Detailed Design Review

Detailed Design Review P09343 Microwave Data II Joel Barry Amanda Kristoff Mia Mujezinovic Michael Pecoraro 1

P09343 Microwave Data II Technical Review Agenda Meeting Purpose: This meeting is to review the system and components to confirm it meets the Functionality, Engineering Specifications, and Customer Needs of the P09343 Microwave Data II KGOCE MSD team. Receive feedback to be incorporated for component/system improvements. Materials to be Reviewed: Project Overview (Rev. A) Customer Needs and Specifications (Rev. E) Concept Generation/System Level Design (Rev. B) Risk Assessment (Rev. B) Engineering Design Process (Rev. B) Component and System Progress Report (Rev. B) Two Quarter Schedule/Milestones (Rev. C) Meeting Date: February 13, 2009 Meeting Location: 70-3445 Meeting time: 12:00 2 PM Timeline: Start Time Topic of Review 12:00 Team Introductions 12:05 Project Overview 12:10 Concept Generation/System Level Design 12:15 Engineering Design Process - Components 12:25 Engineering Design Process - System 12:35 Component Results: 45 Schiffman Phase Shifter 12:45 Component Results: 90 Hybrid Coupler 12:55 Component Results: 180 Knochel Hybrid Coupler 1:05 Component Results: SMA Connector Launch 1:15 System Results: 4x4 Butler Matrix System A 1:25 System Results: 4x4 Butler Matrix System B 1:35 Layout Progress 1:40 Questions 2

Introduction Members: Mia Mujezinovic (Team Lead) Michael Pecoraro Amanda Kristoff Joel Barry Breakdown of Roles/Tasks: 90 Branchline Hybrid Coupler: Michael Pecoraro 180 Knöchel Model Hybrid Coupler: Joel Barry 45 Schiffman Phase Shifter: Mia Mujezinovic 4x4 Butler Matrix System A: Amanda Kristoff 4x4 Butler Matrix System B: Amanda Kristoff Vertical Launch: Anaren, Michael Pecoraro EDGE Updates: Amanda Kristoff Ansoft Designer Ansoft HFSS IEEE Xplorer Database AutoCAD Michael Enders (Anaren, customer) Resources Utilized: 3

Project Overview Customer Michael Enders RF Engineer Anaren Microwave Incorporated Space and Defense Group MEnders@anaren.com 315-362-0273 Project To design, build and verify the operation of two 4x4 Butler Matrices for use in antenna beamforming Figure 1: Block Diagram of System A Figure 2: Block Diagram of System B Purpose Butler Matrices are networks that allow antenna beams to be electrically (as opposed to physically) steered. The necessary condition for electrically steering an antenna beam is that the antenna elements must be given different phase progressions. This is what a Butler Matrix does depending on which input is used, a different phase progression is realized at the output. This project is more of a research type project for Anaren. They have two main goals that they would like us to attain: o Both Butler Matrices should be wideband covering a bandwidth of 10-12GHz o Anaren has never used a Knöchel hybrid - they would like to see how it performs individually as well as in a system 4

Customer Needs/Specifications Revision #: 5 Project #: P09343 Customer Need # Importance Description Comments/Status CN1 1 Layout stackup should be designed as specified by Anaren CS1 CN2 2 Suppress higher order TE and TM modes @ operating frequency built into design spec CS1, CS2 CN3 1 180 Degree Coupler Project 1 Title CN3a 1 Ease of integration with current RF products CS3, CS4, CS15 CN3c 2 Power Division Ratio of: mag of 1:1, phases separate properly CS5, CS6, CS7, CS8 CN3d 3 Narrow profile (small size) for future use in products CS16, CS17 CN4 1 90 Ddegree Hybrid Coupler Project 2 Title CN4a 1 Ease of integration with current RF products CS3, CS4, CS15 CN4b 2 Power Division Ratio of: mag of 1:1, phases separate properly CS9, CS10, CS11, CS12 CN4c 3 Narrow profile (small size) for future use in products CS16, CS17 CN5 1 45 Degree Phase Shifter Project 3 Title CN5a 1 Ease of integration with current RF products CS3, CS4, CS15 CN4b 2 Minimal Power Loss with Output CS13, CS14 CN4c 3 Narrow profile (small size) for future use in products CS16, CS17 CN6 SMA Launch for Re-Usable Test Fixture Project 4 Title CN6a 1 Ease of Use CS3, CS4, CS15, CS18 CN6b 1 Reduce testing costs from current testing methods CS16, CS18 CH6c 2 Achieves Wide Band operation CS15 CN7 1 Integration Butler Matrix A - 90 Degree Couplers Project 5 Title CN7a 3 Narrow profile (small size) for future use in products CS16, CS17 CN7b 1 Ease of Manufacturing CS1 CN7c 2 Meets all customer Specifications CS - Same as defined for components CN8 1 Integration Butler Matrix B - 180 Degree Couplers Project 6 Title CN8a 3 Narrow profile (small size) for future use in products CS16, CS17 CN8b 1 Ease of Manufacturing CS1, CS2, CS16, CS17 CN8c 2 Meets all customer Specifications CS - Same as defined for components CN9 1 All components must be manufacturable in given board size CS17 Cust. Need #: enables cross-referencing (traceability) with specifications Importance: Sample scale (1=must have, 2=nice to have, 3=preference only). Comment/Status: allows tracking of questions, proposed changes, etc; indicate if you are meeting the need ("met") or not ("not met") 5

