Home

Receive Sensitivity Characterization of the PolySat Satellite

Contents

1. e e e 70 23221 Overview f Testing pU a OU e e hg pae ER PES AE ERE UE Rr DE Pell 70 5 2 2 CDH Testing Measuring the Receive Sensitivity essere nennen eene 71 5 2 3 CDH Rev 4 Measured Sensitivity Testing eese eee enne 71 5 2 4 CDH Rev 5 Measured Sensitivity Testing esses eene eene entente trennen teens 72 3 2 9 CP CDE Flight Candid te e n ERE e PERRO EP PEE 74 3 2 7 YaesuiE 1 647 Transcelver ee e eo RS Sade ER ERE SIL EE SRE 75 CHAPTER 6 ADDITIONAL TESTING eeeeeeeee eee en seen ene tn sins tn stata seta suse sesto sees suse tasse sensn aen 78 6 1 RECEIVER NOISE FLOOR COMPARISON eese en natns theta tns tein strata sihi tasa tain strata sessanta sata 78 6 1 1 Receive Line of CC1000 Noise Floor at RX TX Frequency esee nennen eene 82 6 1 2 Receive Line of CC1000 Broadband Noise Floor eene nenne 85 6 1 3 Receive Line of CC1000 Noise Floor over 70 CM Amateur Band seen 68 6 1 4 Receive Line of CC1000 Noise Floor Characteristics eese eene 89 6 2 TESTING FOR REDUCED COMM SENSITIVITY FROM SWITCHING NOISE eee eene 90 6 2 1 Switching Noise from DC DC Converters eese eene eene nennen rennen trennen nenne en rennen 90 6 2 2 Switching Noise from PC B ssc coii eno eosam coves dete
2. Connection to back of VNA ENR dB pot Cold Noise Source Figure 25 The test setup used to measure the Noise Figure of the Low Noise Amplifier 22 39 The measured noise figure Table 12 remains close to the datasheets typical values up through 1 GHz but increases by almost 2 dB at 2 GHz At 3 GHz the noise figure was measured at 7 298 dB However the measured noise figure remains at approximately 3 dB throughout the 430 450 MHz amateur radio band which is very close to the typical values listed in the ERA 3 datasheet 20 Noise Figure Typical Values datasheet Measured Values Table 12 Typical and measured Noise Figures from 0 010 3 GHz 40 NOISE FIGURE NOISE FIGURE MEAS CH 1 NOISE FIG REFERENCE PLANE B BBBB mm LOG HRGNITUDE REF 8 888 dB 8 588 dB DIV 1 43 MHz 3 161 dB 2 437 288888 MHz 3 817 dB 3 445 288888 MHz 3 842 dB 4 45 MHz 3 033 dB MARKER TO PEAK MORE gt H 430 000 B8B MHz 458 888 B8B MHz Figure 26 Measured noise figure over the 430 450 MHz range 4 7 HP8566A Spectrum Analyzer System calibration consists of characterizing the signal reaching the DUT To do this a splitter divides the signal and directs it to the DUT and a Spectrum Analyzer The LNA attached to the front of the Spectrum Analyzer increases sensitivity With a Resolution Bandwidth RBW of 3 kHz the Disp
3. 1W Splitter Attenuation Variable Attenuation Yaesu FT 847 Radio Faraday Cage Oscilloscope DC Power Sensitivity setup for finding the threshold of sensitivity of the PolySat COMM system The LNA is disconnected from the power splitter because the satellite response exceeds the maximum input power of the LNA and would cause permanent damage Record the resting RSSI value of COMMA using the oscilloscope hit quick meas for the average of Chl If the Yaesu FT 847 is outputting around 0 dBm of power start with the attenuation at 44dB Otherwise if the radio output 20 dBm start at 24dB Send 0201 at each of the 12 attenuation settings and record the RSSI values Make sure to measure the RSSI during the command since the CC1000 monitors the RSSI in real time Also note at which setting the satellite stops responding Adjust the attenuation to the previous setting before the satellite stopped responding Now increasing the attenuation in 1dB increments send 0201 commands until the satellite stops responding Make sure that you stop at an attenuation which the satellite CONSISTENTLY responds This is the threshold of sensitivity 124 Sensitivity Response Threshold of Sensitivity Actual Signal Successful Response dBm No Response Attenuator Setting dB The last successfully decoded command indicated by satellite response is the sensitivity threshold Once this point is
4. 43 Table 14 Comparison of the attenuation of each Screen Room Each room does not provide enough isolation at UHF ia ete detinere oon bU n o asi 47 Table 15 Variables affecting skin depth of material esee 48 Table 16 Feedthrough capacitors capable of passing DC power Ground clips soldered E 53 Table 17 Comparison of attenuation of each screen room to the custom built Faraday Cage The cage offers 10 dB more isolation which was enough to isolate the satellite Irom stray TAG AU ONS AEE Mud 55 Table 18 CC1000 receive sensitivity is a function of frequency data format data rate and frequency separation 8 This table is obtained from the CC1000 datasheet The red highlighting indicates that the uplink to the CC1000 is 600 baud data rate 67 Table 19 The measured sensitivity of CDH Rev 4 for both COMMA and COMMB tested using the sensitivity measurement setup in Section 4 11 essen 71 Table 20 Resting RSSI of CC1000 of CDH Rev essere 72 Table 21 RSSI Characterization of CDH Rev 4 eite 72 Table 22 Sensitivity of CDH Rev 5 with preamp tested using the sensitivity measurement setup in Section 4 11 reel nU est Ee Mesa 73 Table 23 Resting RSSI of CDH Rev edidit
5. S ERI D ERE E E Et 49 4 8 4 FeatHres c eu cte c e E TOR TREES HERI e e YER TENET EER ea ESSE Ve Yee Spe Een 51 4 8 5 Attenuation Performance of Faraday Cage esee eene nennen rene 54 4 9 OTHER CONSIDERATIONS gra it CO ve SE Eg Mu lee xr MEE 55 4 OARE Leakage ERE 25 4 9 2 Yaesu Radio RF Output Power secet iste tte be Fate E EE 56 4 10 SENSITIVITY MEASUREMENT SETUP CHARACTERIZATION ene ene tnter enne nen 56 4 11 TEST SETUP MEASURING SENSITIVITY OF THE POLYSAT COMM SYSTEM see eee 59 CHAPTER 5 SENSITIVITY TESTING eeeeeeee eese sto thats tosta suse ta sons sesso sees suse ta sene 64 3 1 POLYSAT SATELLITE COMM SYSTEM ioter n en oe t dye rede 64 DLL Overview of COMM Setup Une Bener A UEQWUE eM PUE 64 CC TIO00 U HE TT0nscelVerss Sess oe m eee RI poer pes oi eee eie he Cem A eda Go 65 5 1 3 Receive Sensitivity versus Frequency Separation and Data Rate eene 66 S E AERSST Ontpit in ou e UE Ee P E PU ADD EE BU Es 68 SE RUN 69 oe e A DY a Revisions e et De E DRE RR et ge xD HEU PER EE eR ae e EEUU 69 3 2 COMM SENSITIVITY TESTING rhe rr DR PAN EU EAEE
6. eese nennen 108 7 2 6 E ER EE EN Dodo couse 108 CHAPTER 8 SUMMARY 110 8 1 RECEIVE SENSITIVITY TEST SETUP iced etae edite besito tie Mee on e e eA EOTEKO ESS 110 8 1 Sensitivity Characterizationg aic eee dee red eR IS eni Dee RR ERE Nene Gables dee eR REA RUIT 110 8 2 CDH SENSITIVITY PERFORMANCE ouye ennt ennt resoan iea E aore koene sVe a e sanata sessanta sna ins 110 8 2 CDH Rev 4 versus Revo E a d Dee Maina waa ees 110 6 2 2 Noise Characterization of the Receive Line 111 6 2 3 Link Budget Compared to Receive Sensitivity of CDH Rev S sse 111 8 3 CC1000 rst tee nee eee tti tht ET WE a Gena ey 111 5 3 T Limitations of the CCT000 mee d rU BR Re 111 5 3 2 Oyerall Sensitivity a de Sek Tg a See ous fh Nv at Deis BAI d RI Es 113 8 4 FUTURE WORK eet perte estt EHE Beet re b e re btt ee S Peor bust tenere dene ee 113 8 41 New Layo tfor GP35 cite gn IR EEUU rem 113 6 4 2 Receive Sensitivity versus Temperature eese 114 6 4 3 Antenna Characterization and Re design eese eene nente nennen eene tne 114 84A Long Duration TestiWgs 2 Ren e S Oerniiseiieen ate 114 6 4 5 Upgrading the Ground Station to Improve Uplink to the Satellite
7. 106 Figure 82 Frequency separation 112 Figure 83 HVL 1100 1000W amplifier s ce petet eds 115 xii LIST OF TABLES Tablet Uplink Bug Get o ore dan ioi tl Deci e tare beet ore o Wheat ade e ens 18 2 E 19 Table 3 Measured sensitivities of two revisions of PolySat COMM system and the ground station LCCC Ver 20 Table 4 Variable Attenuator Characteristics Be 27 Table 5 Characteristics of ZFRSC 42 S Power Splitter esee 28 Table 6 LNA Specifications required voltage bias typical operating current and absolute TAR AMIN PO WET TuS e SN ose QV Lee UR dae tu redi 33 Table 7 across the 70 CM Amateur Radio band eene 34 Tables LNA a dh tte itd t e mera E hu Bl a edu A de A ON eoa 35 Table 9 S across the 70 CM Amateur Radio band eene 36 Table 10 S1 across the 70 CM Amateur Radio band esee 37 Table 11 1 dB compression point of LNA output seen 38 Table 12 Typical and measured Noise Figures from 0 010 3 GHz eee 40 Table 13 Noise Figure and Gain of individual stages in dB and linear equivalents used to calculate the Noise Figure of cascaded Spectrum Analyzer and
8. 42 Fisure 29 The screen room in rin e as 44 Figure 30 Screen room in 118 left and accompanying RF passthrough right 45 Figure 31 EMC Chamber at the RFID lab in Building 4 46 Figure 32 Testing each screen room for isolation a handheld radio transmitting at 437 MHz was placed outside each screen room and the HP8566A Spectrum Analyzer with antenna was used to measure the power with the screen room door open and closed The difference in measured power with the screen room door open and closed provided an approximate level of auenuatlOti s iae Go 47 Figure 33 Concept of Faraday Cage construction duced t eq corsa MA DU e 49 Figure 34 Unetched PCB being soldered together to form the enclosure 50 Figure 35 Completed cage using clamps to seal the lid Coax is seen attached to each RF pass through 51 Figure 36 Spacious interior allows testing on larger satellites 3U if necessary CP3 TestSat is attached via the U FL antenna connector with CDH Rev 5 with the LNA amplifier same CDH flown on CPG 52 Figure 37 N Type and SMA pass through ports When not in use each port should be terminated with 500 termination to prevent the possibility of
9. 52 Figure 38 SMB Connectors used to pass DC power into the Faraday 54 Figure 39 Testing the Faraday Cage for isolation A handheld radio was placed inside the cage transmitting at 437 MHz The HP8566A Spectrum Analyzer equipped with an antenna was used to measure the signal strength with the cage closed and open The difference between these two measurements provides an approximate level of isolation 54 Figure 40 Characterizing the receive sensitivity setup sese eene 57 Figure 41 Actual signal reaching DUT dBm versus attenuator setting The attenuation was adjusted from 40 to 110 dB and the signal reaching the Faraday Cage Section E of Figure 12 was measured using the Spectrum Analyzer Section D of Figure 12 A strong linear response indicates that the test setup Figure 12 is capable of accurately measuring receive SENSI VILY ssid ERE EE 58 Figure 42 Sensitivity Measurement setup during a satellite response to a command The PolySat COMM system Section E outputs IW 30 dBm and the resistive splitter Section C only offers 6 dB port to port isolation A signal of 24 dBm will permanently damage the LNA attached to the Spectrum Analyzer Section D sees 60 Figure 43 Sensitivity setup for finding the threshold of sensitivity of the PolySat COMM system The LNA is disconnected from the power splitter becau
10. Calculating the expected receive sensitivity of the Yaesu FT 847 The datasheet states that the minimum sensitivity is 0 125 uV so the minimum sensitivity in power dBm can be calculated The filter width of the SSB is 2 2kHz so the thermal noise seen can be calculated Equation 10 174dBm Hz 1010g49 2200 140 6 dump Equation 10 Calculating the thermal noise of the Yaesu FT 847 receiver in Single Side Band SSB With a theoretical noise floor at 140 dBm and a 10 dB SNR at 125 dBm measuring the receiver sensitivity will provide a good baseline comparison to the CP Bus COMM system To test the sensitivity the testing procedure was reversed TestSat now the RF source was programmed to beacon every 30 seconds providing packet information for the Yaesu Transceiver to decode Attenuation was increased to reduce the signal received by the Yaesu transceiver Once MixW Figure 12 could no longer decode the beacon packets the signal strength was measured using the LNA and Spectrum Analyzer at the last attenuation setting MixW could properly decode packets 76 Variable Attenuation Attenuation Yaesu FT 847 Radio Faraday Cage TestSat TX Figure 52 Test Setup to measure the receive sensitivity of the Yaesu FT 847 ground station receiver The satellite is set to transmit every 30 seconds providing the Yaesu FT 847 data packets to receive and decode By adjusting the variable attenuator the
11. ee 115 RESINID AID MEN M LEE 116 8 3 1 COnNClUSTON ME 116 RAE H ANII vr 116 CHAPTER 9 WORKS CITED esee eese s Eet eana 118 APPENDICES APPENDIX A SENSITIVITY MEASUREMENT PROCEDURE eene eene nennen nennen nennen enne inneren nnn 120 APPENDIX B LIST OFAACRONYMS te RC RI DURER Ur HD ERO redet bee repe 128 viii LIST OF FIGURES Figure 1 CubeSat Standard 10x10x10 cm cube 1 33 kg maximum weight 3 Figure 2 PPOD on left 3 CubeSats on right 4 Figure 3 CPL satelite eee EEK Ea eae 6 2 oi bei ale ton dne on tes 7 Ligure 5 CPS Fight Unit acie pepe de Ia volt aa ae ded e do alt 8 Figure 6 Engineering unit of CP6 in the 1 10 Figure 7 CP7 development platform particle damper beams on left high voltage board for driving piezo crystals on right 12 Both boards interface to a modified CDH and Figure 8 John Abel shown left flying the CP7 development platform on the NASA Zero Gravity Flight Three particle dampers were tested driven with different amplitudes and Treue nee s p ate eph i HP arte ur tu IPod 12 Figure 9 Antenna Reclproelty
12. EMC Chamber at the RFID lab in Building 4 Testing the EMC Chamber at 437 MHz showed attenuation at 437 MHz slightly higher as compared to the other screen rooms but not high enough to warrant use Some of the modifications done to the chamber may have compromised the attenuation at UHF Furthermore the room is inconveniently located in Building 4 a substantial walk from the PolySat lab in the MSTL Although this sounds trivial it would require the use of a vehicle to bring all the test equipment radio power supplies test setup etc which would be very inconvenient It would also require an RF pass through installed Although each room is a valuable asset to the EE Department none of them offered enough attenuation at 437 MHz A HP8566A Spectrum 46 Analyzer with antenna attached was placed inside the screen room and a handheld radio was placed outside transmitting at 437 MHz The difference in measured power with the screen room doors open and closed provide an approximate measure of isolation A block diagram of the setup is shown Screen Room Spectrum Analyzer with antenna RF Source transmitting at 437 MHz Figure 32 Testing each screen room for isolation a handheld radio transmitting at 437 MHz was placed outside each screen room and the HP8566A Spectrum Analyzer with antenna was used to measure the power with the screen room door open and closed The difference in measured power with the screen room door op
13. SCL 2 and data line SDA CDH 2 originate from the CDH processor ues 92 Figure 76 Beacon commands sent every 60 seconds and the satellite responses monitored The horizontal axis shows how many satellite responses there were for each command from the group station The vertical axis is divided into bins showing that not every command received TES PONS 95 Figure 77 Variations of the IF due to the crystal can cause a reduction in the sensitivity ofthe CC TODO erheben utat voice be codi pei tH ot a E A taa ertet ee e reti peor tius 102 Figure 78 RF Chain testing of CDH Rev 5 Using the RF source a signal at 437 MHz was applied to the antenna connector The CC1000 is not receiving any signal through the antenna connector indicating a manufacturing defect in the RF 104 Figure 79 Male U FL connector om CDH Rev 105 Figure 80 U FL female connector with open circtuit defect center conductor prongs are spaced too far apart shown on left new U FL connector shown on right The connector on right shows the proper center pin spacing sssseeseeeeeeeeeerenenen nenne nennen 106 Figure 81 A worn out U FL connector on left behaves an open circuit By adjusting the spacing of the two prongs a proper connection was established shown on right
14. amp produces 1W Gry Isotropic Antenna gain 2 to 10 dBi Depends on orientation Les Free space path loss 142 2 to 153 4 dB 50 ft length Lm Miscellaneous loss 1 dB Estimated additional loss Connectors radio etc Lex 5 8 hardline loss 0 22 dB 50 ft length 18 Ggx Antenna gain 18 95 dBi Antenna Gain Total 90 to 113 dBm Approximate strength after antenna rounded Table 2 Downlink Budget The antenna system has an SSB Electronics preamplifier before it reaches the Yaesu FT 847 transceiver but this was not included in the link budget because it does not increase the sensitivity of the transceiver 6 The preamplifier s purpose is to increase the signal after the antenna to overcome losses in the feedline to the transceiver The ground station receiver must have a sensitivity of 123 dBm to provide a 10 dB worst case link margin 3 3 Addressing the Unreliable Uplink In order to shed light on the unreliable uplink problem an accurate method of measuring receive sensitivity was developed Characterizing the PolySat communications system helped the team gain insight to the unreliable problem 3 3 1 Preview of Results Receiver Sensitivity Testing After the test setup was developed two revisions of the CP Bus communication system were tested for receive sensitivity CDH Rev 4 had a sensitivity of around 90 dBm CDH Rev 5 which includes an LNA
15. approximately 3 dB The power of the measured signal is only twice that of the noise power significantly reducing the accuracy of the measurement Furthermore the marker used to read the amplitude of the signal starts to fluctuate significantly varying by as much as 1 2 dB increasing the error in the measured signal Although 1 dB does not seem like much variation the system was characterized in 2 dB increments so an error of 1 2 dB corresponds to a percent error of 50 100 The accuracy of the system behaves as expected even as the signal power becomes less than 120 dBm 4 11 Test Setup Measuring Sensitivity of the PolySat COMM System The sensitivity test setup shown in Figure 12 was designed to measure the receive sensitivity of the PolySat COMM system Recalling the back to the definition listed in Section 4 1 the receive sensitivity of the PolySat COMM system is the weakest signal command measured in dBm in which the COMM system could reply to The PolySat COMM system 59 outputs approximately 1W of RF power 30 dBm 3 This means that a response to a command will result in 30 dBm of RF power entering the power splitter D Computer Spectrum Analyzer HP 8566B 43 2 dB LNA Rigblaster Figure 42 Sensitivity Measurement setup during a satellite response to a command The PolySat COMM system Section E outputs 1W 30 dBm and the resistive splitter Section C only offers 6 dB port to port isolat
16. enough to prevent stray radiation from triggering the satellite This was discovered by placing the satellite in the screen room and sending commands from outside with the Yaesu radio Even with the minimum RF power and the output of the FT 847 terminated with a 50Q load the satellite consistently responded This indicated that some sort of RF leak was present most likely due to the ceiling vents or light switch Or the screen room wasn t designed for attenuation at UHF Ultra High Frequencies Figure 30 Screen room in RM 118 left and accompanying RF passthrough right The screen room in RM 118 offers comparable attenuation to the cage in room 116 The RF passthrough is a PL 258 bulkhead This connector although labeled UHF isn t suitable for frequencies over 300 MHz and wasn t designed for the typical 50Q characteristic impedance of 45 most connectors and cables used in today s RF applications it does not have a constant 5092 impedance Although the insufficient attenuation was the limiting factor the RF passthrough was a good indicator that the room would not offer enough attenuation at UHF The third screen room on campus located in Professor Dean Arakaki s lab near the RFID lab in Building 4 is the EMC Chamber designed and operated by Professor Dean Arakaki In addition to RF isolation ferrous ceramic tiles line the interior of the room allowing for conducted and radiated emissions testing Figure 31
17. sueco pest edet Ie ta ce dece dise t de cng Do idees 14 Figure 10 Field testing of CP6 The satellite was taken halfway up Bishop s Peak approximately 2 miles away Attenuators were placed on the ground station reducing the BLO Mal SEPSIS Uy NP c 16 Figure 11 Link budget Using typical orbital parameters and the above figure to calculate the slant range the Friss Free Space equation determines the attenuation from path loss The attenuation of the orbital path varies between 142 2 153 1 dB 17 Figure 12 System level overview of the setup developed used to measure the receive sensitivity of the PolySat COMM system 24 Figure 13 Sensitivity Measurement setup this shows all the necessary components required for measuring sensitivity This setup was used to measure the sensitivity of two revisions of the PolySat COMM system and the ground station receiver 26 Figure 14 Attenuation Setup allowing signal strength to be adjusted from 0 to 110 dBm 27 Figure 15 The resistive power splitter divides the signal equally to both the DUT and AM Aly Ze Livi EMT HX 28 Figure 16 ZFRSC 42 S Power Splitter Signals 6 dB down from the original signal at port S will appear at ports 1 amp 2 With signals of equal magnitude reaching the Spectrum Analyzer and the DUT the signal strength at the D
