the satellite is over the Monterey, 
CA area. To conserve electricity the satellite receiver is only activated when its GPS and 
orbit propagator predict it to be over the Monterey, CA area. If the NPSATl controller 
reboots and does not have orbit position awareness then the “No Nav” branch in 
Figure 16 is followed. This causes the receive antenna to turn on for thirty seconds of 
every two minute period. The receiver remains on if the ground antenna is successfully 
transmitting to it. If the onboard GPS has failed, then the priority after a reboot should be 
to upload new predictions to NPSATl making it easier to acquire with the ground 
antenna. This process is displayed in the block diagram in Figure 16. 


52 



ACS & Comm Process 



Figure 16. 


cmd[comm=nadir] 

NFS ATI Communications Contingencies. 


53 

















THIS PAGE INTENTIONALLY LEFT BLANK 


54 



VI. CONCLUSION AND RECOMMENDATIONS 


The link budget proves that the downlink is very sensitive to the pointing 
accuracy of the ground antenna. The single most important conclusion of this work is that 
a more accurate ground antenna pointing control scheme should be implemented before 
NFS ATI is launched. It is no surprise that the link with MidSTAR is intermittent, 
because the antenna controller and software at the Naval Academy is identical to that 
tested here. Pointing accuracy could be improved by splitting the elevation control and 
azimuth control with the separate elevation and azimuth controllers. This would facilitate 
two individual 9600 baud RS-232 serial connections. Dr. Michael Owen (personal 
communication, July 19, 2007) of Northern Lights Software estimated that their Nova 
software could be modified to support two separate controllers with eight hours of their 
programming time which was quoted at a rate of $100 an hour. Alternatively, use of a 
high-fidelity SGP4 algorithm would allow more precise control of the antenna rotors with 
in house configuration of custom software. 

The large parabolic ground segment dish antenna can be installed on a 
commercial mount when wind speed data sent to the computer has the controller place 
the dish in a safe configuration. 

Software radio features of both NPSATl and the demodulator card in the ground 
computer enhance our capabilities. The software controlled radio in NPSATl allows it to 
lower the data rate as a communications contingency and the ground computer software 
radio card compensates for Doppler shift with AFC. The PCI software radio card on 
order will allow communications with both MidSTARl and NPSATl. Frequency 
modulation (FM) is employed on NPSATl but the MidSTARl uses phase modulation 
(PM). Fortunately, the PCI radio card that was ordered can demodulate either FM or PM. 

The ability to track the beacon on MidSTARl will prove that the antenna control 
system will function with NPSATl. Another recommendation is to conduct far field tests 
with the parabolic dish antenna to obtain an empirical value for its efficiency which could 
be used in the link budget instead of the estimate. 


55 



A follow on study should be completed to detail ground segment operation. This 
would establish the procedures for sending commands and receiving telemetry from 
NPSATl. MidSTAR should be used, if it is still operational, to test sending commands 
and receiving telemetry from a satellite. Work with MidSTAR 1 could be used as a basis 
for the future operation of the NPSATl ground segment. 


56 



LIST OF REFERENCES 


Antenna Systems and Solutions Inc. 25BRGM Assembly Drawing; 25BRGM Ballast 
Chart. (1999). Retrieved 1 September 2007 from: 
http://w w w. antennasystems .com/rohn/mounts .html . 

Erel, Mahmut. Design of Microstrip Patch Antenna for the NPSATl, (2002) Monterey, 
CA: Naval Postgraduate School. 

Gokben, Hahn. Prototype Fabrication and Measurements of Uplink and Downlink 
Microstrip Patch Antennas for NP SAT-1, (2003) Monterey, CA: Naval 
Postgraduate School. 

Gordon, Gary; Morgan, Walter. Principles of Communications Satellite, (1993) Hoboken, 
NJ: John Wiley & Sons, Inc. 

Larson, Wiley; Wertz, James. Space Mission Analysis and Design Third Edition, (1999) 
El Segundo, CA: Microcosm Press. 

L3 Communications Telemetry-East (2004). Technical Bulletin PCI-2070. Retrieved 31 
August 2007 from http://www.l-3com.eom/TE/PDE/PCI-2070.pdf . 

Reed, Jeffrey H. Software Radio a Modern Approach to Radio Engineering, (2002) 

Upper Saddle River, NJ: Prentice Hall PTR. 

Ricardo, R. Installation Discussion (2001) Retrieved 1 September 2007 from: 
http://www.geo-orbit.org/sizepgs/gmdpole.htnil . 

Sakoda, Daniel; Homing, James. Overview oftheNPS Spacecraft Architecture and 
Technology Demonstration Satellite, NPSATl (2002) Proceedings of the 16* 
Annual AIAA/USU Conference on Small Satellites. Logan, Utah. 


57 





THIS PAGE INTENTIONALLY LEFT BLANK 


58 



BIGLIOGRAPHY 


Meeks, M. Ball, John. Hull, Anthony. The Pointing Calibration of the Haystack Antenna, 
November (1968) IEEE Transactions on Antennas and Propagation, VOL. AP-16 
N0.6. 

Rodin, A. Sekido. M. Pulsar VLBI Observations, (2002) Proceedings of the 6th European 
VLBI Network Symposium. Bonn, Germany. 

Tomasi, Wayne. Electronic Communications Systems Fundamental Through Advanced, 
fifth edition, (2004), Upper Saddle River, NJ: Pearson Education, Inc. 


59 



THIS PAGE INTENTIONALLY LEET BLANK 


60 



INITIAL DISTRIBUTION LIST 


1. Defense Technical Information Center 
Ft. Belvoir, Virginia 

2. Dudley Knox Library 
Naval Postgraduate School 
Monterey, California 

3. Professor Rudolf Panholzer 
Naval Postgraduate School 
Monterey, California 

4. Professor William Smith 

U. S. Naval Academy Physics Department 
Annapolis, Maryland 

5. Daniel Sakoda 

Naval Postgraduate School 
Monterey, California 

6. James Horning 

Naval Postgraduate School 
Monterey, California 

7. David Rigmaiden 

Naval Postgraduate School 
Monterey, California 

8. Professor Alan Ross 
Naval Postgraduate School 
Monterey, California 

9. Ronald Aikins 

Naval Postgraduate School 
Monterey, California 


61