Revision #: 1 Custom Impo er Spec. rtanc # e 6 Description Customer Needs/Specifications Spec. Value Units Achieved Physical Value CS1 1 0.060 Layout Stackup should be constructed of RO3003 +/- 0 Inches 0.060 Inches CS2a 1 Add Through-Stack vias to 'break' continuity of stackup to suppress TE and TM modes 46.5 mil Dia. 46.5 mil Dia. CS2b 1 Vias must be close enough to the traces to suppress modes but must not change impedances >77.8 mil 86.1 mil CS2c 1 Vias must be spaced close to each other to form continuous barrier against higher modes ~70 mil 66-80 mil CS3 1 Character Impedance 50 +/- 0 Ohms 50 +/- 0 Ohms CS4 1 Frequency Bandwidth 10-12 GHz 10-12 GHz X-Band CS5 2 180 Degree Coupler output Magnitude Difference 0 +/- 0.2 db 0 +/- 0.3 db CS6 2 180 Degree Coupler Return Loss (All Ports - MAX) -20 db -17.5 db CS7 2 180 Degree Coupler Output Phase Difference (Port 1 - Input) 0 +/- 1 Degrees 0 +/- 0.5 Degrees 180 +/- CS8 2 180 Degree Coupler Output Phase Difference (Port 4 - Input) 1 Degrees 180 +/- 0.5 Degrees CS9 2 90 Degree Coupler output Magnitude Difference 0 +/- 0.2 db 0 +/- 0.13 db CS10 2 90 Degree Coupler Return Loss (All Ports - MAX) -20 db -24.8 db CS11 2 90 Degree Coupler Output Phase Difference (Port 1 - Input) 90 +/- 1 Degrees 90 +/- 0.5 Degrees CS12 2 90 Degree Coupler Output Phase Difference (Port 4 - Input) 90 +/- 1 Degrees 90 +/- 0.5 Degrees 45 Degree Phase Shifter Phase Difference between Ports 1 44.57 +/- CS13 2 and 2 45 +/- 0 Degrees 1.5 Degrees CS14 2 45 Degree Coupler Return Loss (All Ports - MAX) -20 db -15.67 db Units Comments/Status Distance between edge of via and edge of trace Distance between vias - varies with bends SMA connector - defined as Return Loss under - 15dB CS15 2 DC to Connectors to provided accurate analysis from DC to 18GHz 18 GHz DC to 17 GHz Minimal spacing between couplers (minimize coupling 0.5 +/- Wavelen CS16 1 between couplers) 0.0 gths 2 Wavelengths CS17 1 Manufacturing Board Length and Width 12x24 inches inches design in process CS18 1 Minimum space between connectors 0.5 Inches 0.75 Inches Cust. Need #: enables cross-referencing (traceability) with specifications Importance: Sample scale (1=must have, 2=nice to have, 3=preference only). Comment/Status: allows tracking of questions, proposed changes, etc; indicate if you are meeting the need ("met") or not ("not met")

Customer Needs/Specifications CN1 CS1 CS2 CS3 CS4 CS5 CS6 CS7 CS8 CS9 CS10 CS11 CS12 CS13 CS14 CS15 CS16 CS17 CS18 X CN2 X X CN3 X X X X X X X X X CN3a X X X CN3c X X X X CN3d X X CN4 X X X X X X X X X CN4a X X X CN4b X X X X CN4c X X CN5 X X X X X X X CN5a X X X CN4b X X CN4c X X CN6 X X X X X CN6a X X X X CN6b X X CN6c X CN7 X X CN7a X X CN7b X CN7c X X X X X X X X X X X X X X X X CN8 CN8a X X CN8b X X X X CN8c X X X X X X X X X X X X X X X X CN9 X 7