18. 4 CP4 After the failure of Dnepr 1 the backup flight unit of CP2 was re flown on Dnepr 2 Although it is actually CP2 the flight required a unique name resulting in a rename from CP2 to CP4 Several months in orbit the satellite failed The C amp DH seems to be locked up but the satellite still switches between the redundant COMMs and will respond to a limited set of commands 2 3 5 CP5 CP5 currently in the design phase consists of a de orbiting mechanism payload Since CubeSats use Consumer Off the Shelf Components COTS the target lifetime is only 3 6 months But CubeSats typically in Low Earth Orbits LEO of approximately 500 700 km will continue to orbit for 25 30 years as space debris De orbiting mechanisms will allow the CubeSat to re enter and burn up in the Earth s atmosphere after completing the primary mission Since CP5 will utilize the standard CP Bus a goal of this thesis is to provide recommendations on bus improvement specifically for improving uplink reliability 2 3 6 CP6 After CP3 and CP4 were flown the PolySat team noticed a significant problem in closing the uplink to both satellites To combat the poor receive sensitivity of the CP Bus a Low Noise Amplifier LNA was added between the antenna and receiver A tri state buffer was added between the payload processor and cameras to ensure data was being sent to only one of the two cameras at a time Compared to CP3 the software received a major overhaul Sign
19. 500 MHz Insertion Loss 0 5 dB max Attenuation Accuracy 0 2 dB or 1 RF Input Power AVG 2W Table 4 Variable Attenuator Characteristics During sensitivity testing a high power dissipation 30 dB attenuator is added to the 433 MHz output of the Yaesu FT 847 transceiver to reduce the power from IW to 1mW This is required since even the minimum output power on the 433 MHz band is approximately IW With the 27 power effectively reduced a variable attenuator is added to further reduce the signal by 0 110 dB This provides the necessary range for determining the minimum detectable signal 4 5 Resistive Power Splitter In order to accurately measure the signal reaching the DUT Section E of Figure 12 the path must be split into two separate but equal paths Section C of Figure 12 Spectrum Analyzer Splitter 0 110 dB RF Source Variable Attenuation Attenuation DUT Figure 15 The resistive power splitter divides the signal equally to both the DUT and Spectrum Analyzer To do this a resistive power splitter was used Both ports are 6 dB down from the input port Half the power 3 dB is dissipated in the resistive network and the remaining power is split equally another 3 dB between the two ports The ZFRSC 42 S Splitter features a wide bandwidth low insertion loss and excellent amplitude imbalance characteristics Frequency DC 4200 MHz In
20. 74 MHz apart are clearly visible Broadband noise is much higher than CC1000 development board Figure 67 549 lees sere a oput aeqne 86 Figure 65 CDH Rev 5 CommA frequency span of 100 500 MHz RBW 86 Figure 66 CDH Rev 5 COMMB frequency span of 100 500MHz RBW 1 2 86 Figure 67 CC1000 Development Board 100 500 MHz RBW 1MHz Crystal harmonics repeated every 14 74 MHz and LO leakage visible at 433 860 87 Figure 68 System Noise Floor 100 500 MHz RBW 1MHz LNA and Spectrum Analyzer ee 87 Figure 69 CDH Rev 4 COMMA RBW 100kHz over the 70 CM amateur band 88 Figure 70 CDH Rev 4 COMMB RBW 100kHz over the 70 CM amateur band 88 Figure 71 CC1000 Development board RBW 100kHz over 70 CM amateur 89 Figure 72 System Noise Floor over 70 CM amateur band eee 89 Figure 73 The CDH Rev 4 had two types of crystal harmonics visible The larger harmonic is spaced 14 74 MHz apart the frequency of the CC1000 crystal and the smaller harmonics were spaced every 1 474 2 eene enne 89 Figure 74 The DC DC converter regulates the battery voltage from 4 2V to 3V for the P lysSat COMNIS ysterm 2 5 aeree nci EM Se EU S Lee 90 Figure 75 PC switching
21. 866 Figure 58 CDH R4 CommB noise floor centered at the PolySat COMM receive frequency 437 365 MHz LO leakage visible at 437 515 MHz RBW 10 kHz Both COMMA Figure 57 and COMMB Figure 58 have very similar noise characteristics but COMMB has slightly lower noise In receive mode both COMMs measure around 70 dBm at the RX frequency of 437 365 MHz comparing closely to the noise of the development board at the RX frequency of 434 010 MHz Figure 59 82 433 518 433 768 434 010 434 260 434 510 Figure 59 CC1000 Development Board centered at receive frequency 434 010 MHz LO Leakage visible at 433 860 MHz RBW 10 kHz she ral Ul Am tbe PP 436 860 437 118 437 360 437 618 432 860 Figure 60 System Noise Floor of the LNA and Spectrum Analyzer RBW 10 kHz The input of the LNA is terminated with a 50 connection This provides a reference noise level at which to compare each noise floor graph to COMM System Noise at RX frequency Comments CDH Rev 4 COMM 65 Measured sensitivity 88 5 dBm CDH Rev 4 COMM B 70 Measured sensitivity 92 4 dBm CC1000 Development Board 75 dBm Figure 59 System Noise Floor 85 dBm Figure 60 Table 27 Results of the noise measurement at the RX frequency of CDH Rev 4 and CC1000 development board The noise of COMMA at the RX frequency is 10 dB greater than the CC1000 development board and 5 dB higher than COMMB possible causing the
22. Analyzer 129
23. COMM system will be performed providing an interesting look at possible causes of the inconsistent uplink and methods of improving the COMM system For future bus development this test setup can be used to accurately measure the receive sensitivity iv ACKNOWLEDGEMENTS Thanks to my family and friends for the unwavering support throughout my academic career Thanks to Dennis Derickson for his support and expertise His guidance played a huge role in the work presented throughout this paper Thanks to Dr P for helping pioneer the CubeSat standard and encouraging both PolySat and CubeSat to dream big Thanks to the PolySat team for support during testing especially the software team for the constant troubleshooting of the software TABLE OF CONTENTS LIST OB FIGURES eA Ix LIST OF TABLES op P D XIII LIST OF EQUATIONS me XV CHAPTER 1 INTRODUCTION SCOPE OF THESIS cscscssssssssscsscssssssescssssessssnsssesssssesssssssssssnessesooes 1 CHAPTER 2 BACKGROUND OVERVIEW OF SATELLITE PROGRAM eene ceste tne tosta tn anat 3 2 1 CUBESAT teer IR astitit RE Ie RERO 3 DAD Cal Poly CubeS tiu si usd eed dete item rue te dre decet 4 2 2 CubeSat Launch HIStory 35 5a t ei ete RW eter ib ee d EG rever d RERO EOS 4 2 2 POLYS AT PROGRAM terere eT em rae eu tu bet OA GE p E recae ces 5 2 3 OATELLITES EM EE 5 DS AICPA iun cuu E REP HSU Rd m RUNE CB II
24. CP4 Falcon 1 2008 Two modified Mk III P PODs with two NASA 3U CubeSats were lost due to launch vehicle failure Minotaur I TacSat3 2008 Two P PODs launched four CubeSats including Cal Poly s CP6 2 2 PolySat Program PolySat Cal Poly s CubeSat development program is a team of multidisciplinary students designing CubeSats for Cal Poly Both PolySat and CubeSat teams are located in the ATL Advanced Technologies Laboratory Building 007 Room 15 Originally founded in 1999 the PolySat team embraces Cal Poly s Learn by doing motto by developing satellites from the ground up Although CubeSat subsystems are commercially available for purchase such as COMM or power systems PolySat emphasizes the student learning experience by developing all hardware and software in house Located in Room 15 Building 7 the lab provides the team with the necessary tools for satellite development electronics bench for hardware development and RF testing a software bench with computers and two independent ground stations for communicating with the satellites during passes CubeSat maintains a cleanroom down the hall used for satellite integration to the P POD and a Thermal Vacuum Chamber TV AC is located in Building 4 2 3 Satellites Over the years PolySat has developed several satellites Three of these are currently in orbit PolySat is also currently involved in several projects 2 3 1 1 Development of Cal Poly s first s
25. RF switches select between RX and TX The RX line goes directly to the CC1000 and the TX is from the RF amplifier vilis mp E I Er E M E ET RT TR 69 Figure 50 RF Chain of COMM A on CDH Rev 5 The only difference is the addition of an LNA associated matching network and high pass filter see 70 Figure 51 CPS Flight Candidate RSSI CUL yeso pietre ott tir pest Crac dato diues 75 Figure 52 Test Setup to measure the receive sensitivity of the Yaesu FT 847 ground station receiver The satellite is set to transmit every 30 seconds providing the Yaesu FT 847 data packets to receive and decode By adjusting the variable attenuator the signal strength to the FT 847 was reduced until it could no longer decode packets providing a threshold oLr eelve SeRSIEVIEV 3 Urea A qu Del eh etes quse debel 77 Figure 53 Test setup to monitor the noise at the receive line of the PolySat COMM system The LNA attached to the Spectrum Analyzer is used to observe the noise floor 79 Figure 54 Test setup showing the system noise floor of the LNA and Spectrum Analyzer This is important in comparisons between the measured noise of each COMM SY SUTIN Pace he unco usted du E E or ac onan etu od 80 Figure 55 LO leakage For the CC1000 transceiver the leakage is 57 dBm 8l Figure 56 The LO will be 150
26. Shift Keying AFSK modulation the separation frequency actually used on the CP Bus is 2 kHz 7 Most likely this reduces the achievable sensitivity but the datasheet does specifically list the expected sensitivity for a separation frequency of 2 kHz at 600 baud With CDH Rev 5 successfully decoding packets down to 101 dBm the CC1000 is performing reasonably well A new layout could reduce broadband noise see Section 6 1 possibly increasing sensitivity by several dB Although the sensitivity of the 98 PolySat COMM system is ultimately determined by the CC1000 the LNA incorporated in Rev 5 is clearly capable of increasing the overall sensitivity By following the manufacturers recommended layout of the CC1000 it could be possible to achieve several dB of sensitivity 7 1 3 PolySat COMM Performance Compared to Future Replacement Transceiver The Axsem AX5042 will replace the CC1000 as the transceiver for the new PolySat COMM system According to the datasheet the stated sensitivity is 122 dBm at 1200 baud 17 Datasheet Sensitivity 111 dBm Conditions FSK modulated data at 1 2 kBaud separation frequency 64 kHz Datasheet Sensitivity 122 dBm Conditions FSK modulated data at 1 2 kBaud Table 34 Sensitivity comparison of CC1000 and the AX5042 future replacement transceiver The datasheet of the AX5042 states that the receive sensitivity is not dependent on FSK frequency separation 17 The listed
27. The feed through capacitor passes DC signals but shorts AC signals to ground An SMB connector allows easy connection to a power supply SMA connectors with semi rigid coax provide connections at the input and output LNA Specifications Voltage 11V Current 72 Gain 437 MHz 42 3 dB Max Input Power 5 dBm Table 6 LNA Specifications required voltage bias typical operating current and absolute maximum power The LNA is a small signal amplifier meaning that it is not designed to amplify larger signals 0 dBm or 1 mW The maximum input power shown in Table 6 is 5 dBm or 0 316 mW and it is important to not to exceed this rating otherwise permanent damage to the LNA would occur 4 6 2 Input Return Loss of LNA Return loss is a measure of the reflected power at a port compared to the power incident upon that power Typically denoted in magnitude form it describes how well matched the LNA is to the 50Q system 33 IS dB frequency GHz Figure 21 Sj of LNA 511 f GHz dB 0 4286 20 778 0 436075 20 559 0 44355 20 356 0 451025 20 395 Table 7 S across the 70 CM Amateur Radio band The input return loss is greater than 20 dB across the 430 450 MHz amateur radio band Below 0 5 GHz the input is well matched to a 50 2 system As the frequency increases past 1 GHz the return loss decreases to approximately 10 dB 34 4 6 3 Outp
28. Uplink Using Derek Huerta s link budget calculator 3 the attenuation of the orbital path was calculated for both best and worst cases Assuming a 700km 98 degree inclination as a typical CubeSat orbit the satellite to ground station distance will be a function of the elevation from the horizon For worst case a 5 degree elevation results in a slant range of over 2 500 km At 90 degrees satellite passing directly overhead the slant range is simply 700 km Thus the attenuation varies from 142 2 to 153 4 dB Figure 11 Link budget Using typical orbital parameters and the above figure to calculate the slant range the Friss Free Space equation determines the attenuation from path loss The attenuation of the orbital path varies between 142 2 153 1 dB 2 zo Path Loss dB 10 log d distance f frequency Hz 3 0 10 m s Equation 1 Calculating the free space path loss The Hertz ground station equipped with dual phased circular polarized Yagi antennas and a 100W amplifier outputs approximately 72 dBm 6 Subtracting off the path loss the signal strength at the satellite will be between 71 to 83 dBm The CP Bus antenna is a V2 17 dipole but with slightly shorter length due to size limitations A characterization of the antenna showed the range of the E Plane radiation pattern to vary 12 dB from the max depending on orientation 13 An ideal dipole has a gain of 2 15 dBi dB relative
29. done in space There are watchdog timers for each microcontroller but a reset only occurs if they get stuck in loop and fail to keep tapping the timer 3 The Smart Fuses only cut power thereby resetting the satellite if the current limit is exceeded 2 Second over a 6 month period of testing CDH Rev 4 and CDH Rev 5 which is the target life span of a CubeSat problem 4 Table 31 occurred during sensitivity testing in the lab approximately 10 times With CP3 and CP4 almost 3 years old now it is highly likely that this behavior could have also occurred at least several times in orbit Problem 4 of Table 31 was first observed while measuring receive sensitivity Section 4 11 of the PolySat COMM system both CDH Rev 4 and Rev5 The satellite was sent beacon commands while reducing the signal strength At some point the satellite could no longer decode the command and respond Occasionally the satellite would stop responding at all even as the power was increased above the dropout threshold Increasing the power back all the way to 50 dBm still did not prompt a response from the CDH under test A hard reset cycling the power of the satellite was the only way to solve this problem This suggests that the uplink problem may be more than just poor sensitivity One characteristic of Problem 4 Table 31 was that it usually occurred at or very near the sensitivity threshold Figure 44 94 6 3 3 Long Duration Communication Test Testing fo
30. has a sensitivity of approximately 100 dBm Comparing this to the link budget a more sensitive transceiver is needed The ground station receiver the Yaesu FT 847 had a measured sensitivity of 115 dBm which is 15 25 dB more sensitive than the PolySat COMM system The PolySat COMM system should have a sensitivity comparable to the sensitivity of the ground station to help increase reliability of the uplink 19 CDH Rev 4 90 dBm CP3 CP4 CDH Rev 5 100 dBm Includes LNA CP6 Yaesu FT 847 115 dBm Ground Station Table 3 Measured sensitivities of two revisions of the PolySat COMM system and the ground station receiver 3 3 2 Additional Testing Performed and Problems Discovered After characterizing the receive sensitivity of the CP Bus additional testing was performed as an effort to find any flaws in the COMM system design that would cause a reduction in sensitivity A DC DC converters used for efficiency were previously believed to be causing desensitization of the COMM system The PC data bus used for data transfer switches at 100kHz This is suspiciously close to the 150 kHz IF frequency of the CC1000 Testing was performed for both theories but it was determined that they are not causing a reduction in sensitivity B Using the setup noise on the receive line of the CP Bus was monitored and it was immediately obvious that poor layout resulted in a huge increase of broadband noise A new layout using proper PCB layout te
31. is included The testing performed is described along with a discussion of the results A procedure of using the test setup to measure sensitivity is included in the appendix 21 Chapter 4 Measuring Receive Sensitivity 4 1 Overview In order to measure the receive sensitivity of the PolySat COMM system a test setup must be developed which allows the accurate adjustment and measurement of the signal reaching the receiver under test The main goal of this project was to develop a method of accurately measuring the receive sensitivity of the PolySat COMM system Since it is typical to find receiver sensitivities ranging from 110 dBm to 130 dBm the setup must be capable of measuring signals of comparably small magnitudes Stray leakage from the ground station transmitter is easily picked up by the receiver under test making it very difficult to reduce the received signal strength below the sensitivity threshold of the receiver A high degree of isolation is required to prevent this from occurring Reducing a signal to such diminutive amplitudes in a controlled manner is not a trivial task It is also very difficult to measure such small signals 4 1 1 Definition of Receive Sensitivity Receive sensitivity is defined as the minimum input signal that produces a desired output A receiver s ability to successfully receive and decode a signal requires a minimum Signal to Noise Ratio SNR The receive sensitivity is measured in dBm specified
32. located measure the power of this signal using the LNA Spectrum Analyzer and resending the command 21 Attach the output of the 42 3dB LNA output to the input of the Spectrum analyzer Using the BNC to SMB cable bias the LNA with 11V set a current limit of 100 mA 22 With the attenuation set to the threshold of sensitivity see step 20 send the 0201 command and press Single to freeze the display Use the peak search to measure the value Repeat this 5 times recording each value The excel sheet should subtract off the LNA gain and average the value This value is the approximate receive sensitivity of the receiver under test 23 Record this value in the Trial 1 box Make sure to comment the file with all pertinent information date CDH version any anomalies that occur etc 24 Flight candidates should be tested 3 different trials for consistency Be sure to cycle the power so that the satellite gets a hard reset between each test 25 Commit the file to XSERV and put all the RF testing hardware away 125 Groundstation Software Install 1 om ND UR 11 12 13 14 17 18 Install MixW 2 18 then copy and paste the three MixW folders into Program Files say yes to overwrite Install the KeySpan USB serial driver Copy the CommEmulDrv3 zip to the desktop Extract it Start gt Control Panel gt Add Remove Hardware This will launch the Hardware Wizard Click Next then click on Yes h
33. on COMMA is almost 20 dB higher This would reduce the SNR resulting in a reduction of receive sensitivity Since both COMMs are identical hardware and software this large increase in broadband noise is most likely attributed to layout differences 6 1 2 Receive Line of CC1000 Broadband Noise Floor A PCB with poor layout typically will see an increase in broadband noise 15 The next noise test was to see if there was broadband noise over a large span 400 MHz 100 000 200 000 368 668 468 688 Figure 63 CDH Rev 4 CommA 100 500 MHz RBW 1MHz Harmonics of crystal spaced 14 74 MHz apart are clearly visible Broadband noise is much higher than CC1000 development board Figure 67 85 100 000 268 666 300 000 400 000 568 666 Figure 64 CDH Rev 4 COMMB 100 500MHz RBW 1MHz Harmonics of crystal spaced 14 74 MHz apart are clearly visible Broadband noise is much higher than CC1000 development board Figure 67 COMMA has an overall noise floor of slightly below 50 dBm Figure 63 with crystal harmonics reaching 35 dBm The crystal harmonics are from the CC1000 s 14 7456 MHz crystal oscillator and are spaced apart by 14 7456 MHz COMMB has a lower noise floor of about 60 dBm and crystal harmonics below 40 dBm Figure 66 CDH Rev 5 COMMB frequency span of 100 500MHz RBW 1MHz 86 The broadband noise of COMMA is typically around 50 dBm Figure 65 but shows a much more controlled at around 65