Concept Generation/System Level Design Layout Stackup Chosen Stackup Copper Ground Plane Alternative Stackup 1.5 mil 60 mil 60 mil Rogers 3003 Arlon 6700 Copper Traces Rogers 3003 Copper Ground Plane 30 mil 30 mil Figure 3: Anaren specified stackup designs for implementation into both Butler Matrices. The only difference between the two stackups is the thickness of the Rogers 3003 dielectric. 1.5 mil The 120 mil stackup will be implemented in the Butler Matrix Designs for some key advantages: o Thicker dielectric yields wider traces for equal impedance lines, and this ensures traces will be greater than the minimum manufacturing tolerance of 10 mil provided by Anaren o Wider traces are less susceptible to variation with the width tolerance of 0.5 mil Topology Selection By unfolding the crossovers of the Butler Matrix A design, a single copper trace layer that mirrors the symmetry of the 90 coupler. Butler Matrix B is similarly unfolded to a single copper trace Vertical Mounting SMA connectors allow for access to all ports without implementation of crossover networks. Figure 4: Selected topologies for the Butler Matrix A Design (Top) and the Butler Matrix B Design (Bottom). 8

Risk Assessment Risks That Have Been Addressed: Understanding and learning the theory and models behind the components Understanding customer expectations/needs Learning Ansoft HFSS Understanding ideal simulation tool in Ansoft Designer Methodology of parameter selections Vertical mount documentation and theory, in a timely manner to incorporate into system design Completion of system design to allow manufacturing lead time of 6 weeks Current Risks/Assumptions: Getting board by Friday April 10 to allow adequate time for test/verification Launch may need adjustments depending on manufacturability Layout methodology will be successful 9

Component Design Process A simplified process (or flow) by which the team has been utilizing to design this project is shown below: Theory Ideal Simulations Designer Simulations HFSS Simulations Figure 5: Individual Component Design Flow Chart This generalized process applies to each component that is designed. Below, the design process illustrated above is explained in detail. 1. Theory - Textbook entries and papers written on each of the components being built, as well as documents regarding design of Butler matrices in general, were used to begin each design. Reading and fully understanding the theories behind each component, and how they will work together, was essential in beginning the designs. 2. Ideal Simulations - This type of simulation is done in Ansoft Designer using the special ideal components. These components do not take into account any losses and show the absolute best case performance of your component or system. Impedance values, electrical lengths and frequencies are determined by the research papers (formulas would be given) and specific specifications for the project. 3. Designer Simulations - Once the ideal simulations were completed, the substrate effects, discontinuities and other losses into the system were included using Ansoft Designer. As a result, the simulation provides a more realistic view of the performance of the components. The ideal simulation is the design goal, used as comparison to see how well the designed system performs. Designer allows for many variables to be tuned in real time. Intelligent tuning takes into account ideal values and theoretical dependencies, and changes are made logically. 4. HFSS Simulations - Once the designer simulation were finished, models were created in HFSS. Using the theoretical dependencies and the ability of Designer to see, magnitude-wise, how much these changes affect the output variables, the final adjustments are made to the model. Based on project development experiences, this proved to be the most efficient method of simulating and a good way of solidifying knowledge of the components and how changing variables affects performance. Having a process in designing the components is important in tracking progress as well as developing ideas to improve the design in order to meet specifications. In this process, there is a significant interdependency between Designer and HFSS, and the changes applied to the designs will depend on the compared results between these two simulations 10

System Design Process The process in Figure 6 is utilized to design the 4x4 Butler Matrix Systems A and B. Component Design Once component design was completed, the N-port data was exported from HFSS and imported into Ansoft Designer. In addition, the component layout from Ansoft Designer was also utilized. Ansoft Designer Simulations The component layouts from Ansoft Designer were assembled to identify additional transmission lines and bends needed to assemble the systems. This data was returned to the component level and developed in HFSS. The N-port data from all components and additional transmission lines and bends was assembled in Ansoft Designer to examine the final model. If components were identified as needing physical adjustments for the layout or improvements for system performance, these were returned to the component design state. AutoCAD Layout Once the layout was completed in Ansoft Designer and all additional transmission lines and bends were developed, the final layout was assembled in AutoCAD to be submitted for manufacturing. Figure 6: System Design Flow Chart 11