34. possible especially at greater power levels over 10 dBm This reduces the reliability of the testing performed The magnitude of the signal reaching the satellite must be accurately measured to assess whether or 15 not the COMM system is capable of overcoming the path loss in orbit It is also difficult to predict how multipath signals reaching antenna via multiple paths caused by reflections off terrain affects the testing Following the addition of the LNA to the CP Bus range testing was performed in a similar manner to verify the improvement in sensitivity 4 Although a 10 dB improvement was observed there were several problems with the testing procedures Like the previous testing attenuation was added to simulate the orbital path loss but the test lacked a conclusive measurement of the actual output power of the ground station The test also lacked a method of measuring the power received at the satellite Using fixed attenuators in 3 and 10 dB increments reduced the resolution of the test A variable attenuator could help resolve sensitivity improvement in greater detail but a controlled method of feeding an RF signal to the satellite is lacking which is needed to accurately measure the receive sensitivity Figure 10 Field testing of CP6 The satellite was taken halfway up Bishop s Peak approximately 2 miles away Attenuators were placed on the ground station reducing the signal strength 16 3 2 3 Link Budget
35. sensitivity of the Yaesu FT 847 is 125 dBm with a measured sensitivity of 115 dBm If the AX5042 can perform similarly the team should have no problem closing the uplink 7 2 Recommendations for Improving Uplink and Future PolySat COMM Based on the results of the testing several recommendations can be made These include recommendations directly applicable to CP5 the last satellite to utilize the current CP Bus and also recommendations to aid future bus development 99 7 2 1 Improved Layout of CDH Rev 5 Crystal harmonics and broadband noise Section 6 1 on the PolySat COMM system could be reduced While it s not clear how much the noise is reducing the sensitivity proper layout techniques could significantly reduce noise seen at the RX line of the CC1000 possibly improving sensitivity This is especially important with CDH Rev 5 as in some cases the LNA of COMMA is amplifying the broadband noise at the receive line at the input of the LNA Measured sensitivities of COMMA and COMMB differed and differences in layout of each COMM could be causing this Application notes for suggested layout of the CC1000 are available from Texas Instruments and detailed information about reducing EMI is available online With CP5 using the same PolySat COMM system a new layout is highly recommended to maximize hardware performance 7 2 2 COMM Software Testing Based on the results of the long duration test it is suggested that the COMM upl
36. the PolySat ground station receiver 110 8 2 2 Noise Characterization of the Receive Line The noise analysis in Section 6 1 1 6 1 4 shows significant differences in the noise floor of CDH Rev 5 Of the two CDH Rev 5 boards measured for receive sensitivity Section 4 11 there was a significant difference in sensitivity between each redundant COMM For both boards broadband noise polluted the receive lines in front of the LNA on COMMA Reducing broadband noise can be achieved by placing a narrowband filter in front of the LNA and following proper layout guidelines 8 2 3 Link Budget Compared to Receive Sensitivity of CDH Rev 5 For both revisions of the PolySat COMM system the sensitivity does not provide enough link margin see Section 3 2 3 and 3 2 4 While the sensitivity of the PolySat COMM system is ultimately limited by the CC1000 a revision of the board could increase sensitivity and additional software testing and modifications could help improve consistency 8 3 CC1000 8 3 1 Limitations of the CC1000 AX 25 an amateur radio packet communication standard was chosen as the communication protocol Bit rates typically 1200 bits s usually do not exceed 9600 bits s and transmission occurs using Audio Frequency Shift Keying AFSK AFSK utilizes a carrier modulated 1200 Hz and 2200 Hz audio tones corresponding to 0 s and 1 s The spacing between the two frequencies denoting 0 s and 1 s is called the separation freque
37. through frictional collisions 12 Particle damper applications could involve reducing vibration in scientific equipment such as optical systems in a zero gravity environment Modeling the performance of particle dampers in microgravity is difficult due to its non linear behavior and dependence on gravity Figure 7 CP7 development platform particle damper beams on left high voltage board for driving piezo crystals on right 12 Both boards interface to a modified CDH and EPS The CP7 team led by John Abel developed an experiment that flew on the NASA Zero G flight in June of 2009 The experiment consisted of cantilever beams with particle dampers with piezo driving elements to vibrate the beams at different frequencies and amplitudes 12 11 Figure 8 John Abel shown left flying the CP7 development platform on the NASA Zero Gravity Flight Three particle dampers were tested driven with different amplitudes and frequencies 2 3 8 Lightsail 1 The Planetary Society is developing a spacecraft propelled by sunlight This is will be done by deploying a large sail and photons hitting the sail will transfer energy to the spacecraft Cal Poly will provide the electronics bus including the CDH EPS and COMM Since Lightsail will be much more ambitious than previous missions it is very important to determine the cause of the unreliable uplink to the PolySat COMM system 12 Chapter 3 Communication Problems 3 1 Unrel
38. with a 5092 termination at the antenna jack The CC1000 of COMMB detects signal strength on the order of 95 dBm corresponding closer to the performance of CDH Rev 4 Section 5 2 3 As seen in CDH Rev 4 the COMM with the lower resting RSSI performed better 73 CC1000 RSSI RSSI COMMB RSSI Datasheet RSSI Table 24 RSSI from both COMMS of CDH Revy 5 compared to datasheet 5 2 5 CP5 CDH Flight Candidate CP5 will utilize the CDH Rev 5 board One of the flight candidate boards was recently completed providing the opportunity to characterize another board The performance of the board is shown below CDH Rev 5 Sensitivity COMMA 96 8 dBm COMMB 100 5 dBm Table 25 CP5 Flight Candidate receive sensitivity performance COMMA performs better than the other CDH Rev 5 candidate showing that manufacturing differences may be causing differences in sensitivity Resting RSSI and Signal Strength COMMA 0 890V 90 dBm 74 COMMB 0 953V 95 dBm Expected 1 1V 105 dBm Table 26 CP5 Flight Candidate resting RSSI values Again the COMM with lower resting RSSI performs better With a lower residual noise the SNR will be greater possibly explaining the better performance It is very interesting that the sensitivity performance varies between each board implying that multiple flight candidates should be tested to determine the bes
39. would 43 be dissipated into the 50 Ohm load but a satellite placed across the room will easily respond to commands sent This simple test shows that a very high degree of isolation is required 4 8 1 Screen Rooms In order to properly characterize the Received Sensitivity RSSI of the satellite a high level of isolation is required to prevent stray radiation from being picked up by the sensitive receiver Cal Poly has three screen rooms on campus all operated by the Electrical Engineering EE department Two are located in the EE department Building 20 One is in the RF Lab room 116 and the other is located in the Communications Lab room 118 The third screen room is located in Professor Dean Arakaki s laboratory near the RFID lab in Building 4 Before the rooms could be utilized each had to be characterized by determining the approximate attenuation at the frequency of interest 437 MHz The original plan was to use one of the existing rooms for sensitivity testing but the isolation of each room proved to be insufficient Figure 29 The screen room in RM 116 44 It was thought screen room in room 116 would be ideal to use because it is conveniently located in the RF lab which contains important test equipment such as spectrum analyzers and network analyzers Additionally an N Type bulkhead connector could be installed which would allow signals to be passed into the cage However the attenuation at 437 MHz was not
40. 24 dB the actual signal reaching the DUT is 50 5 dBm The characterization is done to determine with a high degree of accuracy the strength of the signal reaching the DUT After characterization a user can easily determine the Minimum Detectable Signal MDS of a transceiver by increasing the attenuation until the transceiver no longer responds Then the receive sensitivity is simply the signal amplitude corresponding to the attenuation setting at which packets were last received However after the output variability of the Yaesu FT 847 was discovered it is highly recommended to actually measure the signal during each sensitivity test 57 Attenuation Characterization Actual Signal dBm 0 990x 27 27 R 0 999 Attenuator Setting dB Figure 41 Actual signal reaching DUT dBm versus attenuator setting The attenuation was adjusted from 40 to 110 dB and the signal reaching the Faraday Cage Section E of Figure 12 was measured using the Spectrum Analyzer Section D of Figure 12 A strong linear response indicates that the test setup Figure 12 is capable of accurately measuring receive sensitivity In Figure 41 there is a strong linear correlation between the actual signal and the attenuator setting This indicates that the user can expect an accurate adjustment of the signal when adjusting the attenuator If the attenuator is adjusted by 10 dB the corresponding signal should be reduced by 10 dB The li
41. 3 did not respond to any commands sent suggesting that problem is more than just poor sensitivity of the COMM system CP4 responded to a few commands but the overall results were very surprising It was expected that the SRI dish would have no problem closing the uplink 6 3 2 Observations from Sensitivity Testing Throughout the testing process from September 2009 to February 2010 several software glitches were observed 1 Extra or repeated responses to commands 2 Broken record During a response to a command the satellite would lock up and continue to respond indefinitely until the satellite was reset 3 Non stop beaconing similar to 2 as the satellite was turned on it it would transmit beacons continuously 4 No response the satellite would not respond to any commands at all Table 31 A list of four common satellite behavior problems observed during sensitivity testing 93 Months of testing showed that TestSat would not always behave as expected Table 31 shows list of common behavioral malfunctions Each of these malfunctions is caused by software bugs The software team is aware of these problems as they were first observed during software development A few times during testing TestSat stopped responding altogether This happened relatively infrequently approximately 10 times but is problematic for several reasons First the only solution was to hard reset the satellite which can t be
42. 4 23 dB 26 485 Gi 42 3 dB 16 982 Table 13 Noise Figure and Gain of individual stages in dB and linear equivalents used to calculate the Noise Figure of cascaded Spectrum Analyzer and LNA F 1 26 485 1 2 0031 3 614 5 58 dB F F f G 16 9824 Equation 5 Noise Figure of the Spectrum Analyzer and the LNA cascaded together The Noise Figure of the cascaded LNA and Spectrum Analyzer Section D of Figure 12 shown in Figure 28 becomes slightly higher than the noise figure of the first stage By adding the LNA to the front end of the Spectrum Analyzer Section D of Figure 12 the sensitivity of the Spectrum Analyzer is greatly increased 4 8 Faraday Cage The Faraday Cage Section E of Figure 12 isolates the satellite or receiver under test from stray RF radiation of the Yaesu FT 847 Previous attempts at measuring the receive sensitivity failed due to a lack of isolation from stray RF or radio leakage 3 4 Without sufficient isolation the satellite will respond regardless of how much attenuation is added to the setup The CC1000 transceiver has a typical receive sensitivity of 110 dBm 2400 baud FSK modulated data which is 0 01 pW of power This illustrates that even very weak RF radiation can seriously pose a problem for sensitivity testing RF leakage from the Yaesu FT 847 is easy to observe by attaching a 50 Ohm load to the 433 MHz output Ideally all the output power
43. 41 dBm 35 497 dB MARKER TO PEAK MORE gt F F 35 00 dBm 436 075 A MHz CH 20 00 dBm Table 11 1 dB compression point of LNA output The 1dB compression point referenced to the output occurs when the input signal amplitude reaches 29 dBm The gain drops from 42 2 dB to 41 2 dB In order to prevent errors during characterization or testing the magnitude of the signal reaching the LNA should be several dB less than 29 dBm In order to prevent permanent damage the maximum signal at the input should be significantly less than 330 mW 24 2 dBm 38 4 6 6 Noise Figure of LNA Noise Figure is a measurement of the degradation of Signal to Noise Ratio SNR For a cascaded system the Noise Figure can be calculated using the following equation R 1 1 F Equation 2 Noise Figure Using typical ERA 3 datasheet values at f 1 GHz 21 db 125 89 Also E 2 6 dB or 1 8197 F 1 8197 1 8262 2 62 dB 12589 7 I Equation 3 Calculated Noise Figure of the Low Noise Amplifier Equation 3 shows the calculated Noise Figure for the LNA at f 1 GHz Since the datasheet value doesn t change much over the 0 01 1GHz range similar performance can be expected at 437 MHz The Noise Figure of the LNA was measuring using the setup shown in Figure 25 F Hot F Cold Dc 0 e
44. 73 Table 24 RSSI from both COMMS of CDH Rev 5 compared to datasheet 74 Table 25 CP5 Flight Candidate receive sensitivity performance esee 74 Table 26 CP5 Flight Candidate resting RSSI values eene 75 Table 27 Results of the noise measurement at the RX frequency of CDH Rev 4 and CC1000 development board The noise of COMMA at the RX frequency is 10 dB greater than the CC1000 development board and 5 dB higher than COMMB possible causing the difference in measured sensitivity between COMMA and 83 Table 28 Results of the noise measurement at the RX frequency of CDH Rev 5 and CC1000 development board The noise of COMMA at the RX frequency is 20 dB greater than COMMB possible causing the difference in measured sensitivity between COMMA xiii COMMB The 512 of the LNA Figure 50 is 30 dB reducing the LO leakage Section 6 1 and any noise between the output of the LNA and the input to the CC1000 Fiste DO pe 84 Table 29 Conclusions from the noise measurements of the receive 88 Table 30 Testing for differences in sensitivity caused by DC DC converters No significant difference in receive sensitivity is observed without the DC DC converters eliminating them as suspect in causing desensitization of the COMM sy
45. MA cable or U FL pigtail connect the CDH antenna jack to the SMA bulkhead for Port 1 of the Power Splitter IMPORTANT If using the U FL pigtail ensure that it actually connects to the antenna jack Continuity check the inner conductor 122 11 The test setup is complete Verify everything is connected It should look like the figure below Spectrum Analyzer HP 8566B LNA Computer 437 MHz 1W Splitter Variable Attenuation Attenuation Yaesu FT 847 Radio Faraday Cage Oscilloscope 12 Remove the RBF pin Set the variable attenuator to 24 Send the satellite a beacon command 0201 and verify that the satellite responds This is to make sure the setup is working before you close the lid 13 Place the lid on top of the box and clamp it down Make sure all unused ports are terminated 14 Send 473c to set the CDH in normal ops Verify that it ACKs Send 4800 to disable the beacon Finally send 410005 to set it to COMMA for approximately 30 minutes Make sure the radio is tuned to COMMA 123 Testing 15 16 17 18 19 20 Open up the original Sensitivity_Test excel spreadsheet and save it to the directory as a new spreadsheet name it with the correct date and board revision Eg 2010 03 02 CDH REVA SensitivityTest Verify the sensitivity test setup is connected properly Spectrum Analyzer HP 8566B LNA Computer 437 MHz
46. Receive Sensitivity Characterization of the PolySat Satellite Communication System A Thesis Presented to the Electrical Engineering Department Faculty of California Polytechnic State University San Luis Obispo In Partial Fulfillment Of the Requirements for the Master of Science Degree in Electrical Engineering By Ivan M Bland March 2010 2010 Ivan Bland ALL RIGHTS RESERVED ii TITLE AUTHOR DATE SUBMITTED COMMITTEE CHAIR COMMITTEE MEMBER COMMITTEE MEMBER COMMITTEE MEMBERSHIP Receive Sensitivity Characterization of the PolySat Satellite Communication System Ivan M Bland March 2010 Dennis Derickson Dr Jordi Puig Suari Dr Bryan Mealy Dr iii ABSTRACT Sensitivity Characterization of the PolySat Satellite Communication System Ivan M Bland Following the successful launch of CP3 and CP4 the PolySat team noticed an unreliable uplink to both satellites A significant problem with the PolySat COMM system is poor receive sensitivity of the communications system Efforts have been made to improve the uplink margin but without proper characterization of the receiver sensitivity the problem cannot be fully addressed By developing an accurate method of measuring receive sensitivity a methodical approach can be used to properly diagnose the communication system and link budget Two revisions of the PolySat COMM system will be measured and compared An in depth study of the PolySat
47. UT is easily measured 29 Figure 17 Low Noise Amplifier used to increase sensitivity of the Spectrum Analyzer The resistive splitter divides the RF path and the Spectrum Analyzer is used to measure the amplitude of the signal reaching the satellite in the Faraday Cage esses 30 Figure 18 Schematic of custom built EN unde pee ote epit eer cia be ake 3l Figure 19 Construction of insane nae Pra P uu etes d e Peau cedi 32 Figure 20 Miniature Faraday Cage surrounding LNA eene 32 Figure 21 gt Sri OPENA C c m 34 Figure 22 Soo Of LNA nusuri o Beta E E EE eae 35 Figure 23 s Forward Gain Sz of LENA au eo 36 Figure 24 Reverse Gain flue one ene one 37 Figure 25 The test setup used to measure the Noise Figure of the Low Noise Amplifier nO hiss E EEE nm 39 Figure 26 Measured noise figure over the 430 450 MHZ range eee 41 1X Figure 27 The Noise Figure of the cascaded system LNA and the Spectrum Analyzer approaches the Noise Figure of the first stage of the cascaded system LNA This increases the sensitivity of the Spectrum Analyzer 42 Figure 28 Noise Figure F of the two stage cascaded system LNA and Spectrum Aa yet feo bus tod idoli veis Fees ll vido itus oe co dus
48. WM 6 DS 24D S saeua UR GROSS Pumas RI POR IUe 6 DS 3 2 CD3 s o SECUN PIS BRE 8 234 CPA a ve E E HU INR Om P BEI ee 8 DS DOD ee sare EE ie E ENEO ee 9 Z0 CPO Senna E ERU PINE te inte Sadan ese dan P BI 9 PST MERC 10 dU De idee hue 12 CHAPTER 3 COMMUNICATION PROBLEMSS c eeeeee eese ente seen enne ta sns tn sens tastes suse ta sone sees sosta sene 13 S LUNRELIABEE UPEINK ederet a ie en ee P OU eee s 13 3 1 1 Inability to Communicate with Satellites eee eese esee eene 13 3 12 DOWRIIRRE seien eo eere 14 3 2 RECEIVE SENSITIVITY doeet torte rl dei ne ie E eo I dde ebore Rok ote E dna eodein gd 14 3 2 T Antenna Reciprocitysa du oo Redes eise e ee be Waning ONIS E reve eee RUE ERA Dresden rete 14 3 2 3 Previous Receive Sensitivity Testing iisdem eet teet deese e ee dee et debent 15 3 2 3 Link Budget Uplink ded etd edm ond d cede e eerte ee det dete 17 3 2 4 Link Budget Downlink iaa tse amete tied deeem re ead 18 3 3 ADDRESSING THE UNRELIABLE UPLINK ieeseeeeeeeeee eene nennen een erre nre nnne 19 3 3 1 Preview of Results Receiver Sensitivity Testing esses eterne nennen teens 19 3 3 2 Additional Testing Performed and Problems Discovered eene 20 3 3 3 Documentation of the Sensitivity Mea
49. a A due to reciprocity allows easy reception of the signal The system can also be described from antenna B s point of view PolySat s ground station can easily decode data dumps from CP3 and CP4 If communication can occur one way why is the uplink so difficult to close Since successful downlink shows that the antenna system is more than adequate this suggests the CP Bus receiver is much less sensitive than the ground station receiver While this supports the theory of poor receive sensitivity without an accurate measurement of both receivers sensitivity no conclusion can be made 3 2 3 Previous Receive Sensitivity Testing After the completion of the CP Bus for CP2 the team performed range testing to verify the performance of the COMM system According the link budget a satellite at around 2500 km will have a path loss of approximately 153 dB The satellite was taken to a location 6 km away approximately 100 dB attenuation and the output of the ground station was attenuated by 50 dB 3 While this test provides valuable information it does not verify the CC1000 as capable of its stated receive sensitivity This type of testing lacks a method of accurately determining the power reaching the satellite providing little information about the sensitivity of the CC1000 Additionally it is difficult to measure the power output of the ground station radio to verify the attenuation is reducing the signal to the desired level RF leakage is
50. able of housing an entire P POD or a 3U satellite with small antennas extended More importantly the box can accommodate a satellite with umbilical box which can provide the 3V and 4 2V voltage rails required to power the CDH 51 Figure 36 Spacious interior allows testing on larger satellites 3U if necessary CP3 TestSat is attached via the U FL antenna connector with CDH Rev 5 with the LNA amplifier same CDH flown on CP6 To pass signals in and out of the enclosure 4 SMA bulkheads and 1 N Type bulkhead were mounted 50Q terminations can be attached to the bulkheads not being used to ensure radiation does not leak through the SMA and N Type bulkheads Figure 37 N Type and SMA pass through ports When not in use each port should be terminated with a 500 termination to prevent the possibility of leakage 52 Although the N Type and SMA connectors can pass DC if necessary dedicated DC pass through ports were also installed Since any modifications to the box could compromise the isolation feedthrough capacitors were used Feedthrough capacitors shunt any AC signal to ground while passing DC signals Two separate pass through ports were necessary since the current CP Bus utilizes two voltage levels 3 0V and 4 2V Table 16 Feedthrough capacitors capable of passing DC power Ground clips soldered to cage To minimize the chance of RF leakage SMB connectors were connected to the feedthrough capacitors This all
51. ansceiver an accurate statement of the COMM system s performance can be made Further testing can identify causes of poor sensitivity providing valuable information for developing a more robust COMM system 13 3 1 2 Downlink While the team found it very difficult to get commands through the ground station could easily decode packets from the satellite indicating the receive sensitivity of the ground station to be more than adequate 3 2 Receive Sensitivity Several observations support the theory that the CP Bus has poor receive sensitivity Antenna reciprocity successful downlink and on orbit data suggest that closing the uplink requires increasing sensitivity of the CP Bus 3 2 1 Antenna Reciprocity Antenna reciprocity states that the gain of an antenna will be the same in transmit and receive directions If a communications link can be established between two antennas due to antenna reciprocity the communication can occur both ways Either antenna can be the transmitter or receiver regardless of differences between the two antennas A tx B rx A B E B tx A rx High Gain Low Gain Yagi Uda Omni directional Figure 9 Antenna Reciprocity 14 Using the Figure 9 to illustrate reciprocity Antenna A is a high gain directional antenna Antenna B is a low gain omni directional antenna With antenna A transmitting the high gain provides a lot of power to antenna B When antenna B transmits the high gain of antenn
52. anually entered in the MixW software inferface The MixW software serves as the Terminal Node Controller TNC and interfaces to 24 the Yaesu FT 847 transciever With a power output of 1W the RF signal is reduced to 0 dBm as it travels through the 30 dB attenuator The variable attenuator can further reduce the signal by up to 110 dB shown in Section B of Figure 12 A resistive power splitter equally divides the signal into two paths Section C of Figure 12 Half of the signal goes to the HP8566A Spectrum Analyzer for measurement where a Low Noise Amplifier LNA is used to increase the sensitivity of Spectrum Analyzer Section D of Figure 12 The other half goes to the isolated satellite in the Faraday Cage Section E of Figure 12 Since the signal is equally split the Spectrum Analyzer is used to determine the signal reaching the satellite A power meter was used to verify the accuracy of the Spectrum Analyzer The accuracy of receive sensitivity measurements obtained with the system shown in Figure 12 is 2 dB The relative accuracy comparing two measurements obtained with the test setup 15 2 dB Using the variable attenuator the signal magnitude can be reduced until the satellite no longer responds establishing an approximate threshold of sensitivity 25 Figure 13 Sensitivity Measurement setup this shows all the necessary components required for measuring sensitivity This setup was used to measure the sensitivity of two
53. ards to the RF input or also to the IF port Since the IF of the CC1000 is 150 kHz the LO leakage is a weak signal 58 dBm 150 kHz away from the RX frequency LO leakage common in all receivers is important to note otherwise it could be mistakenly identified as noise on the PolySat COMM system RX mode gt f a low side far fio high side Lo Receive frequency Lo d f gt q fe Figure 56 The LO will be 150kHz above the RX frequency for high side injection and 150 KHz below RX for low side injection For CDH Rev 4 and 5 with high side LO injection the LO leakage occurs at 437 515MHz The CC1000 development board programmed at 434 010 MHz utilizes low side injection so the LO leakage is at 433 860 MHz 81 6 1 1 Receive Line of CC1000 Noise Floor at RX TX Frequency The first test performed was a comparison of the different COMMs at the receive frequency of interest For the CDH boards this was 437 365 MHz The CC1000 development board was set to 434 010 MHz By using a small span of 1MHz noise at the receive frequency could be compared The CC1000 development board provides an excellent baseline standard to compare CDH COMMSs to 436 860 437 118 437 368 437 610 437 860 Figure 57 CDH R4 CommA noise floor centered at the PolySat COMM receive frequency 437 365 MHz LO leakage visible at 437 515 MHz RBW 10 kHz Crystal harmonic also visible 436 868 437 1108 437 368 437 616 437
54. at 2 GHz and less than 25 dB at 3 GHz this gain is still great enough to greatly increase the Spectrum Analyzer s sensitivity If the setup is used at a 36 higher frequency the change in forward gain must be noted This is because the gain is subtracted off to determine the actual signal going to the Faraday Cage during system calibration 4 6 5 Reverse Gain of LNA 0 5 1 1 5 2 2 5 3 dB frequency GHz Figure 24 Reverse Gain 81 of LNA se f GHz dB 0 4286 51 056 0 436075 51 034 0 44355 51 236 0 451025 51 368 Table 10 S across the 70 CM Amateur Radio band The Reverse Gain is less than 50 dB across the 430 450 MHz amateur band indicating a high degree of reverse isolation 4 6 5 1 dB Compression Point 37 As the input signal to an amplifier is linearly increased at some point the output no longer increases linearly If the input signal continues to increase the amplifier saturates or can no longer amplify the signal The 1 dB compression point indicates the power level which causes the amplifier gain to decrease by 1 dB 1 This can be referred to either the input or output 21 TRANSMISSION TRRNSHISSION REFLECTIQN CH 1 321 REFERENCE PLANE 8 86008 mm LOG MAGNITUDE REF215 888 dB 3 888 dB DIV 1 35 00 dBm 42 219 dB Ru XII Sg ORAE Ls PPS 41 223 dB 3 25 59 dBm 39 248 dB 4 28
55. at a particular data rate Higher data rates of a given receiver result in reduced sensitivity For the testing performed in this thesis the receive sensitivity is defined as the weakest signal command dBm the PolySat COMM system can decode and respond to During testing the signal strength was incrementally reduced until the COMM could no longer decode the signal After this threshold was determined the strength of the last successfully decoded command is considered the receive sensitivity For 22 the senstivity measurements of this thesis the receive senstivity is defined as the weakest signal command dBm of which the PolySat COMM system could respond to 4 2 Test Setup Components The receive sensitivity test setup consists of six main components an RF source an attenuation stage with variable attenuator a resistive splitter a Low Noise Amplifier LNA a spectrum analyzer and a custom built Faraday Cage An overview of the system is shown in Figure 12 23 Computer Spectrum Analyzer HP 8566B 43 2 dB LNA Rigblaster A C B 30 dB 0 110 dB 437 MHz 1 1 1 Attenuation Variable Attenuation Yaesu FT 847 Radio Faraday Cage Oscilloscope Figure 12 System level overview of the setup developed used to measure the receive sensitivity of the PolySat COMM system The Yaesu radio connected to a computer via a RigBlaster sends a command to the satellite Section A of Figure 12 Commands are m
56. atellite CP1 began in 2000 The payload included sun sensor donated by Optical Energy Technologies and a magnetic torquer magnatorquer embedded in a side panel 19 Although the Dnepr launch vehicle failed CP1 proved that a team of students with no satellite building experience could complete the entire cycle of satellite development Additionally the satellite was completely built with Consumer Off The Shelf COTS components This is a typical characteristic of CubeSats to reduce development cost Figure 3 CP1 satellite 2 3 2 CP2 CP2 marked Cal Poly s second satellite and first attempt at standardizing the satellite bus A satellite bus is the infrastructure of the satellite encompassing all major systems Command and Data Handling C amp DH Electrical Power System EPS and the Communication System COMM Called the CP Bus it consisted of the C amp DH board the EPS board and side panels with solar cells By standardizing the bus payload development and integration would be easier CP2 is considered the first version of the standardized CP Bus and built upon the lessons 6 learned from CP1 The satellite bus contains all the major subsystems communication system command and data handling power generation and storage and the mechanical structure The satellite payload is interfaced to the bus For CP2 bus improvements include triple junction solar cells dual 1950 mAh Li Ion batteries and Magnetorquers attitud
57. ature range 14 The RX TX frequencies are set by a Phase Lock Loop PLL set from the crystal oscillator so any drift of the crystal will cause the receiver and transmitter center frequencies to drift Crystal Inputs Initial tolerance 30 ppm Temperature drift 50 ppm Aging 5 ppm Load error 2 ppm Total ppm Table 35 Total variation of crystal over a large temperature range calculated from application note AN019 available from Texas Instruments Total frequency error possible for TX kHz Total frequency error possible for RX kHz Worst case IF frequency error kHz Table 36 Resulting IF error caused by temperature variations Table 36 shows that a variation of up to 70 3 kHz of the IF is possible The IF tracks the LO with a 150 kHz offset and crystal frequency variations caused by temperature will While this 101 represents the worst case the bandwidth of the IF is 175 kHz and a variation of 70 kHz could shift the 150 kHz IF closer to the edges of the IF filter Sensitivity vs IF frequency for CC1000 CC1010 typical Sensitivity loss dB O Relative IF frequency error kHz Figure 77 Variations of the IF due to the crystal can cause a reduction in the sensitivity of the CC1000 Figure 77 shows that sensitivity loss can occur with large variations of the IF This chart is for a frequency
58. ave already connected the hardware and click next again Ad Sere or s Click Add a new hardware device and click next Select Install the hardware that manually select from a list Advanced click next Click on Select ComEmulDrv inf from the folder you extracted it into Click Next gt fat the next screen where MixW serial port bridge is selected and click _Next gt sain to begin the installation progress Click on Continue Anyway when Windows complains about the MixW serial port bridge Click Save files for future reinstall Open Start Settings Control Panel System and click on the Hardware tab 15 16 Click the Device Manage Jputton Expand Multi port serial adapters and double click on Felix serial port bridge will bring up the MixW serial port bridge Properties Select the Properties tab You will be presented with the MixW Serial Port Bridge Properties Set Pair 1 to COM5 and COME Set Pair 2 to COM7 COM8 This 126 MixW serial port bridge Properties General Proves Driver Deals COMB 19 Reboot the computer 20 Add Uplink and Downlink shortcuts for easy access 127 Appendix List of Acronyms AFSK Audio Frequency Shift Keying ATL Advanced Technologies Laboratory CDH C amp DH Computer and Data Handling COMM Communications system COTS Consumer Off the Shelf Components CP Bus Cal P
59. ceiver The CC1000 is a single chip UHF transceiver designed for low power applications Mainly intended for Short Range Devices SRD in the Industrial Scientific and Medical ISM bands it can be programmed for operation from 300 1000 MHz 8 Figure 47 High level circuit diagram of the CC1000 In receive mode the CC1000 is configured as a traditional superheterodyne receiver The RF signal is amplified by an LNA and mixed down to the 150 kHz Intermediate Frequency IF stage for filtering and demodulation Although not explicitly stated in the datasheet it is implied that the LNA is broadband 300 1000MHz and no filtering occurs until the IF stage The 65 Receive Signal Strength Indicator RSSI is determined by measuring the raw power independent of modulation format at the IF stage Two main control lines DIO and DCLK interface to a microcontroller for the exchange of demodulated digital data The frequency synthesizer used to set the LO and RF output frequencies consists of a Phase Lock Loop PLL 8 The PLL uses a 14 7456 MHz crystal oscillator as a reference 5 1 3 Receive Sensitivity versus Frequency Separation and Data Rate The receive sensitivity of the CC1000 is a function of several variables frequency data format data rate and Frequency Shift Keying FSK frequency separation Before a conclusion on the performance of the transceiver can be made the typical or expected performance must be gleaned from th
60. ch means it won t load down the circuit being measured at higher frequencies Verification of the RF chain could be easily accomplished by applying an RF source at the antenna connector and probing at different points A used probe can be bought off of eBay for around 1 200 109 8 Summary 8 1 Receive Sensitivity Test Setup 8 1 1 Sensitivity Characterizations The test setup developed to measure receive sensitivity Figure 12 worked very well It is considered successful for several reasons First the setup is capable of making accurate and consistent sensitivity measurements Second the setup was used to characterize the COMM system of the CP Bus which was incompletely tested on earlier satellites The data collected provided valuable information to assess the capability of the COMM system Finally the test setup is versatile capable of being utilized in future receiver development It can also act as a tool for selecting the most sensitive board among flight candidates 8 2 CDH Sensitivity Performance 8 2 1 CDH Rev 4 versus Rev5 The addition of the LNA on CDH Rev 5 successfully increased the sensitivity of the CP Bus Compared to the link budget however the sensitivity of the CP Bus is still not enough A more sensitive transceiver is needed CDH Rev 4 90 dBm CP3 CP4 CDH Rev 5 100 dBm Includes LNA CP6 Yaesu FT 847 115 dBm Ground Station Table 38 Measured receive sensitivities of CDH Rev 4 Rev 5 and
61. chniques could greatly decrease this noise and increase the receive sensitivity C The current testing procedures for qualifying a CDH board as fully functional are not sufficient During testing it was found that a fully functional board had a serious electrical failure The U FL connector which connects the antenna to the COMM system has a tendency to fail resulting in an open circuit The current testing procedures 20 should be updated to include measuring the sensitivity of each CDH board and to eliminate the possibility of manufacturing defects of the COMM system Measuring the sensitivity of a completed satellite can verify system performance before launch D A long duration COMM test showed that the uplink is only approximately 70 successful even with favorable links conditions Software testing should be done to determine the cause of this E The test setup outlined in this paper should be used as part of developmental testing of the new bus and final system check out It is critical to identify sensitivity problems during development rather than after the satellite is completed and launched By using this setup potential problems affecting COMM sensitivity can be identified and resolved early in development 3 3 3 Documentation of the Sensitivity Measurement Setup Design and Testing Performed The remainder of this thesis builds upon on the aforementioned preview of conclusions A detailed discussion of the test setup
62. dBm Figure 66 While this actually performs better than the CC1000 development board Figure 67 it s important to note that the S12 of the onboard LNA of CDH Rev 5 is 30 dB 4 possibly blocking a lot of that noise The LO leakage missing from the noise profile of COMMB in Figure 66 is also suppressed by 30 dB 200 000 309 000 400 000 509 000 Figure 67 CC1000 Development Board 100 500 MHz RBW 1MHz Crystal harmonics repeated every 14 74 MHz and LO leakage visible at 433 860 MHz 109 000 200 000 300 000 400 000 500 000 Figure 68 System Noise Floor 100 500 MHz RBW 1MHz LNA and Spectrum Analyzer The average broadband noise of the CC1000 development board is around 60 dBm slightly higher than the system noise floor 1 Broadband noise was measured to be higher on CDH Rev 4 and Rev 5 compared to the CC1000 Development board 2 The measured noise levels at the RX frequency of COMMA were greater compared to COMMB indicating a possible explanation of better measured sensitivity of COMMB 3 In some cases the crystal harmonics of CDH Rev 4 were 15 20 dB greater than the crystal 87 harmonics of the CC1000 development board This is most likely caused by not following the recommended layout Table 29 Conclusions from the noise measurements of the receive line Table 29 shows conclusions derived from measuring the noise floor of CDH Rev 4 Rev 5 and the CC1000 development board A new layo
63. difference in measured sensitivity between COMMA and COMMB 83 For comparison purposes the system noise floor is included By terminating the LNA with 50Q load the noise floor of the Spectrum Analyzer and LNA can be observed Figure 60 437 116 437 360 437 610 437 860 436 860 437 118 437 368 437 616 437 866 Figure 62 CDH Rev 5 Comm centered at receive frequency 437 365 MHz RBW 10 kHz COMM System Noise at RX frequency Comments CDH Rev 5 COMM A 70 dBm Measured sensitivity 94 4 dBm CDH Rev 5 COMM B 90 dBm Measured sensitivity 101 9 dBm CC1000 Development Board 75 dBm Figure 61 System Noise Floor 85 dBm Figure 62 Table 28 Results of the noise measurement at the frequency of CDH Rev 5 and CC1000 development board The noise of COMMA at the RX frequency is 20 dB greater than COMMB possible causing the difference in measured sensitivity between COMMA and COMMB The of the LNA Figure 50 is 30 dB reducing the LO leakage Section 6 1 and any noise between the output of the LNA and the input to the CC1000 Figure 50 84 Recalling from the measured receive sensitivity testing Section 5 2 3 it was noted that the measured sensitivity of CDH Rev 5 COMMB was 7 5 dB more sensitive than COMMA Comparing the noise profiles centered at 437 365MHz of COMMA and COMMB Figure 61 and Figure 62 respectively a possible explanation of decreased sensitivity is that the noise floor