Component Schiffman Phase Shifter Ideal Physical HFSS The ideal model was developed using Schiffman s original paper published in the IEEE Microwave Techniques Journal. B.M. Schiffman, A New Class of Broadband Microwave 90-Degree Phase Shifters, IRE Trans. Microwave Theory Tech., vol. MTT-6, no. 4, pp.232-237. To obtain the even and odd mode impedances that will yield the desired phase outcome were obtained by starting with a base model using ideal components, where Z oe =Z oo =50Ω with an electrical length of 90 for each coupled section. Adjusting the even and odd mode impedances until a flat phase difference between 10 and 12GHz is obtained, keeping in mind that Z oe >Z oo, results in Z oe =66Ω and Z oo =42Ω. The Schiffman phase shifter is a differential phase shifter, in that the resulting phase is compared to a reference line. For a 45 phase shift, the reference line is 180 (total electrical length of the phase shifter) plus a 45 extra reference length. The results are: Over 10 12GHz, the phase difference between the reference line and the Schiffman phase shifter is 45 flat, with a return loss of -34.88dB at worst over the bandwidth. The even and odd mode impedance values obtained from the ideal model are applied to a model with includes the 121.5 mil substrate. The phase difference between the Schiffman phase shifter and the reference line are compared between 10GHz and 12GHz. There is a slight change in the return loss, with -31.07dB at worst, and no significant change in the phase difference (flat at 45 ). The Designer model yields widths and lengths for the copper traces that will be applied to the HFSS model The Schiffman phase shifter is designed primarily in HFSS. To speed up the analysis process, the reference line is not included in simulation since straight transmission line characteristics are well known and standard. The Designer model gives the length and width of the coupled lines. Port lines and the strip connecting the coupled lines is manually added and designed in HFSS. Chamfering was utilized to reduce reflection. The result was a phase shift centered at 45 ±2 and with a return loss of -15.69dB at worst over the bandwidth. The phase difference changed from the Designer model slightly, and the greatest change was seen in the reflection due to the additional copper. Port1 Port3 F=11GHz E=90deg ZO=42 ZE=66 Port2 Port4 E=(180+45) deg F=11GHz Z=50 Figure 7: Schematic View of Ideal Schiffman Phase Shifter Figure 8: Schiffman phase shifter model in HFSS, with vias. E 12

Component Schiffman Phase Shifter Figure 9: Ideal model phase difference, markers showing 45 at 10GHz, 11GHz, and 12GHz. Figure 10: Physical Designer model phase difference, markers showing 45 at 10GHz, 11GHz, and 12GHz. Figure 11: HFSS model phase difference, with markers showing 46.14 at 10GHz, 45.08 at 11GHz, and 42.99 at 12GHz. 13

Component Schiffman Phase Shifter Figure 12: Ideal model reflection. Markers at 10GHz (-34.88dB), 11GHz (-138.14dB), and 12GHz (-34.88dB). Figure 13: Physical Designer model reflection. Markers at 10GHz (-31.07dB), 11GHz (-73.76dB), and 12GHz (-31.07dB). Figure 14: HFSS model reflection. Markers at 10GHz (-15.69dB), 11GHz (-16.96dB), and 12GHz (-18.25dB). 14

Component Branchline Hybrid Ideal Ideal values were obtained from: A Multisection Broadband Impedance Transforming Branch-Line Hybrid; IEEE Transaction on Microwave Theory and Techniques. Formulas were given for theoretical values of impedances these formulas were applied for performance specifications and a line impedance of 50Ω. Results are as expected: over the entire bandwidth of 10-12 GHz there is approximately equal power division between ports two and three (-3.028 db ±0.139 db) and there is an exactly 90 phase difference between any two output ports. Figure 15: Schematic View of Ideal 90 Branchline Designer HFSS From the ideal model, a physical model was generated. Ansoft Designer s tune feature was used to tweak lengths and widths of transmission lines to yield the best approximation to the ideal model. The overall transmission for this model is -3.028dB ±0.139dB. Because of losses, the shape of the transmission was tuned to be different. This yielded results that are very close to the ideal. The Phase Differences at the output ports ranged from 90.04 to 90.36 across the bandwidth. Return Loss and Isolation, due to symmetry, were very alike. Both were, maximum, -25dB across the bandwidth. A final model of the Branchline was created in HFSS. This model was iteratively tuned between HFSS and Ansoft Figure 16: Layout View of HFSS 90 Branchline Designer. Through-holes were added to suppress higher order modes. The overall transmission for this model is -3.176 db ±0.125 db. Compared to both the Ideal and Physical Designer models, this is low; when run with lossless materials (substrate loss tangent set to zero, all copper replaced with perfect electric conductors, etc.), transmission is 3.03 db ±0.150d B. The Phase Differences at the output ports ranged from 89.75 to 90.06 across the bandwidth. Compared to the Physical Designer model, this is slightly lower, but still quite close to 90. Return Loss and Isolation, compared to the Physical Designer model, decreased only slightly. Across the 10-12GHz bandwidth these parameters -24.7 db or lower. 15

Y1 Component Branchline Hybrid Figure 17: Ideal Model Transmission Parameters (-3.025dB ±0.145dB) Figure 18: Physical Designer Model Transmission Parameters (-3.028dB ±0.139dB) Ansoft Corporation Transmission Tweaked 2.425-2.60-2.80-3.00 m2 m6 m4 Figure 19: HFSS Model Transmission Parameters (-3.176dB ±0.125dB) -3.20-3.40-3.60-3.80-4.00 m1 Curve Info db(s(1,2)) db(s(1,3)) db(s(2,1)) db(s(2,4)) db(s(3,1)) Name X Y m1 10.0000-3.2131 m2 10.0000-3.0836 m3 12.0000-3.3007 m4 12.0000-3.1028 m5 11.0000-3.2982 m6 11.0000-3.0569 m5 m3 db(s(3,4)) -4.20 db(s(4,2)) db(s(4,3)) -4.40 9.00 9.50 10.00 10.50 11.00 11.50 12.00 12.50 13.00 Freq [GHz] 16

Y1 [deg] Component Branchline Hybrid Figure 21: Ideal Model Output Port Phase Differences (+90.00 ±0.00 ) Figure 22: Physical Designer Model Output Port Phase Differences (+90.183 ±0.177 ) Ansoft Corporation Phase Differences Tweaked 2.425 92.00 Curve Info Name X Y 91.00 cang_deg(s(2,1))-cang_deg(s(3,1)) cang_deg(s(3,4))-cang_deg(s(2,4)) cang_deg(s(1,2))-cang_deg(s(4,2)) cang_deg(s(4,3))-cang_deg(s(1,3)) m1 10.0000 90.0175 m2 10.0000 89.8898 m3 11.0000 89.8684 m4 11.0000 89.7463 m5 12.0000 90.0662 m6 12.0000 89.8847 Figure 20: HFSS Designer Model Output Port Phase Differences (+89.9012 ±0.166 ) 90.00 m1 m2 m3 m4 m5 m6 89.00 88.00 9.00 9.50 10.00 10.50 11.00 11.50 12.00 12.50 13.00 Freq [GHz] 17

Y1 Component Branchline Hybrid Figure 24: Ideal Model Return Loss (Below -28.93dB) Figure 23: Physical Designer Model Return Loss (Below -25.01dB) Ansoft Corporation Return Loss Tweaked 2.425-15.00-20.00 Figure 25: HFSS Model Return Loss (Below -24.82dB) -25.00-30.00-35.00 m2 Curve Info db(s(1,1)) db(s(2,2)) db(s(3,3)) m3 Name X Y m1 10.0000-37.9792 m2 11.0000-26.7556 m3 12.0000-24.8218 db(s(4,4)) m1-40.00-45.00-50.00 9.00 9.50 10.00 10.50 11.00 11.50 12.00 12.50 13.00 Freq [GHz] 18

Y1 Component Branchline Hybrid Figure 27: Ideal Model Isolation (Below -29.22dB) Figure 26: Physical Designer Model Isolation (Below -25.03dB) Ansoft Corporation Isolation Tweaked 2.425-15.00-20.00-25.00 m3 Figure 28: HFSS Model Isolation (Below -24.73dB) -30.00-35.00 m1 m2 Curve Info db(s(1,4)) db(s(2,3)) db(s(3,2)) db(s(4,1)) Name X Y m1 10.0000-37.5629 m2 11.0000-26.6985 m3 12.0000-24.7360-40.00-45.00-50.00 9.00 9.50 10.00 10.50 11.00 11.50 12.00 12.50 13.00 Freq [GHz] 19