64. e Attenuation Screen Room Room 116 60 Screen Room Room 118 50 EMC Chamber 60 Custom Built Faraday Cage 70 Table 17 Comparison of attenuation of each screen room to the custom built Faraday Cage The cage offers 10 dB more isolation which was enough to isolate the satellite from stray radiation The Faraday Cage effectively attenuates signals at 437 MHz by at approximately 70 dB At first it was thought that the Faraday Cage must attenuate signals by at least 153 dB which is the worst case path loss for a typical CubeSat orbit However the Faraday Cage s purpose is not to simulate the orbital path loss It is merely to prevent stray leakage from the radio connectors coax etc from triggering the COMM system For this purpose the Faraday Cage performs very well 4 9 Other Considerations 4 9 1 RF Leakage All the components of the setup are connected with lengths of semi rigid coax Compared to regular braided coax the chance of RF leakage is smaller Semi rigid coax offers 100 shielding whereas braided coax does not Previously 10 dB and 20 dB attenuators were connected back to back to reduce the approximately 30 dBm output of the Yaesu radio to around 55 0 dBm However after characterization some of the data appeared erroneous After re characterizing the system shown in Figure 12 it was found that RF leaks can bypass attenuators placed back to back This was only observed to be a problem at higher power l
65. e control performed by magnets generating a torque against the Earth s magnetic field in each side panel The C amp DH featured a redundant COMM system and an PC bus for data transfer 3 A PIC microcontroller served as the brain The EPS contains several Smart Fuses which are fuses designed to reset once a fault is cleared protecting from single event failures Independent DC DC converters provide regulated voltage rails for each system 2 To date the current CP Bus is still based off this design but with slight hardware revisions made to both C amp DH and the EPS Hardware revisions have ranged from simple updating layout for new components to system level improvement adding a Low Noise Amplifier to the COMM 4 Figure 4 CP2 Flight unit Besides flight testing of the CP Bus CP2 contained an energy dissipation experiment CP2 never reached orbit due to the Dnepr 1 launch vehicle failure 2 3 3 CP3 The primary mission of CP3 was attitude determination and control 19 The side panels were equipped with 2 axis Magnetometers allowing measurement of the Earth s magnetic field Coils of wire embedded in the inner layers of the PCB served as Magnetorquers By passing current through the coils a magnetic field could be generated to torque against the Earth s magnetic field CP3 s payload also contained imaging sensors to photograph the Earth CP3 was launched in 2007 on Dnepr2 Figure 5 CP3 Flight unit 2 3
66. e datasheet For many of the possible configurations the expected receive sensitivity is listed in tabular form in the datasheet With the PolySat COMM system utilizing the 433 MHz band and using Non Return to Zero Inverse NRZI encoding the data rate and frequency separation are the final variables which determine the actual receive sensitivity 8 66 Data rate Separation 433 MHz kB aud NRZ made LI LI LI LI l i pem 74 de c Data rate kBaud i qp inte E C J ra c ok 3 Table 18 CC1000 receive sensitivity is a function of frequency data format data rate and frequency separation 8 This table is obtained from the CC1000 datasheet The red highlighting indicates that the uplink to the CC1000 is 600 baud data rate Since the COMM system of the CP Bus uses Audio Frequency Shift Keying AFSK rather than true FSK the frequency separation is 2 kHz rather than the recommended 64 kHz Using a frequency separation of 2 kHz is not a recommended configuration of the CC1000 the datasheet recommends for best sensitivity to keep the frequency separation as high as possible 7 8 Without datasheet information for 600 baud data rate and 2 kHz frequency separation an intuitive guess must be made In the table above a comparable data rate to frequency separation ratio is 4 8 kbaud and 20 kHz separation resulting in a receive sensitivit
67. e of the satellite a small enclosure could be made which provides portability advantages A concept of the enclosure is shown in Figure 33 Lid PCB Sides Figure 33 Concept of Faraday Cage construction 49 Constructing RF proof enclosure is relatively easy but usually the opening is the most difficult part in terms of RF isolation Any RF penetrating the enclosure will typically find its way in through the opening or lid For simplicity it was determined that an effective and sealable opening could be implemented by constructing a rectangular box with no top and adding small flanges perpendicular to the sides Then a lid could be rested on top of the enclosure and sealed by attaching clamps to hold the lid tightly to the enclosure s flanges Finally small clamps could apply pressure providing a tight seal Figure 34 Unetched PCB being soldered together to form the enclosure Each seam was soldered completely and copper tape applied to the top flanges and most of the seams 50 Figure 35 Completed cage using clamps to seal the lid Coax is seen attached to each RF pass through Since a larger sheet of PCB has some flex attaching clamps helps provide a better seal With the lid totally sealed the isolation at 437 MHz is easily sufficient to isolate a satellite 4 8 4 Features The box was built to be 2 long by 1 wide and 1 tall This is small enough to allow desktop operation but still cap
68. en and closed provided an approximate level of attenuation Screen Room Approximate Isolation dB Room 116 Building 20 60 Room 118 Building 20 50 EMC Chamber Arakaki s Lab 60 Table 14 Comparison of the attenuation of each Screen Room Each room does not provide enough isolation at UHF The only alternative was to build a custom Faraday Cage After building the cage a huge advantage was noted sensitivity testing requires a lot of resources from the PolySat lab Sensitivity testing is most efficiently conducted inside the PolySat lab In other words it would have been very inconvenient to conduct testing outside of the PolySat lab 47 4 8 2 Custom Faraday Cage According to Gauss Law an electric field will not penetrate a perfectly conducting enclosure Instead the current will reside at the top of the conductor depending on the frequency The higher the frequency the more concentrated the current will be at the surface of the conductor This is known as the skin effect The skin depth dependent on frequency can be calculated using the equation below 1 p Tho Urf Equation 6 Skin depth formula skin depth in meters Ho 4nx10 7 H m relative permeability 0 999994 for copper p resistivity of the medium in Q m 1 72 x 10 for copper f frequency of the wave in Hz Table 15 Variables affecting skin depth of material After simplifying the above e
69. er visual inspection of the connector should also be performed see Figure 80 1 Visual inspection of female U FL connector See Figure 80 2 Limiting the number of connections made 3 Use proper tool to mate male and female U FL connectors 4 Continuity check with multi meter Table 37 A list of possible ways to mitigate open circuit failure of the female U FL connector Although past testing procedures may not have detected this testing each finished board using the receive sensitivity test setup Figure 12 could easily identify a problem in the RF chain The U FL connector s small size makes it desirable to use in CubeSat applications The desirable features warrant continued use as long as precautions are taken to prevent failures 107 7 2 4 Contract Manufacturing of Assembled boards Each CDH board is soldered by students in the PolySat lab This is a several step process First the necessary parts are gathered a process called kitting Then each component is hand soldered using a fine tip soldering iron and a microscope called population Finally the board is tested to make sure it works There are several drawbacks to this approach the biggest of which is reliability It can be very easy to make errors during component population of the board such as using the wrong capacitor Troubleshooting a non working CDH board is also very difficult As discovered through sensitivity testing the current testing pr
70. er 4 69 RF_TXA_IO gt Figure 50 RF Chain of COMM A on CDH Rev 5 The only difference is the addition of an LNA associated matching network and high pass filter Comparing CDH Rev 5 Figure 50 to CDH Rev 4 Figure 49 the differences on the receive line include the LNA matching elements and a high pass filter 5 2 COMM Sensitivity Testing 5 2 1 Overview of Testing Using the sensitivity test set up developed shown in Figure 12 the receive sensitivity of both CDH Rev 4 and Rev 5 boards was measured During CP2 and CP6 development only estimates of sensitivity could be made 4 so a direct measurement is a huge step forward in gauging performance of the COMM system Comparing sensitivity measurements to the expected performance of the CC1000 the actual performance of the transceiver can be evaluated Directly comparing the measured sensitivity between CDH revision 4 and 5 shows the increase in sensitivity due to the LNA For both revisions the sensitivity of COMM A and B were individually measured The Yaesu FT 847 transceiver utilized during satellite passes was also tested for receive sensitivity 70 5 2 2 CDH Testing Measuring the Receive Sensitivity Following the procedure developed for testing receive sensitivity the CDH boards were tested for receive sensitivity The testing procedure for measuring receive sensitivity of the PolySat COMM system CDH Rev 4 and 5 is outlined in Section 4 11 and detailed
71. er decode the commands from the Yaesu FT 847 Section A of Figure 12 An example sensitivity test is shown in Figure 44 61 Sensitivity Response Threshold of Sensitivity Actual Signal Successful Response dBm No Response Attenuator Setting dB Figure 44 Sensitivity test of the PolySat COMM system The sensitivity threshold is 100 7 dBm corresponding to the last successfully decoded command indicated by a satellite response Once the threshold of sensitivity of the PolySat COMM system is found the LNA and Spectrum Analyzer Section D of Figure 42 can be reconnected to the sensitivity test setup as shown in Figure 45 62 Computer Spectrum Analyzer HP 8566B 43 2 dB LNA Rigblaster Splitter 30 dB 0 110 dB 437 MHz 1W Variable Attenuation Attenuation Yaesu FT 847 Radio Faraday Cage DC Power Figure 45 Sensitivity measurement after the threshold of sensitivity is found The LNA is used to verify the signal strength of the Minimum Detectable Signal MDS which is the weakest signal the PolySat COMM system is capable of responding to This is the measured receive sensitivity and is accurate to 2 dB With the LNA and Spectrum Analyzer connected to the sensitivity test setup as shown in Figure 45 the LNA is used to verify the signal strength dBm of the minimum detectable signal establishing a measured receive sensitivit
72. erformed the other COMM CC1000 RSSI 9 COMMA RSSI i RSSI Datasheet RSSI Table 21 RSSI Characterization of CDH Rev 4 The RSSI values correspond closely to the datasheet until the signal strength drops below approximately 80 dBm 5 2 4 CDH Rev 5 Measured Sensitivity Testing The CDH Rev 5 used for testing was one of the several boards assembled as flight candidates for CP6 Additional sensitivity data was obtained by testing a flight candidate board 72 built for 5 a satellite testing a de orbiting mechanism 5 is currently under development allowing time for possible COMM system upgrades for improved sensitivity CDH Rev 5 Sensitivity COMMA 94 4 dBm COMMB 101 9 dBm Table 22 Sensitivity of CDH Rev 5 with preamp tested using the sensitivity measurement setup in Section There is a significant difference between COMMA and COMMB as COMMB is 7 5 dB more sensitive A possible reason for this is discussed in Section 6 1 1 By observing the RSSI value with a 500 termination at the input of the CDH dubbed the resting RSST any large deviation from the expected indicates noise seen by the CC1000 Resting RSSI and Signal Strength COMMA 0 775 85 dBm COMMB 0 998 V 100 dBm Expected 1 1V 105 dBm Table 23 Resting RSSI of CDH Rev 5 Table 23 shows the CC1000 of COMMA detects signal strength of 85 dBm even
73. eshold of sensitivity to recreate the error Problem 4 in Table 31 during the long duration test This may provide more information It is not clear what is causing the satellite to not 96 respond to commands More long duration tests could possibly reveal a pattern Additional software testing should be performed to help troubleshoot this problem 97 7 Results of Sensitivity Testing 7 1 Comparison of COMM Sensitivity 7 1 1 Overall COMM Performance CDH Rev 4 typically responded to packets down to around 89 to 92 dBm while CDH Rev 5 performed significantly better responding to commands down to 100 to 101 dBm The LNA provided a significant increase in sensitivity A summary of each board tested can be seen in Table 33 Sensitivity Board Comments COMMA COMMB CDH Rev 4 89 dBm 92 dBm CDH Rev 5 1 94 dBm 101 dBm LNA increases sensitivity CDH Rev 5 2 96 dBm 100 dBm CP5 Flight Candidate Table 33 Overview of the performance of COMM tested for sensitivity 7 1 2 CC1000 Performance The datasheet sensitivity for the CC1000 is 110 dBm FSK modulation data rate of 2 4 kBaud with a separation frequency of 64 kHz The wide separation frequency is a potential problem with the transceiver For best sensitivity the separation frequency needs to be 64 kHz and the datasheet explicitly states this 8 Since the AX 25 standard is used for packet communication with Audio Frequency
74. evels greater than 10 dBm A single 30 dB attenuator was used instead 4 9 2 Yaesu Radio RF Output Power The output power of the Yaesu FT 847 radio at minimum power setting was observed to be constant at one of two power levels 0 dBm or 19 5 dBm The output wouldn t change throughout testing but would be at either power level at startup Because of this it is very important to measure the output of the radio during each test This is considered an anomoly of this particular unit and is not yet fully understand 4 10 Sensitivity Measurement Setup Characterization In order to determine the amplitude of the signal reaching the DUT the sensitivity measurement setup shown in Figure 12 must be characterized Using the HP 8640B RF source a signal of amplitude 20 dBm at 437 MHz provided was fed directly to the setup The Spectrum Analyzer with LNA was monitored as the variable attenuator was varied from 30 dB to 110 dB The signal reaching the Spectrum Analyzer is the same amplitude of the signal reaching the DUT after subtracting off the LNA gain By characterizing the system the attenuator setting will correspond to a signal of a specific amplitude The characterization setup is shown in Figure 40 56 Spectrum Analyzer Splitter 0 110 dB HP 8640B RF Source 20 dBm Variable Attenuation Faraday Cage Figure 40 Characterizing the receive sensitivity setup For example if the attenuator is set to
75. ff the LNA gain This allows the user to accurately measure the RF signal going into the Faraday Cage Although the test setup does not change it is important to measure the signal strength during every test since variations in the output power of the Yaesu FT 847 can occur 4 6 1 Design of Low Noise Amplifier 30 12 OUT 1000pF 1000pF Figure 18 Schematic of the custom built LNA Using two drop in monolithic amplifiers and biasing components a two stage LNA was built Monolithic amplifiers are gain blocks with input and output internally matched to 500 Providing a gain of 42 3 dB it increases the sensitivity of the Spectrum Analyzer tremendously In Figure 18 capacitors provide a DC block while Cgypass filters any noise of the power supply The Rgyas resistors provide the required current and RF Chokes were added to prevent AC loading on the output Using Mini Circuits application circuit two gain blocks were cascaded to obtain a gain of 42 3 dB 31 Figure 20 Miniature Faraday Cage surrounding LNA To prevent amplification of unwanted noise or signals the LNA was completely enclosed with unetched FR4 PCB This is important since the actual signals are of such small magnitude ranging from 50 dBm to 110 dBm that any RF leakage from the RF source or radio could be 32 inadvertently amplified In order to supply the proper DC biasing a feed through capacitor was installed
76. g or software problems etc can be discovered early in the development process rather than after satellite development and launch The performance of flight candidate boards can be directly compared before building flight unit satellites for launch Chapter 1 discusses the scope of this paper and outlines the goals of the thesis Chapter 2 provides the reader with a background of the CubeSat program and an overview of Cal Poly s PolySat program Cal Poly s satellites and current projects are briefly described Chapter 3 outlines the motivation for receive sensitivity Evidence of poor receive sensitivity is described The link budget for a successful uplink is discussed giving the reader an idea of what sensitivity is necessary for a successful uplink Previews of the sensitivity testing results are included providing the reader with a quick look at the conclusions achieved This chapter serves to set the stage for an in depth discussion of the characterization that follows Chapter 4 describes the test setup developed to measure the receive sensitivity A discussion of the entire system provides the reader with a detailed analysis of the components of the test setup The difficulties of measuring receive sensitivity are also discussed A procedure for executing a receive sensitivity test can be found in the appendix In Chapter 5 Cal Poly s COMM systems are measured for receive sensitivity The first setup CDH Rev 4 is the COMM system flo
77. he board testing procedures especially for the RF Chain COMM Through RSSI characterization of the CC1000 a manufacturing defect was found in a board that was labeled as fully functional After each hand soldered board is completed student developed testing procedures are used to check each subsystem After each subsystem is tested the board is programmed with code and the COMM system is tested After the CDH Rev 5 was completed the testing procedures were not updated resulting in out of date procedures The outdated procedures were skipped and in this case a serious manufacturing defect was not detected A board s COMM system is deemed working if it can respond to a beacon command However even with 30 dB of attenuation on the radio the output power is high enough to trigger the response from the CC1000 just through spurious emissions from the transmitter inducing RF currents through the PCB traces 103 CC1000 5 Measured RSSI Datasheet RSSI 110 100 Figure 78 RF Chain testing of CDH Rev 5 Using the RF source a signal at 437 MHz was applied to the antenna connector The CC1000 is not receiving any signal through the antenna connector indicating a manufacturing defect in the RF chain Based on these testing procedures the CDH Rev 5 characterized in this thesis was qualified as a fully functional board However once placed in the Faraday Cage for sensitivity characterization a
78. iable Uplink 3 1 1 Inability to Communicate with Satellites After the successful launch of CP3 and CP4 the team experienced difficulty in sending commands to both satellites The main motivation for the development of a reliable test setup is spurred by the team s unreliable uplink to CP3 and CP4 During operations the team found it very difficult to get commands through but very easy to receive data from the satellite CP3 and CP4 pass over the PolySat ground station twice a day and a typical window for communication is approximately 6 10 minutes During this window the team sends a command every 3 5 seconds listening for a response between commands This means that during a typical pass approximately 140 160 commands are sent A good pass would be getting the satellite to respond to at least one command Getting a command through is random with no observable pattern In some cases the satellites would only respond once or twice in a two week period The communications system was revised and a more consistent uplink was observed for CP6 but the uplink was still marginal All three satellites show that the communication system is not adequate The poor uplink of the satellites in orbit is blamed on poor receive sensitivity of the CP Bus Without measuring the receive sensitivity the team lacks conclusive data on how well the COMM system is actually performing By measuring the sensitivity and comparing it to the expected performance of the tr
79. ier eese 39 Equation 4 Calculated Noise Figure of the Spectrum Analyzer see 41 Equation 5 Noise Figure of the Spectrum Analyzer and the LNA cascaded together 43 Eguatton6 Skin depth formula on noie ott ee etia P 48 Equation 7 Calculating the skin depth of copper of a 437 MHZ 1 48 Equation 8 Equation used to calculate Percent Error sese 58 Equation 9 Calculating the expected receive sensitivity of the Yaesu FT 847 The datasheet states that the minimum sensitivity is 0 125 uV so the minimum sensitivity in power dBm can be calculated uide e an ei a n e n d AVISAR TE RERO e aU dose ER ced 76 Equation 10 Calculating the thermal noise of the Yaesu FT 847 receiver in Single Side Band SSB Peera 76 XV Chapter 1 INTRODUCTION Scope of Thesis This paper will specifically address the characterization of receive sensitivity of the PolySat COMM system With increasingly demanding payloads on CubeSats a reliable COMM system is of the utmost importance By developing a method of testing receive sensitivity different revisions of the PolySat COMM system can be specifically measured and compared to the expected sensitivity Receive sensitivity problems caused by desensitivity poor layout transceiver interfacin
80. ificant software changes included more efficient uplink commands checksum value and error statistics of the PC bus and the ability to pause the sensor snapshot circular buffer 23 Figure 6 Engineering unit of CP6 in the cleanroom Launched May 2009 CP6 is considered a huge success More data was collected in the first few months from CP6 than the amount of data collected from CP3 in over a two year period Uplink to the satellite was more consistent suggesting that the LNA helped improve the receive sensitivity Before satellite succumbed to a CDH hardware failure in September 2009 the team was successful in de tumbling the satellite reducing the spin rate using its onboard attitude determination and control This is accomplished by using magnetometers to measure the Earth s magnetic field while coils of wire imbedded in the side panels can provide a torque against it The team also demonstrated the ability to spin up the satellite by reversing the process to increase the spin rate Additionally the team was able to turn on the payload and run a test to check the status of the payload Several pictures were taken 2 3 7 CP7 CP7 a particle damping experiment sponsored by Northop Grumman is currently under development Particle dampers consist of small metal cavities filled with tiny particles 10 reduced or zero gravity environment the particles would create a damping effect as the particles transfer momentum
81. in labeled 1 30 in Figure 76 shows the number of ACKs that occurred in less than 1 minute and 30 seconds Since the beacon command was sent every 60 seconds it is expected that the most responses occurred during the first bin 701 responses occurred between 1 30 and 2 30 indicating that not all commands received a response One response occurred during the 6 30 bin showing that at some point the satellite did not respond for over five and a half minutes Test Duration 70 hours Commands Sent 4199 Acknowledgements 2952 Successful Uplink 70 3 Table 32 Long duration COMM test results From this test it is clear that the satellite does not respond to every command Even within the link margin the uplink is only 70 3 successful Poor sensitivity may not be the only cause of problems seen in orbit It also offers a possible explanation of why an uplink couldn t be established with CP3 while using the SRI dish However this was not the exact problem observed during sensitivity testing because Figure 76 shows the COMM recovered usually in only a few minutes During sensitivity testing the satellite would not recover from the software error labeled Problem 4 in Table 31 even after waiting 20 30 minutes Only a hard reset of the satellite would clear the error The long duration testing occurred without the sensitivity measurement test setup shown in Figure 12 Perhaps it is necessary to attenuate the signal to the thr
82. in Appendix A Both COMMA and COMMB Figure 46 were measured for comparison The results of the sensitivity measurements of CDH Rev 4 and Rev 5 are shown in Sections 5 2 3 and 5 2 4 5 2 3 CDH Rev 4 Measured Sensitivity Testing The CDH Rev 4 used for sensitivity testing was from CP3 TestSat a model of the CP3 in orbit used for bench testing Due to hardware shortages in the lab only one CDH Rev 4 was available for testing CDH Rev 4 Sensitivity COMMA 88 5 dBm COMMB 92 4 dBm Table 19 The measured sensitivity of CDH Rev 4 for both COMMA and COMMB tested using the sensitivity measurement setup in Section 4 11 An interesting thing to note in Table 19 is that COMMB outperforms COMMA by almost 4 dB From on orbit data it has been noted that COMMB performs better than COMMA The RSSI pin of each COMM was monitored allowing verification of the signal strength received The resting RSSI value Table 20 is the RSSI output from the CC1000 with a 50Q termination at the antenna input of the CDH being tested Figure 46 Resting RSSI and Signal Strength COMMA 0 909 V 95 dBm 71 COMMB 0 997 V 100 dBm Expected 1 1V 105 dBm Table 20 Resting RSSI of CC1000 of CDH Rev With a 500 termination at the antenna input CC1000 resting RSSI values indicate residual noise on the CDH From the testing it seems that the COMM with the lower resting RSSI typical outp
83. ing room for very careful layout of the RF circuitry on the top side of the board Removing the redundant COMM would also simplify the software Application notes and recommended layouts of the CC1000 are available from the manufacturer 113 8 4 2 Receive Sensitivity versus Temperature The recommended crystal for the CC1000 varies tremendously over large temperature ranges With such a narrow bandwidth any deviation from the TX RX frequency could possible cause problems closing the uplink Although frequency shift due to Doppler is accounted for frequency shift caused by extreme temperature is not All the receiver sensitivity testing done was performed at room temperature A detailed analysis could provide valuable insight to whether this could be a problem Instead of using a crystal a Temperature Controlled Crystal Oscillators TCXOs would provide a more stable reference This is recommended for future bus development and could eliminate problems caused by shifting receive and transmit frequencies 8 4 3 Antenna Characterization and Re design The receive sensitivity determined with the test setup relies on a direct connection to the U FL RF Connector on the satellite This doesn t take into account the antenna gain or loss so a detailed analysis of the antenna system would improve link budget accuracy Since the CubeSat s small size prevents using the proper length dipole the antenna is slightly shorter than it should be It is un
84. ink is only approximately 70 even in favorable conditions Software testing should be done to determine there is any way increase reliability of the COMM Long duration testing should be performed using the test setup to reduce the signal strength close to the threshold This will closely replicate the orbital conditions allowing a better indication of COMM performance 7 2 2 Temperature Controlled Crystal Oscillator The CC1000 datasheet recommends using a very inexpensive 14 7456 MHz crystal to keep costs down The downside is that a cheap crystal can fluctuate significantly up to 50 parts 100 per million over a large temperature range Since most SRD Short Range Device applications are mass produced this is understandable because reducing cost is a high priority Additionally the assumption is made that both the transmitter and receiver exist in environments of similar temperatures Since PolySat is not mass producing CubeSats using a more expensive crystal to gain better stability over temperature is not an issue In CubeSat applications the temperature can vary greatly so a cheap crystal could be a problem if the transceiver s PLL Phase Lock Loop needs a very stable reference From on orbit data the side panels of CP3 experienced temperature swings from 40 to 30 4 Application Note 0019 available from Texas Instruments discusses problems with the receive sensitivity due to crystal frequency variation over a large temper
85. ion A signal of 24 dBm will permanently damage the LNA attached to the Spectrum Analyzer Section D The power splitter discussed in Section 4 5 is a resistive splitter so the port to port isolation is only 6 dB During a satellite response the input power to the LNA Section D of Figure 42 will be approximately 24 dBm This exceeds the maximum input power to the LNA and would cause permanent damage 20 To prevent damaging the LNA during sensitivity testing of the 60 PolySat COMM system the LNA amplifier is disconnected from the setup Figure 43 This setup is the standard setup used to measure sensitivity shown in Figure 12 and Figure 42 Spectrum Analyzer 43 2 dB HP 8566B LNA Computer 0 110 dB 437 MHz 1W Splitter Variable Attenuation Attenuation Yaesu FT 847 Radio Faraday Cage Oscilloscope Figure 43 Sensitivity setup for finding the threshold of sensitivity of the PolySat COMM system The LNA is disconnected from the power splitter because the satellite response exceeds the maximum input power of the LNA and would cause permanent damage Once the LNA is disconnected from the splitter Figure 43 the user can send commands while increasing the variable attenuator Section B of Figure 12 Eventually the signal will be reduced below the sensitivity of the PolySat COMM system and the threshold of sensitivity will be the point at which the COMM system can no long
86. kHz above the RX frequency for high side injection and 150 kHz below RX for low side injection eese nennen 8l Figure 57 CDH R4 Comma noise floor centered at the PolySat COMM receive frequency 437 365 MHz LO leakage visible at 437 515 MHz RBW 10 kHz Crystal 82 Figure 58 CDH R4 CommB noise floor centered at the PolySat COMM receive frequency 437 365 MHz LO leakage visible at 437 515 MHz RBW 10 KHZ eee 82 Figure 59 CC1000 Development Board centered at receive frequency 434 010 MHz LO Leakage visible at 433 860 MHz RBW 10 KHZ eee sees eene nennen 83 Figure 60 System Noise Floor of the LNA and Spectrum Analyzer RBW 10 kHz The input of the LNA is terminated with a 50 Q connection This provides a reference noise level at which to compare each noise floor graph to e ee eeeeceeeeceeceeeeeeeeeeceeseceesaeceeeeeeenteeeenaeeees 83 Figure 61 CDH Rev 5 CommA centered at receive frequency 436 365 MHz RBW 10 j 84 Figure 62 CDH Rev 5 CommB centered at receive frequency 437 365 MHz RBW 10 opp 84 Figure 63 CDH Rev 4 CommA 100 500 MHz RBW 1MHz Harmonics of crystal spaced 14 74 MHz apart are clearly visible Broadband noise is much higher than CC1000 development board Figure ane aide HO ete 85 xi Figure 64 CDH Rev 4 COMMB 100 500MHz RBW 1MHz Harmonics of crystal spaced 14
87. known how much this is affecting the link budget In order to increase the antenna length the deployment system would need to be redesigned and tested With the link budget so small a characterization of the antenna system could help identify whether or not the antenna is a weak link in the COMM system 8 4 4 Long Duration Testing Long duration testing can reveal problems previously unnoticed More testing could help identify software bugs and provide more information about the inconsistent uplink of the CP 114 Bus During development of the new bus long duration testing of the COMM system provide very useful information on how reliable the uplink is 8 4 5 Upgrading the Ground Station to Improve Uplink to the Satellite The 100W amplifier could be replaced with a more powerful amplifier The HLV 1100 a 1000W amplifier made specifically for the 70 CM amateur band would provide an additional 10 dB of uplink margin This is a very practical way of increasing the uplink margin but expensive The HLV 1100 is a built to order amplifier and costs approximately 6 000 Figure 83 HVL 1100 1000W amplifier AAUSAT II a Danish student satellite launched April 2008 also noticed a lower uplink margin than originally expected By adding the HLV 1100 they were able to increase the uplink margin and as of October 2008 uplink and downlink occurred on a daily basis 11 A very useful project would be to collaborate with Aalborg Uni
88. layed Average Noise Level DANL is approximately 95 dBm 3kHz or 129 77 dBm Hz The noise figure of the Spectrum Analyzer is the difference of the noise floor dBm Hz and the thermal noise power 174 dBm Hz NFspectrum Analyzer 129 77 dBm Hz 174 dBm Hz 44 23 dB Equation 4 Calculated Noise Figure of the Spectrum Analyzer 41 It is important to note that the above noise figure was determined with the attenuation set to 10 dB on the Spectrum Analyzer Without the 10 dB attenuation the noise figure would be 34 23 dB The noise figure of the Spectrum Analyzer can be significantly reduced by placing a large gain block on the front end shown in Section D of Figure 12 and in Figure 27 NF cascaded 5 58 dB NF 44 23 dB NF 3 127dB 7 7 spect Spectrum Analyzer 43 2 dB LNA i J I I 1 l i Faraday Cage Isolated Device Under Test Figure 27 The Noise Figure of the cascaded system LNA and the Spectrum Analyzer approaches the Noise Figure of the first stage of the cascaded system LNA This increases the sensitivity of the Spectrum Analyzer F2 Spectrum Analyzer m ae ee oo oe Figure 28 Noise Figure F of the two stage cascaded system LNA and Spectrum Analyzer The theoretical Noise Figure of the cascaded system Figure 28 is calculated using Table 13 and Equation 5 42 Linear Fi 3 127 dB 2 0545 F 4
89. ncy 111 Bee f fur f f 0 1 Lower FSK Center frequency Upper FSK sk frequency frequency f gt Figure 82 Frequency separation The CC1000 was not originally designed to receive and transmit 1200 baud AX 25 packets at amateur frequencies using AFSK 7 Because the CC1000 is a single chip transceiver only digital information can be passed to the transceiver on the DIO line Then the CC1000 outputs a modulated RF signal Since the CC 1000 is not designed for AFSK for a successful downlink a translation must be made from FSK to AFSK This is done by using LSB lower side band mode of the Yaesu FT 847 which only demodulates half the FSK signal The FSK tones from the CC1000 correspond closely to the 1200 2200 Hz tones of the AX 25 AFSK standard allowing packets to be decoded by MixW software As mentioned before a major limitation of the CC1000 is the data rate dependency on separation frequency In order to properly receive a packet the frequency separation must be at least twice the bit rate For a 1200 baud uplink a minimum separation of 3 kHz is required However the bandwidth of the Yaesu s SSB single side band filter is 400 Hz to 2 2 kHz at 6 dB harshly limiting the FSK separation frequency of the uplink As a compromise the uplink data rate was reduced to 600 bits s and the separation slightly increased to 2 kHz Luckily the Yaesu transceiver does not have a data rate limitation based on separati
90. near correlation shows that this does indeed happen For example when the attenuator is adjusted from 50 to 60 dB the measured signal is reduced from amplitude of 76 7 dBm to 86 8 dBm The percent error calculated in Equation 8 is only 1 Measured Expected _ 76 7 86 8 10 P tE x 100 x 100 19 ercent Error Expected x 10 x Equation 8 Equation used to calculate Percent Error 58 Throughout most of the characterization the linear relationship remains strong with a percent error of less than 10 But with the attenuator set to 94 dB or higher the linear relationship seems to deteriorate Percent error varies greatly in some cases reaching 90 Although this suggests that the setup isn t accurate for signals with a magnitude of less than 120 dBm this isn t the case The accuracy of the measurement signal is limited by the sensitivity of the Spectrum Analyzer The displayed average noise floor of the Spectrum Analyzer is approximately 94 dBm RBW set to 30 kHz default attenuation of 10 dB A RBW of 30 kHz was chosen because it is the system noise bandwidth of the transceiver utilized in the PolySat COMM system With measured signals falling below approximately 64 dBm corresponding to actual signal strength of 106 3 the Signal to Noise Ratio SNR becomes less than 10 dB At measured signals below 77 7 dBM corresponding to a magnitude of 120 dBm attenuator setting of 94 dB the SNR is very low
91. ns set the Spectrum Analyzer to a center frequency of 437 365MHz and a span of 5MHz Measuring the output power of the FT 847 Connect the output of the radio with 30 dB attenuator to the input of the spectrum analyzer Send a beacon command 0201 in MixW and observed the magnitude on the Spectrum Analyzer Use the marker peak search button to measure the power It should either be around 0 to 2 dBm or 19 to 20dBm Note which one it is as this dictates what attenuation settings to use The table below shows the attenuation setting and the approximate signal strength reaching the receiver under test FT 847 Output 0 dBm FT 847 Output 20 dBm Attenuator Power Attenuator Power Setting dBm Setting dBm 79 105 99 105 73 100 93 100 69 95 89 95 121 64 90 84 90 59 85 79 85 54 80 74 80 49 75 69 75 44 70 64 70 39 65 59 65 34 60 54 60 29 55 49 55 24 50 44 50 8 Connect Port 1 of the Power Splitter to an SMA Port of the Faraday Cage using semi rigid coax Make sure Port 2 is terminated with a 50Q termination gold cap with chain 9 Connect two BNC Cables to the Oscilloscope add SMA BNC adapters and hook them up to two SMA ports on the Faraday Cage From inside the cage use alligator clips to connect to pin 28 of the CC1000 Hook up COMMA to Chl and COMMB to Ch2 10 Using a small S
92. nt switch es should be compatible with Z contact points ADDITIONAL NOTES No external components other than the rails shall touch the inside of the P POD Shall incorporate Remove Before Flight pin OR launch with batteries fully discharged Rails shall be aluminum hard anodized CONTACT DETAIL FOR SIDE 7 At least one 1 deployment switch shall be incorporated on 11 CubeSats Center of grabity shall be located within a sphere of 2 cm from its geometric center Separation springs can be found at McMaster Carr P N 84985A76 Figure 1 CubeSat Standard 10x10x10 cm cube 1 33 kg maximum weight 2 1 1 Cal Poly CubeSat In addition to maintaining the CubeSat standard Cal Poly s CubeSat program designs and manufactures the Poly Picosatellite Orbital Deployer P POD which interfaces to the launch vehicle and deploys up to three CubeSats into orbit Launch provider coordination and satellite integration is also led by CubeSat Figure 2 PPOD on left 3 CubeSats on right 2 1 2 CubeSat Launch History Eurockot 2003 First CubeSat launch Two Cal Poly Mk I P PODs successfully deployed four CubeSats Dnepr 1 2006 Five Mk II P PODs with fourteen CubeSats were lost due to launch vehicle failure Minotaur I TacSat2 2006 One modified Mk II P POD deployed GeneSat 1 a NASA Ames CubeSat Dnepr 2 2007 Three Mk II P PODs successfully launch seven Cubesats including Cal Poly s CP3 and
93. ocedures aren t capable of identifying all possible problems Fully assembling and testing a CDH board requires significant resources tying up students with a rather mundane task With PolySat s limited student resources this is not the best use of their time Software and hardware development is also hindered Because the process of assembling a CDH is resource intensive it seems the lab is always experiencing a shortage of hardware A solution to improving reliability and conserving student resources is to have the boards assembled by a company specializing in PCB assembly The drawback is added expense but the increase in reliability and resources saved would offset this expense With professionally assembled boards in lab both hardware and software development would be less hindered by hardware shortages and reliability issues 7 2 6 High Frequency Probe To facilitate COMM system development the lab should be equipped with a high frequency probe Similar to a scope probe this would allow troubleshooting at RF frequencies Several of the difficulties encountered during this thesis could have been avoided with a high frequency scope probe The HP 85024 High Frequency Probe capable of measuring signals from 300kHz to 3GHz is the preferred solution It is directly compatible with the Spectrum Analyzer 108 in lab except it does require the separate power supply The probe features high sensitivity and a very low shunt capacitance whi