Port3 E E Z=$Z8 E=180deg F=11GHz E Port2 E E=360deg F=11GHz Z=50 E E=360deg F=11GHz Z=50 E=90deg F=11GHz Z=$Z1 E E=180deg F=11GHz Z=$Z3 E E=180deg F=11GHz Z=$Z5 E=90deg F=11GHz Z=$Z1 Component Knöchel Hybrid Ideal Ideal values were obtained from the paper: Broadband Printed Circuit 0 /180 Couplers and High Power Inphase Power Dividers; IEEE Transaction on Microwave Theory and Techniques. Ideal simulations are presented for a 50 line impedance. The impedance values were tuned slightly to minimize the return losses and maximize the transmitted power. Port1 E E=360deg F=11GHz Z=50 E E=90deg F=11GHz Z=$Z7 E E=90deg F=11GHz Z=$Z2 E E=90deg F=11GHz Z=$Z2 E E=90deg F=11GHz Z=$Z4 E E=90deg F=11GHz Z=$Z4 E E=360deg F=11GHz Z=50 The results are as expected. Over the entire desired bandwidth of 10-12GHz, Figure 29: Schematic View of Ideal 180 Hybrid Coupler there is equal power division between the output ports (-3dB) and a 180 phase difference between the output ports when power is applied to port four. All return losses are suppressed below -30dB. Port4 Physical HFSS When the Ideal model was exported to a physical model in Designer, all parameters became completely unacceptable. Through an iterative piecewise tuning process, a final model is presented. The transmission values of this model centered around -3.2dB with an error of ±0.15 db. The phase error is approximately 0.5 for both input ports. Return losses and isolation values between all ports are below -23dB over the full bandwidth. When the Designer Model was exported to HFSS, the discontinuities at all of the corners and losses experienced in all materials created a generally unacceptable component. Again, iterative piecewise tuning was implemented to improve the design. The final HFSS model has transmission values that are centered around -3.22dB with an error of ±0.75dB. The phase error is approximately 0.5. Chamfers and Through-holes were added to suppress reflections and higher order modes. Figure 30: Three Dimensional Model of Simulated HFSS 180 Coupler 20

Magnitude (db) Component Knöchel Hybrid Figure 33: Ideal Model Transmission Parameters (-3.02dB ±0.5dB) Figure 32: Physical Designer Model Transmission Parameters (-3.15dB ±0.15dB) Ansoft Corporation -2.00 Transmission Final Model -2.50 Curve Info db(s(waveport2,waveport1)) db(s(waveport2,waveport4)) db(s(waveport3,waveport1)) db(s(waveport3,waveport4)) Name X Y m1 10.00-3.15 m2 10.00-3.32 m3 11.00-3.16 m4 11.00-3.21 m5 12.00-3.18 m6 12.00-3.30 Figure 31: HFSS Model Transmission Parameters (-3.22dB ±0.75dB) -3.00 m1 m2 m3 m4 m5 m6-3.50-4.00 9.00 10.00 11.00 12.00 13.00 Freq [GHz] 21

Y1 [deg] Component Knöchel Hybrid Figure 35: Ideal Model Phase Parameters (± 0.4 Phase Error) Figure 34: Physical Designer Model Phase Parameters (± 0.5 Phase Error) Ansoft Corporation 2.00 1.50 HFSS Phase Differences Final Model Curve Info phasediff1 phasediff4 1.00 m3 m4 m5 0.50 m2 Figure 36: HFSS Model Phase Parameters (± 0.4 Phase Error with a 0.6 Phase Offset) 0.00-0.50-1.00-1.50 m1 Name X Y m1 10.00 0.09 m2 10.01 0.51 m3 11.00 0.77 m4 12.00 0.91 m5 12.00 0.81 22-2.00 8.00 9.00 10.00 11.00 12.00 13.00 14.00 Freq [GHz]

Y1 Component Knöchel Hybrid Figure 39: Ideal Model Return Loss Parameters (Below -30.98dB) Figure 38: Physical Designer Model Return Loss Parameters (Below -23.5dB) Ansoft Corporation -15.00 HFSS Return Losses Final Model m1 m3-20.00-25.00 m2 Figure 37: HFSS Model Return Loss Parameters (Below -17.25dB) -30.00-35.00-40.00 Curve Info db(s(waveport1,waveport1)) db(s(waveport2,waveport2)) db(s(waveport3,waveport3)) db(s(waveport4,waveport4)) -45.00 Name X Y m1 10.00-17.25 m2 11.00-25.34 23 m3 12.00-18.41-50.00 9.00 10.00 11.00 12.00 13.00 Freq [GHz]