94. oly Bus standardized bus developed for CP2 revised for CP3 and CP6 DANL Displayed Average Noise Level dB Decibel dBi Decibel referenced to an isotropic antenna dBd Decibel referenced to a dipole antenna dBm Decibel referenced to mW of power DESENS Desensitization DUT Device Under Test EMC Electromagnetic Compatibility EPS Electrical Power System Woven fiber glass with resin principle component of circuit board FSK Frequency Shift Keying GHz 10 Hz rc Inter Integrated Circuit a low speed two wire bus developed by Philips IF Intermediate Frequency ISM Industrial Scientific and Medical LNA Low Noise Amplifier MHz 105 Hz NPS Naval Postgraduate School PIC Programmable Interface Controller PCB Printed Circuit Board P POD Poly Picosatellite Orbital Deployer LEO Low Earth Orbit NF Noise Figure LO Local Oscillator PLL Phase Lock Loop NRZI Non Return to Zero Inverse SMA SubMiniature Version A 500 connector DC 18 GHz SMB Sub Miniature version B 50Q connector DC 4 GHz RBF Remove Before Flight RF Radio Frequency RFID Radio Frequency Identification Device RSSI Received Signal Strength Indication Rx Tx Receive Transmit TVAC Thermal Vacuum Chamber SNR Signal to Noise Ratio SRD Short Range Device UHF Ultra High Frequency U FL Miniature RF connector 128 Very High Frequency Vector Network
95. ommended layout the crystal harmonics could be significantly reduced 6 2 Testing for Reduced COMM Sensitivity from Switching Noise After the unreliable uplink of CP3 and CP4 was noticed several theories suggested noise on the PolySat COMM system as a possible reason for reduced sensitivity 6 2 1 Switching Noise from DC DC Converters The satellite power system uses DC DC converters to regulate the voltage from the 4 2V batteries to 3V rails used for the CDH microprocessor and COMM system This is shown in Figure 74 Figure 74 The DC DC converter regulates the battery voltage from 4 2V to 3V for the PolySat COMM System 2 90 DC DC converters typically provide a greater efficiency than linear regulators To achieve greater efficiency DC DC converters switch on and off the output of the regulator at a specific duty cycle to achieve the required average output 2 Due to this on off switching noise can be a significant problem It has been thought that the harmonics caused by the DC DC regulators are introducing noise on the bus reducing COMM sensitivity 4 To test for a reduction in sensitivity caused by switching noise of the DC DC converters the sensitivity of each CDH was measured with and without the DC DC converters Since the Faraday Cage has two ports for passing through DC power each COMM CDH Rev 4 and Rev 5 was powered from a linear power supply for the CDH The DC DC converters were physically removed from
96. on frequency allowing 1200 bits s at a spacing of 2 kHz 112 8 3 2 Overall Sensitivity A separation frequency of 64 kHz is recommended for the CC1000 The PolySat COMM system utilizing AFSK rather than true FSK has a separation of 2 kHz Although the datasheet doesn t list receive sensitivity information for a setup of 0 6 kBaud at a separation of 2 kHz a comparable ratio is 4 8 kBaud at 20 kHz The listed sensitivity is 104 dBm for NRZ encoding With CDH Rev 5 responding to packets down to 100 dBm the CC1000 is for the most part performing as expected Noise could be reduced through a new layout but the sensitivity of the bus is ultimately limited by the CC1000 To further increase the link budget future bus designs will need to utilize transceivers of greater sensitivity The AX5042 transceiver chosen as the replacement transceiver for the new bus has a stated sensitivity of 122 dBm at 1200 baud Compared to the CC1000 this is a huge performance increase of approximately 20 dB Most importantly though is the fact that the AX5042 s sensitivity is not dependent on the FSK separation frequency 17 8 4 Future Work 8 4 1 New Layout for CP5 As discovered during the noise characterization poor layout has resulted in a large increase in broadband noise By removing the redundant COMM the board layout will be simplified greatly With more board space the CDH digital circuitry can be placed on the backside of the board provid
97. ows the use of shielded coaxial cables to connect from the power supply to the Faraday Cage Compared to standard leads shielded coax is less likely to act as an antenna and carry signals to the Faraday Cage 53 Figure 38 SMB Connectors used to pass DC power into the Faraday Cage 4 8 5 Attenuation Performance of Faraday Cage After completing the construction of the Faraday cage a test was performed to evaluate the enclosure s attenuation at 437 MHz The first test consisted of placing a source transmitting at 437 MHz inside the cage Using a spectrum analyzer with an antenna attached the power of the source was measured while the cage was open and closed By observing the difference of the two measured powers an approximate measure of attenuation was obtained This test is shown in Figure 39 Faraday Cage 6 Spectrum Analyzer with antenna RF Source transmitting at437 MHz Figure 39 Testing the Faraday Cage for isolation A handheld radio was placed inside the cage transmitting at 437 MHz The HP8566A Spectrum Analyzer equipped with an antenna was used to measure the signal 54 strength with the cage closed and open The difference between these two measurements provides approximate level of isolation It is important to note that this is an approximate attenuation because this does not take into account other factors such as multipath or fading due to reflections Isolation Devic
98. poly edu earthstation equipment index php 20 Drop In Monolithic Amplifier ERA 3 Data sheet Available from http www minicircuits com pdfs ERA 3 pdf 21 User Manual Yaesu FT 847 Transceiver Available from http www yaesu com 22 Anritsu Application Note Noise Figure Scorpion Option 4 April 2000 23 McCabe Keith Enhancements to the CPX I2C Bus Cal Poly Senior Project De cember 2007 119 Appendices Appendix A Sensitivity Measurement Procedure Ground station Installation Setup 1 Gather required equipment Faraday Cage Attenuation Block and the box of connectors All of this is located in the gray cabinet listed RF Testing NI gt F 2 Double check MacDoppler on Marconi to ensure there isn t a pass coming up for at least an hour 3 Turn on the Spectrum Analyzer It takes several minutes to warm up and is not ready to display signals until a distinct click is heard WARNING 30dBm 1W MAX AT THE INPUT OF THE SPECTRUM ANALYZER DO NOT EXCEED THIS 4 The computer at the RF bench should have Mix W shortcuts on the desktop and is already setup with the FT 847 transceiver Turn the transceiver on 120 GaGa 43184941 Verify there is a 30dB attenuator at the 433MHz port of the Yaesu FT 847 and attach it to the input of the Spectrum Analyzer Verify the RF Power is turned all the down CCW Using the Center Frequency and Frequency Span butto
99. quation and converting from meters to inches the skin of an signal at 437 MHz penetrating copper is calculated in Equation 7 poe 6 787 x 10770 in a ees UE CEP ee ee 124 microns 2 3 19 x 10 8 47 437 x 10 Hz Equation 7 Calculating the skin depth of copper of 437 MHz signal At 437 MHz the thickness of the Faraday Cage material is not an issue With such little penetration the cage can be constructed with conducting material as thin or thick as convenient In order to isolate the satellite from stray RF radiation a Faraday Cage can be built based on the 48 principles described above In principle any high conductivity box would work fine The thickness of the metal is not of great importance because at the frequency of interest 437 MHz the skin depth of copper is very small 4 8 3 Construction Aluminum Steel or any other high conductivity metal could be used but copper was chosen because the enclosure could be easily soldered rather than welded Although welding is also practical it would require specialized equipment and a person skilled in welding Since copper sheeting is quite expensive double sided blank FR4 PCB 1 OZ copper was used The board provided the structural rigidity and the unetched copper served as the solid conductor As previously shown the skin depth of a signal at 437 MHz is 0 124 mils whereas 1 OZ copper has an average thickness of 1 4 mils Due to the small siz
100. r Lastly interference can be intentionally introduced onto the bus while monitoring the RSSI pin This provides information about the susceptibility to DESENS receiver desensitization of the CC1000 presented in Chapter 6 68 5 4 3 CDH Revision 4 CDH Rev 4 flown on CP3 is the standard CP Bus without the LNA 2 3 The RF chain is shown in Figure 49 RF switches are used to select each COMM and for switching between Rx Tx for the antenna RF_TXA_10 gt SEL 10 RF RXA 104 SEL RXA 0 C114 u20 C112 U FL R SMT 2200F SW 425 220pF LEN ce i RF TXB 105 3 amp SEL TXB 10 COMMA 10 3 RF RXB 10 SEL_RXB_10 232 MBROS20L U32 wii7s7na Figure 49 RF Chain of CDH Rev 4 The antenna connector is AC coupled to the main RF switch which selects COMM A or B Individual RF switches select between RX and TX The RX line goes directly to the CC1000 and the TX is from the RF amplifier output 5 4 4 CDH Revision 5 CDH Rev 5 is the most current CDH Past revisions from 2 4 have included mostly minor layout changes such as fixing wire modifications and replacing obsolete components The CC1000 s noise figure is 12 dB which is pretty poor By adding a gain block to the front such as an LNA the sensitivity can significantly be improved theoretically With CDH Rev 5 an LNA was added to the receive line The required matching elements were added along with a high pass filt
101. r the problematic behavior Table 31 was difficult because of how infrequently it happened There was no specific pattern so reproducing the behavior was difficult to do In order to quantify how frequently the erratic behavior occurs a long duration test was performed TestSat an exact model of CP3 in orbit used for software development was left on for three days It was programmed with flight code to ensure behavior as close as possible to the satellites in orbit Beacon commands were sent to the satellite every 60 seconds and each response was recorded The approximate power reaching TestSat was 50 dBm eliminating poor sensitivity as a culprit Comm Test 60 second 0201 ping responses 0 06 30 1 neas 0 05 30 11 D Z 2 0 04 30 35 u o 2 gt 0 02 20 197 4 ee ix ee didi 0 500 1000 1500 Number of responses 2 3 Figure 76 Beacon commands sent every 60 seconds and the satellite responses monitored The horizontal axis shows how many satellite responses there were for each command from the group station The vertical axis is divided into bins showing that not every command received a response 95 Figure 76 shows a histogram of the results of the three day communication test The X Axis indicates how many ACKs acknowledgements the satellite made in response to the beacon command and the Y Axis shows the time period is divided into 1 5 minute sections The first b
102. revisions of the PolySat COMM system and the ground station receiver 4 3 RF Source The first component of the setup is the RF source Section A of Figure 12 For calibration of the test setup discussed in Section 4 10 the HP 8640B RF Source was used By providing a signal of variable magnitude at 437 MHz a Spectrum Analyzer can be used to quantify the signal amplitude as a function of attenuation applied For the actual sensitivity measurements Section 4 11 the Yaesu FT 847 Amateur Transceiver is used with a laptop to send actual commands to the satellite or receiver under test Commands can be sent from MixW 26 interfaced to the FT 874 radio Section A of Figure 12 and responses from the satellite are also monitored with MixW 4 4 Attenuation Stage In order to vary the amplitude of the signal reaching the DUT a variable attenuator is placed directly after the 30 dB attenuator The JEW Model 50DR 001 variable attenuator can be adjusted from 0 110 dB in 1 dB steps 10 437 MHz 1 W 30 dBm Fixed Variable Attenuation Attenuation Yaesu FT 847 Radio Figure 14 Attenuation Setup allowing signal strength to be adjusted from 0 to 110 dBm A table outlining the specifications of the variable attenuator is presented below Full specifications can be found in the datasheet online JFK 50DR 001 N Variable Attenuator Frequency Range DC 1000 MHz Attenuation Range 0 110 dB VSWR 1 2 1 max DC
103. riginate from the CDH processor 3 The bus originates from the CDH microprocessor Figure 75 so without it on there is no clock or data line Then with the CDH microprocessor programmed and powered the sensitivity was measured again No difference in receive sensitivity was found eliminating switching noise as a cause of reduced sensitivity While the LNA internal to the CC1000 is broadband it most likely would not amplify signals in the hundreds of kilohertz range Therefore any interference at the IF could possibly be caused by leakage into the IF from the RSSI pin A 150 kHz square wave 1 5Vp 1 5V offset was applied to the RSSI pin of the CC1000 with the goal of intentionally introducing interference at the IF The receive sensitivity was measured again Section 4 11 and no reduction in sensitivity was observed The I C is not reducing sensitivity of the COMM system 92 6 3 Additional Issues Relating to Inconsistent Uplink Several observations suggest that the satellite s non responsive behavior is not just a sensitivity problem Other problems could include software bugs and problems with the way the microcontroller interfaces to the transceiver 6 3 1 Observations from CP3 and CP4 When the unreliable uplink was first diagnosed as poor sensitivity SRI s 60 meter dish was used for uplink to CP3 and CP4 Compared to PolySat s ground station the large dish offers a tremendous increase in gain However CP
104. se seen at the antenna connection the receive line of the CC1000 Figure 49 and Figure 50 78 Spectrum Analyzer 43 2 dB LNA Faraday Cage PolySat COMM Figure 53 Test setup to monitor the noise at the receive line of the PolySat COMM system The LNA attached to the Spectrum Analyzer is used to observe the noise floor 79 Spectrum Analyzer 43 2 dB LNA Faraday Cage Figure 54 Test setup showing the system noise floor of the LNA and Spectrum Analyzer This is important in comparisons between the measured noise of each COMM system For comparison the system noise floor setup shown in Figure 54 is included It is important to note that each graph obtained from the noise floor is 42 3 dB higher than the actual signal due to the gain of the LNA However since all measurements include this gain it s not necessary to subtract off this gain A high gain stage in front of the Spectrum Analyzer Figure 54 is required to increase sensitivity of the Spectrum Analyzer discussed in Section 4 7 Before diving into a discussion of the noise performance LO leakage should be mentioned to reduce confusing normal LO leakage for noise on at the receive line of the CC1000 80 Figure 55 LO leakage For the CC1000 transceiver the leakage is 57 dBm In a superheterodyne receiver the RF signal is mixed down to the IF with the LO A non ideal mixer exhibits LO leakage which is when the LO signal leaks backw
105. se the satellite response exceeds the maximum input power of the LNA and would cause permanent apium 61 Figure 44 Sensitivity test of the PolySat COMM system The sensitivity threshold is 100 7 dBm corresponding to the last successfully decoded command indicated by a Satellite TESPONSE MH 62 Figure 45 Sensitivity measurement after the threshold of sensitivity is found The LNA is used to verify the signal strength of the Minimum Detectable Signal MDS which is the weakest signal PolySat COMM system is capable of responding to This is the measured receive sensitivity and is accurate to 2 dB sse 63 Figure 46 Main components of the redundant COMM system CC1000 Transceiver on left with RF amplifier on right Most of the additional components are supporting circuitry for the transceiver and amplifier matching power decoupling etc Each redundant COMM 1s labeled Ac and B aon ea oen 64 Figure 47 High level circuit diagram of the CC1000 esee 65 Figure 48 CC1000 RSSI pin voltage versus receive power This is a graph from the CC1000 datasheet providing a reference voltage as a function of received signal strength 68 Figure 49 RF Chain of CDH Rev 4 The antenna connector is AC coupled to the main RF switch which selects COMM or B Individual
106. separation of 64 kHz so the PolySat COMM system may see less of a reduction in sensitivity since a 2 kHz separation frequency is used This is because the lower frequency separation will have more tolerance to frequency errors However it does emphasize the fact that a reduction in sensitivity could be caused by a low quality crystal Instead of using a crystal as a reference a Temperature Controlled Crystal Oscillator TCXO could be used to maintain stability over temperature Using a TCXO could reduce crystal frequency variation to less than 5 PPM compared to the current crystal varying 50 PPM Even though a wider frequency separation is more susceptible to decreased sensitivity from crystal variation the application note recommends against reducing the frequency separation since maximum sensitivity is achieved with a 64 kHz separation This is further evidence that using a 2 kHz frequency separation is possibly causing reduced sensitivity of the CC1000 After the potential problem of reduced sensitivity caused by crystal variations over temperature was discovered a 102 heat gun was used to heat CC1000 the crystal to approximately 50 C The LO leakage see Figure 55 monitored with the Spectrum Analyzer was observed to shift by 30 kHz verifying that the cheap crystal is capable of causing a significant shift of the IF 7 2 3 Proper Testing Procedures to Verify the RF Chain Significant improvements can be made to t
107. sertion Loss 0 1 dB Amplitude Imbalance 0 02 dB Power 0 75 W Table 5 Characteristics of ZFRSC 42 S Power Splitter 28 Figure 16 ZFRSC 42 S Power Splitter Signals 6 dB down from the original signal at port S will appear at ports 1 amp 2 With signals of equal magnitude reaching the Spectrum Analyzer and the DUT the signal strength at the DUT is easily measured 4 6 Low Noise Amplifier Following the attenuation path Section B of Figure 12 a resistive splitter divides the signal into two paths Section C of Figure 12 The first path used to measure signal strength feeds directly to a Low Noise Amplifier LNA attached directly to the Spectrum Analyzer Section D of Figure 12 By using a gain block the Spectrum Analyzer s sensitivity is significantly increased 29 Cascaded Noise Figure 5 dB Spectrum Analyzer 43 2 dB LNA NF 44 dB Faraday Cage Isolated Device Under Test Figure 17 Low Noise Amplifier used to increase sensitivity of the Spectrum Analyzer The resistive splitter divides the RF path and the Spectrum Analyzer is used to measure the amplitude of the signal reaching the satellite in the Faraday Cage The other path used during sensitivity measurement goes directly to the Faraday Cage Section E of Figure 12 It is important to note that the signal reaching the Faraday Cage is the same amplitude of the signal reaching the Spectrum Analyzer after subtracting o
108. signal gives way after too much use allowing separation of the prongs Figure 80 Since the female connector will still mate to the surface mount male connector even with an open circuit failure it is impossible to visually detect the problem Figure 81 A worn out U FL connector on left behaves an open circuit By adjusting the spacing of the two prongs a proper connection was established shown on right 106 Using the network analyzer to view the reflection coefficient S11 it is shown that a faulty U FL connector behaves as an open circuit Figure 81 Slightly bending the two prongs back to the original spacing ensures a proper connection evident by a reflection coefficient at the center of the Smith Chart The U FL connector an RF connector with a 50 2 characteristic impedance is a one time connector It is not meant to be repeatedly connected and disconnected as it is rated for a maximum of 30 connections 18 After a failure of one of the connectors was noted all the connectors currently used in lab were checked and almost every connector had significant damage The open circuit failure is caused by disconnecting and reconnecting past the manufacturer s specified maximum and also by applying too much force while connecting or disconnecting A special tool is available for correctly connecting the connector A simple way to verify the U FL connector is behaving correctly is to perform a continuity check with a multi met