Y1 Component Knöchel Hybrid Figure 42: Ideal Model Isolation Parameters (Below -33.48dB) Figure 41: Physical Designer Model Isolation Parameters (Below -19.56dB) Ansoft Corporation 0.00-10.00 HFSS Isolation Curve Info db(s(waveport1,waveport4)) db(s(waveport2,waveport3)) db(s(waveport3,waveport2)) Name X Y m1 10.00-17.89 m2 11.00-27.58 m3 12.00-23.69 Final Model db(s(waveport4,waveport1)) m1-20.00 m3 Figure 40: HFSS Model Isolation Parameters (Below -17.89dB) -30.00-40.00 m2-50.00 24-60.00 9.00 10.00 11.00 12.00 13.00 Freq [GHz]

SMA Launch HFSS The SMA Launch is used as a 50Ω interface from stripline to coax and is designed over the frequency range of DC to 18GHz. The original HFSS model of the SMA connector was provided by Anaren, as was the proposed assembly concept. The assembly concept was modeled in HFSS and simulated while varying different parameters to optimize transmission across the bandwidth. The higher frequencies, 15-18GHz, were the hardest to lower. Final results yielded Return Losses of -12.5dB or lower across the entire bandwidth of DC to 18GHz. This translates to a minimum of 94.38% of the power being transmitted through the launch. Looking at the 10-12GHz bandwidth for the components, we have Return Losses of -32dB or lower. This translates to a minimum 99.94% of the power being transmitted through the launch. Figure 43: Assembly Concept for SMA Launch Figure 44: HFSS Model of SMA Connector, provided by Anaren Figure 45: Modified Connector for Assembly Simulations 25

SMA Launch Figure 47: Full Assembly Model Figure 46: Full Assembly Model, Front View Figure 49: Closer View of Signal Path Figure 48: Front View of Full Assembly with Through-Holes 26

db(s(1,1)) db(s(1,1)) SMA Launch Ansoft Corporation 0.00 Return Loss 59646-0001 Final Launch Curve Info Name X Y db(s(1,1)) Setup1 : Sw eep1 m1 10.0000-32.4156 m2 11.0000-37.4734-10.00 m3 12.0000-41.9564 m4 18.0000-12.6610 m5 6.2900-24.5719 m4-20.00 m5 Figure 51: SMA Launch, Return Loss up to 18GHz. -30.00 m1 Results show at least 12.5dB down across the entire bandwidth. Sub - 30dB return loss from 10-12GHz m2-40.00 m3-50.00-60.00 0.00 5.00 10.00 15.00 20.00 Freq [GHz] Ansoft Corporation -25.00 Return Loss 59646-0001 Final Launch1-30.00 Curve Info db(s(1,1)) Setup1 : Sw eep1 Name X Y m1 10.0000-32.4156 m2 11.0000-37.4788 m3 12.0000-41.9575 m1-35.00 m2 Figure 50: SMA Launch, Return Loss, 9-13GHz -40.00 m3-45.00-50.00 27-55.00 9.00 9.50 10.00 10.50 11.00 11.50 12.00 12.50 13.00 Freq [GHz]

System A The Schiffman phase shifter was noted to have decreased isolation, especially at higher frequencies. It also causes a decrease in transmission and an increase in return loss. The majority of the phase error is attributed to the Schiffman phase shifter with error no greater than 4. The 90 coupler was noted to have decreased isolation especially at lower frequencies. It also causes a decrease in transmission and an increase in return loss. The 90 coupler did not have a significant impact on the phase error. The SMA Connector Launch model did not have a significant impact on isolation, return loss, transmission error or phase error. It did cause a decrease in transmission. Overall, the system has maintained frequencies close to the ideal at the boundaries of the frequency range but has greatly decreased isolation around 11GHz. The transmission has decreased by more than -0.5dB. The system also performs with the expected performance for the phase progression with error attributed to the Schiffman phase shifter of no greater than 4. Figure 52: Ideal Isolation for System A Figure 53: Isolation for System A Figure 54: Ideal Return Loss for System A Figure 55: Return Loss for System A 28

System A Figure 56: Ideal Transmission for System A Figure 57: Transmission for System A Figure 58: Ideal Transmission Error for output port 1 for System A Figure 59: Transmission Error for output port 1 for System A Figure 60: Ideal Transmission Error for output port 2 for System A Figure 61: Transmission Error for output port 2 for System A 29