109. signal strength to the FT 847 was reduced until it could no longer decode packets providing a threshold of receive sensitivity The Yaesu radio successfully decoded packets down to 115 dBm indicating that the Yaesu radio outperforms the CDH Rev 5 by about 15 dB It outperforms CDH Rev 4 by approximately 25 dB In order to consistently close the uplink margin the CP Bus needs a more sensitive receiver 77 Additional Testing As an attempt to dig deeper into the sensitivity issue several additional tests were performed to explore possible explanations of reduced sensitivity After determining the receive sensitivity the Spectrum Analyzer was used to observe noise levels at the receive line of the CC1000 Each COMM of both CDH boards was compared with the CC1000 development board used as a baseline standard More specific tests were conducted to confirm or disprove theories speculating the cause of poor receive sensitivity The bus was tested with and without C and DC DC converters which are switching elements believed to be introducing noise on the bus These tests were performed on both Rev 4 and Rev 5 CDH boards 6 1 Receiver Noise Floor Comparison The noise floor of each system was measured using the test setup shown in Figure 53 This was done by placing each COMM system CC1000 Development board CDH R4 and CDH R5 boards in the Faraday Cage for isolation from outside noise and using the LNA to amplify any noi
110. significant problem was observed none of the RF power reaching the transceiver was going through the U FL connector The test setup s high degree of isolation see Figure 12 ensured that the signal was reaching the CC1000 only through the RF chain allowing a much more accurate assessment of COMM functionality Systematic testing narrowed the problem down to a specific manufacturing defect the LNA of COMMA was not receiving power causing an attenuation of over 40 dB Since past testing procedures qualified a partially functioning board as a flight candidate an updated qualifying procedure is needed including a direct measurement of the receive sensitivity using the setup in Figure 12 104 7 2 5 U FL Connector Problems The U FL female connector Figure 80 which mates with the on board antenna connector of the RF Chain see Figure 49 and Figure 50 has a tendency to fail as an open circuit With an open circuit failure any RF power will be completely reflected back to the source An open circuit failure of the U FL connector would essentially disconnect the antenna from the RF chain rendering the COMM system useless Figure 79 Male U FL connector on CDH Rev 5 105 Figure 80 U FL female connector with open circtuit defect center conductor prongs are spaced too far apart shown on left new U FL connector shown on right The connector on right shows the proper center pin spacing The soft plastic separating the ground and
111. stem 91 Table 31 A list of four common satellite behavior problems observed during sensitivity testing iS 93 Table 32 Long duration COMM test results 96 Table 33 Overview of the performance of COMM tested for sensitivity 98 Table 34 Sensitivity comparison of CC1000 and the AX5042 future replacement transceiver The datasheet of the AX5042 states that the receive sensitivity is not dependent on PSK frequency separation 17 at ete eren ime eaa edes cte een ra een 99 Table 35 Total variation of crystal over a large temperature range calculated from application note ANO19 available from Texas Instruments eene 101 Table 36 Resulting IF error caused by temperature variations eene 101 Table 37 A list of possible ways to mitigate open circuit failure of the female U FL CONDE CLOES aetas bd 107 Table 38 Measured receive sensitivities of CDH Rev 4 Rev 5 and the PolySat ground station s o VOL 110 xiv LIST OF EQUATIONS Equation 1 Calculating the free space path 1056 1 onere teri diede 17 Euq atton 2 Noise Fi gure fits ed std en iui iu edid thu a eet caudate eden 39 Equation 3 Calculated Noise Figure of the Low Noise Amplif
112. still produces a marginal uplink solution Evidence shows that the unreliable uplink is a combination of hardware limitations and problematic software By using the testing done in this project as a template for developmental testing of a new bus progress can be made while learning from the past The sensitivity study performed in this thesis comes at the beginning of development of the new COMM system providing a method to verify that the new system will have greater sensitivity 116 Regardless of what the future holds PolySat has made amazing progress in the development of CubeSats and will continue innovation by learning from the past 117 9 Works Cited 1 National Instruments dB Gain Compression Measurement P1dB 2009 November 30th Retrieved from http zone ni com devzone cda tut p id 2952 2 Day Chris The Design of an Efficient Elegant and Cubic Pico Satellite Electronics System Cal Poly Masters Thesis December 2004 3 Huera Derek Development of a Highly Integrated Communication System for use in Low Power Space Applications Cal Poly Masters Thesis April 2006 4 Klofas Bryan Improving Receive Sensitivity of the CPX Bus Cal Poly Senior Project June 2008 5 Arakaki Dean Experiment 4 Noise Figure and 1dB Compression Point of an Amplifer Using the Vector Network Analyzer EE480 Lab Manual Cal Poly Electrical Engineering Department 6 PolySat Grounds
113. surement Setup Design and Testing Performed 21 CHAPTER 4 MEASURING RECEIVE SENSITIVITY 22 d EONERVIEW s Mcr e E eI I EE 22 4 2 LEST SETUP COMPONENTIS ai Ara ur teer tertium bn ee EE E erede 23 TS RESOURCE Sete AN A bh ene ttn Eu tee 26 4 4 ATTENUATION S TAGE eene couse eire UR e E E Y LUPO ree RN EE RR UU HE EOD 27 4 5 RESISTIVE POWER SPLITTER apei a r eerte ne ee mb e Ee tbe eR e ns 28 4 6 T OWANOISE AMPLIFIER dep cept use EU etu eaten beet t eet RP eis 29 4 6 1 Design of Low Noise Amplifier 30 4 6 2 Input Return Loss of LNA nn eheur 33 4 6 3 Output Return Loss Of LNA nein eet rennen nene eene enne en 35 vi 4 64 Forward Gain of LNA oe tdsep ar e t e n cd fe ORE E PI tede tpe 33 4 6 3 Reverse Gain of LNA y eee he RR eR PE REP UR PR PEOR tete aree Perret 37 46 5 EAB Compression ie ee SUPE 37 4 6 6 Noise Figure of LNA sin ioo e EO UIS PERO pe 39 HP85606A SPECTRUM ANALYZER DELE TREE en rr oe IRE IHE Rr etr Revier 41 4 8 FARADAY CAGE i otio pre PHP Oe Pete art node 43 ZG D Screen ROOms aee te t e Oe rb RO E Roe te Mte e P EE a RU UE OR eee 44 4 8 2 Custom Faraday Cage REP PERSE EP IPERPEU ERU 48 48 2 CONSIFUCTION Sc eee te C SERES OR ER ERE
114. t performing board CC1000 RSSI RSSI f COMMB RSSI H Datasheet RSSI 105 85 65 Figure 51 CP5 Flight Candidate RSSI curve 5 2 7 Yaesu FT 847 Transceiver PolySat has two independent ground stations named Marconi and Hertz The rotor mounted Yagi antennas can be seen atop the roof of the ATL Marconi consists of a Yaesu FT 847 amateur transceiver connected to dual phased M squared 436CP42Yagi antennas A preamp is connected directly to the antenna with LMR 400 coax which is fed through the roof with a 50 ft section of 5 8 heliax to the Yaesu FT 847 radio 6 With the CC1000 characterized its performance can be judged by comparing it to the ground station receiver By reversing the 75 standard receive sensitivity test the Yaesu FT 847 can be characterized The reversed sensitivity test is shown in Figure 52 The earlier discussion of antenna reciprocity in Section 3 2 1 suggested that the inconsistent uplink issues could be caused by significant sensitivity differences in the two receivers Since downlinks can be reliably decoded a significant difference in sensitivity between both receivers would suggest a more sensitive receiver is required for the PolySat COMM system From the user s manual the FT 847 has a stated sensitivity of 0 125uV at the 430MHz band with a 10 dB SNR 21 This is shown in Equation 9 0 125 x 1076 v 18y7 _ 500 312 5x10 W 125 dBm Equation 9
115. tation Equipment 2006 May 26th Retrieved from http polysat calpoly edu earthstation equipment index php 7 Noe Chris Design and Implementation of the Communications Subsystem for the Cal Poly CP2 CubeSat Project Cal Poly Senior Project June 2004 8 CC1000 Data Sheet Available at http www ti com 9 ZFRSC 42 Power Splitter Data Sheet Available at http www minicircuits com pdfs ZFRSC 42 pdf 10 JEW MODEL 50DR 001 Data Sheet Available at http www jfwindustries com 11 Jens Dalsgaard Nielsen and Jesper A Larsen a Danish Student Satellite October 2008 12 Abel John Evaluation of Particle Dampers in a Microgravity Environment June 2009 13 Lim Tiffany Antenna Characteristics of the CubeSat Dipole Antenna Cal Poly Senior Project July 2007 14 Application Note ANO19 Crystal oscillator issues for CC1000 Available at http focus ti com lit an swra072 swra072 pdf 15 Kobeissi Imad Noise Reduction Techniques for Microcontroller Based Systems Application Note 1705 118 16 User Manual CC1000DK Development Kit Application Note Available from http www ti com 17 AXSEM AX5042 Data sheet Data sheet Available from http www axsem com 18 Ultra Small Surface Mount Coaxial Connectors U FL Series Data sheet Available from http www hirose com 19 PolySat Projects 2009 Retrieved from http polysat cal
116. tenta ded etu MERE 92 6 3 ADDITIONAL ISSUES RELATING TO INCONSISTENT UPLINK 93 6 3 Observations from CP3 and pU Ue d qr EDEN 93 6 3 2 Observations from Sensitivity Testing eese eene teen een een eterne entente nennen teens 95 6 3 3 Long Duration Communication 95 CHAPTER 7 RESULTS OF SENSITIVITY TESTINQG eere ceste eene testes enses enata tuse tatnen sens enses tuae 98 7 1 COMPARISON OF COMM SENSITIVITY esee eee nennen enne neas K Eee OSa nre 98 Zl I Overall ede be d eed eee tere er dee e Re ete oe 98 7 1 2 CCL000 Performance cess hess decid ates A ive te steeds re eee ee eter eee eet antes 98 7 1 3 PolySat COMM Performance Compared to Future Replacement Transceiver esee 99 7 2 RECOMMENDATIONS FOR IMPROVING UPLINK AND FUTURE POLYSAT COMM 99 7 2 I Improved Layout Of Rey 5 sese ettet eee Dedi eo te eger Rie eei eee tak der 100 7 2 2 COMM Software Testing io eode ed darte aee rete eid edu Heese Meets 100 7 2 2 Temperature Controlled Crystal Oscillator eese eese ener eene 100 7 2 3 Proper Testing Procedures to Verify the RF Chain esee nennen 103 7 2 9 UEL Connector Problems e iei t ra tues Rl RR REP te Ehre tie Eee Re Pe nta eT 105 7 2 4 Contract Manufacturing of Assembled boards
117. the EPS ensuring that the only 3V rail was from the linear power supply without the associated noise from switching converters The test was repeated again this time using the DC DC converters to power the COMM and CDH No significant difference in sensitivity was noted Table 30 eliminating switching noise of the DC DC converters as a cause poor sensitivity In retrospect it is unlikely that the DC DC converters would reduce sensitivity since switching occurs at 750 kHz The harmonics at 437 MHz are likely to be very small and the IF of the CC1000 is at 150 kHz DC DC Converters Linear Regulator CDH Rev 4 COMMA 89 dBm 88 dBm COMMB 94 dBm 93 dBm CDH Rev 5 COMMA 92 dBm 94 dBm COMMB 102 dBm 101 dBm Table 30 Testing for differences in sensitivity caused by DC DC converters No significant difference in receive sensitivity is observed without the DC DC converters eliminating them as suspect in causing desensitization of the COMM system 91 6 2 2 Switching Noise from Bus The data bus consists of two lines SCL and SDA The SCL is the clock which switches at 100kHz suspiciously close to the CC1000 s 150 kHz IF In order to test for interference both COMMA and COMMB were tested individually without the CDH microprocessor powered 116 25 LAE Y Figure 75 PC switching SCL CDH 2 and data line SDA CDH 2 o
118. to isotropic so an antenna gain of 2 to 10 dBi is a reasonable approximation Parameter Magnitude Comments Pry Amplifier Output 50 100W Gant Isotropic Antenna gain 18 95 dBi 16 8 2 15 dB 18 95 dBi Gpus tx Gain from dual phasing 3 Dual phased antennas provide additional gain 5 8 hardline loss 0 22 dB 50 ft length Lm Miscellaneous loss 1 dB Estimated additional loss Connectors radio etc Les Free space path loss 142 2 to 153 4 dB Actual distance to satellite varies during pass Grx Antenna gain of satellite 2 to 10 dBi Depends on orientation Total 70 to 93 dBm Approximate strength after antenna rounded Table 1 Uplink Budget A sensitivity of 103 dBm would provide a worst case margin of 10 dB To ensure a reliable uplink the margin should be 20 dB In the next revision bus the AX5042 transceiver chosen as the replacement for the CC1000 will provide much better sensitivity allowing a greater margin Increasing the ground station transmit power will also insure a greater degree of uplink margin A receiver sensitivity of 103 to 113 dBm would provide a 10 to 20 dB worst case link margin 3 2 4 Link Budget Downlink The downlink budget is an important reference for when the sensitivity of the ground station receiver is characterized Parameter Magnitude Comments Amplifier Output 30
119. ut Return Loss of LNA IS 0 0 5 1 1 5 2 2 5 3 5 10 15 20 25 frequency GHz Figure 22 S of LNA 0 f GHz dB 0 4286 20 746 0 436075 20 73 0 44355 20 608 0 451025 20 341 Table 8 5 of LNA The output of the LNA has a return loss of greater than 20 dB across the 430 450 MHz amateur band It slowly approaches 10 dB around 2 5 GHz 4 6 4 Forward Gain of LNA 35 IS 50 45 4 35 30 dB 25 20 15 10 0 0 5 1 1 5 2 2 5 3 frequency GHz Figure 23 Forward Gain S of LNA 55 f GHz dB 0 4286 42 292 0 436075 42 296 0 44355 42 207 0 451025 42 195 Table 9 across the 70 CM Amateur Radio band To improve the sensitivity of the Spectrum Analyzer a large gain is required From the datasheet each MMIC block has a typical gain of 22 1 dB at 0 1 GHz and 21 dB at 1 GHz With two gain blocks in series the expected gain is approximately 42 dB below 1 GHz As the frequency increases to 2 and 3 GHz the typical gain of each block decreases to 18 7 dB and 16 4 dB respectively The expected total gain is approximately 36 dB at 2 GHz and 32 dB at 3 GHz Using the Vector Network Analyzer the forward gain of the LNA was measured to be greater than 42 dB across the 430 450 MHz amateur band Even as the gain drops off significantly to approximately 30 dB
120. ut of the PolySat COMM system using the recommended layout of the CC1000 could reduce broadband noise 16 possibly improving sensitivity 6 1 3 Receive Line of CC1000 Noise Floor over 70 CM Amateur Band After observing the broadband noise noise across the 70 CM amateur band was measured 435 000 440 000 445 000 430 000 435 000 440 000 445 000 450 000 Figure 70 CDH Rev 4 COMMB RBW 100kHz over the 70 CM amateur band 88 Only a slight difference approximately 3dB is noticeable between COMMA Figure 69 and COMMB Figure 70 but crystal harmonics are much greater 10 15dB than the development board Figure 71 439 000 Figure 72 System Noise Floor over 70 CM amateur band 6 1 4 Receive Line of CC1000 Noise Floor Characteristics 334 400 338 568 342 600 346 708 358 860 Figure 73 The CDH Rev 4 had two types of crystal harmonics visible The larger harmonic is spaced 14 74 MHz apart the frequency of the CC1000 crystal and the smaller harmonics were spaced every 1 474 MHz 89 Figure 73 shows crystal harmonics observed on the receive line of CDH Rev 4 The CC1000 s 14 74 MHz crystal has two types of harmonics visible The larger harmonic reoccurs every 14 74 MHz while the smaller harmonics repeat every 1 474 MHz Both harmonics are present from 100 500 MHz The CC1000 development board which uses the same crystal shows a much more controlled and cleaner noise spectrum By following the rec
121. versity see if PolySat could use their ground station remotely This would allow PolySat to test drive the amplifier before committing to the large expense However with PolySat projects becoming more and more ambitious the cost of the new amplifier is justified by an instant link margin gain of 10 dB 115 8 5 Summary 8 5 1 Conclusion Throughout the duration of this thesis an accurate method of testing receive sensitivity was developed and documented This was a significant development to the testing capabilities of the PolySat lab COMM systems can be tested for receive sensitivity noise analysis performed and verification of the RF chain can be confirmed as a result of this test system development Testing can occur alongside development for system verification identifying sensitivity problems early in the development stage CubeSat developers can easily replicate the test setup for sensitivity testing purposes helping advance the entire CubeSat community Using the test setup two revisions of Cal Poly s bus were thoroughly tested Based on the results of this testing it was determined that the CP Bus would greatly benefit from a more sensitive transceiver 8 5 2 Final Words At the beginning of this project a solution to the poor uplink problems of CP3 4 was sought Several unanticipated problems were uncovered and huge developments were made in sensitivity testing The test results show that our current receiver design
122. wn on CP3 and CP4 CDH Rev 5 includes a Low Noise Amplifier LNA and was flown on CP6 For a baseline comparison the Yaesu FT 847 ground station transceiver is measured for sensitivity Chapter 6 discusses additional testing performed in order to further investigate possible causes of poor sensitivity Chapter 7 provides observations and test results of each COMM system tested Comparisons of each receive sensitivity are made and possible methods of increasing receive sensitivity are discussed Chapter 8 concludes the thesis with conclusions and recommendations for future work in the area of receive sensitivity Chapter 2 Background Overview of Satellite Program 2 1 CubeSat The CubeSat standard was developed as a joint project between Stanford University s Space Systems Development Laboratory and Cal Poly s Multi disciplinary Space Technologies Laboratory Stanford professor Bob Twiggs and Cal Poly professor Dr Jordi Puig Suari led the way and now Cal Poly s CubeSat program maintains the specifications for the 10x10x10 cm satellites 3 CubeSat developers such as other universities or corporations must adhere to the 1 33 kg mass limit and dimensions of a 10cm cube Other specifications such as the Remove Before Flight RBF pin diagnostic port location and deployment switches are contained in the specification document NOTE Deployment switch and separation spring placement schemes shown in Option A and B Deployme
123. y of 104 dBm Under this assumption the CC1000 on the CP Bus should have a similar receive sensitivity 67 5 1 4 RSSI Output The CC1000 has a dedicated pin for Received Signal Strength Indication RSSI The voltage at this pin is inversely proportional to the received signal Therefore a voltage of 1 1V corresponds to a very weak signal 105 dBm and 0 1V indicates a strong signal 50 dBm 13 12 1 1 e 433Mhz H 3 868 08 S 07 06 05 04 0 3 02 0 1 0 105 100 95 90 85 80 75 70 65 60 55 50 dBm Figure 48 CC1000 RSSI pin voltage versus receive power This is a graph from the CC1000 datasheet providing a reference voltage as a function of received signal strength The RSSI is measured at the IF stage and only depends on the raw power not a modulated signal with data This was confirmed by sending both modulated and non modulated signals and observing no difference in the RSSI The RSSI pin of the CC1000 proved to be very useful in evaluating receiver sensitivity First the RSSI was monitored and compared to the magnitude of the signal reaching the isolated DUT This is useful in discovering problems in the RF chain Also data collected from the RSSI pin of each COMM system was compared providing an interesting look at what the CC1000 sees in terms of received powe
124. y of the PolySat COMM system 63 5 Sensitivity Testing 5 1 PolySat Satellite COMM System To evaluate the COMM performance of the satellites in orbit the receive sensitivity of the PolySat COMM systems of CP3 CP4 and CP6 were measured 5 1 1 Overview of COMM Setup Starting with CP2 a standardized bus was developed Chris Day and Derek Huerta s theses 2 3 discuss the entire system in great detail For the scope of this project only a brief description of the PolySat COMM system is necessary The satellite features redundant COMMS to protect against hardware failure Each COMM consists of several main components CC1000 transceiver RF amplifier RX TX switches and an antenna jack Figure 46 Main components of the redundant COMM system CC1000 Transceiver on left with RF amplifier on right Most of the additional components are supporting circuitry for the transceiver and amplifier matching power decoupling etc Each redundant COMM is labeled A and B 64 Two different revisions of the CP Bus COMM system were tested CDH Rev 4 is the COMM system flown on CP3 launched in 2007 CDH Rev 5 flown on CP6 in 2009 contains a Low Noise Amplifier on the receive line of each COMM Bryan Klofas documented the addition of the LNA in his senior project 4 For the remainder of this paper Rev 4 refers to the CDH without the preamplifier and Rev 5 refers to the CDH with the preamplifier 5 1 2 CC1000 UHF Trans

Download Pdf Manuals

Related Search

Related Contents

LC32DH510E Service Manual, 00 Version    guide logiciel twido  Bookeen Cybook OPUS  CADEIRA DE RODAS - World Health Organization  Orphee Mythic 22-CT Hematology Analyzer  Maxx Ice MIM250 Use and Care Manual  1. - Vivitek  HadCalc manual  

Copyright © All rights reserved. Failed to retrieve file