System A Figure 62: Ideal Transmission Error for output port 3 for System A Figure 63: Transmission Error for output port 1 for System A Figure 64: Ideal Transmission Error for output port 4 for System A Figure 65: Transmission Error for output port 4 for System A Figure 66: Ideal Phase Error for input port 1 for System A Figure 67: Phase Error for input port 1 for System A 30

System A Figure 68: Ideal Phase Error for input port 2 for System A Figure 69: Phase Error for input port 1 for System A Figure 70: Ideal Phase Error for input port 3 for System A Figure 71: Phase Error for input port 3 for System A Figure 72: Ideal Phase Error for input port 4 for System A Figure 73: Phase Error for input port 4 for System A 31

System A Figure 74: Ideal Phase Difference for input port 1 for System A Figure 75: Phase Difference for input port 1 for System A Figure 76: Ideal Phase Difference for input port 2 for System A Figure 77: Phase Difference for input port 2 for System A Figure 78: Ideal Phase Difference for input port 3 for System A Figure 79: Phase Difference for input port 3 for System A 32

System A Figure 80: Ideal Phase Difference for input port 4 for System A Figure 81: Phase Difference for input port 4 for System A 33

System B The 180 coupler resulted in the majority of the isolation loss and also increased the return loss and transmission error. The coupler was noted to have decreased transmission except from input ports 3 and 4; this is due to the topology of the system. The majority of the phase error can be attributed to the 180 coupler. The 90 coupler did not have a significant impact on the isolation loss or phase error. The coupler increased return loss, specifically at higher frequencies, and transmission error. The coupler decreased transmission from input ports 3 and 4; this is due to the topology of the system. The SMA Connector Launch model did not have a significant impact on isolation, return loss, transmission error or phase error. It did cause a decrease in transmission. Overall, the transmission has decreased by more than -0.5 db. The system also performs with the expected performance for the phase progression with error attributed to the 180 coupler of no greater than 1. Some of the loss of system performance is caused by the bends used to connect the components. Figure 82: Ideal Isolation for System B Figure 83: Isolation for System B Figure 84: Ideal Return Loss for System B Figure 85: Return Loss for System B 34

System B Figure 86: Ideal Transmission for System B Figure 87: Transmission for System B Figure 88: Ideal Transmission Error for output port 1 for System B Figure 89: Transmission Error for output port 1 for System B Figure 90: Ideal Transmission Error for output port 2 for System B Figure 91: Transmission Error for output port 2 for System B 35

System B Figure 92: Ideal Transmission Error for output port 3 for System B Figure 93: Transmission Error for output port 3 for System B Figure 94: Ideal Transmission Error for output port 4 for System B Figure 95: Transmission Error for output port 4 for System B Figure 96: Ideal Phase Error for input port 1 for System B Figure 97: Phase Error for input port 1 for System B 36

System B Figure 98: Ideal Phase Error for input port 2 for System B Figure 99: Phase Error for input port 2 for System B Figure 100: Ideal Phase Error for input port 3 for System B Figure 101: Phase Error for input port 3 for System B Figure 102: Ideal Phase Error for input port 4 for System B Figure 103: Phase Error for input port 4 for System B 37

System B Figure 104: Ideal Phase Progression for input port 1 for System B Figure 105: Phase Progression for input port 1 for System B Figure 106: Ideal Phase Progression for input port 2 for System B Figure 107: Phase Progression for input port 2 for System B Figure 108: Ideal Phase Progression for input port 3 for System B Figure 109: Phase Progression for input port 3 for System B 38

System B Figure 110: Ideal Phase Progression for input port 4 for System B Figure 111: Phase Progression for input port 4 for System B 39

Two Quarter Schedule/Milestones Major Milestone Chart (Milestones from 11 week schedule): Task Expected Date Responsible Team Member Modified Completion Date Comments Initial Research 12/18/08 Butler Matrices All 12/12/08 Benchmarking All 12/12/08 180 Degree Coupler Joel 12/12/08 90 Degree Coupler Michael 12/12/08 Phase Shifter Mia/Amanda 12/12/08 Miniaturization Amanda Time permitting Presentation All Presenting to each other Ideal Simulations 180 Degree Coupler 1/6/09 Joel 1/6/09 90 Degree Coupler 1/12/09 Michael 1/12/09 Phase Shifter 1/12/09 Mia/Amanda 1/12/09 System A 1/16/09 Mia 1/16/09 System B 1/23/09 Amanda 1/18/09 Detailed Design Simulations of Components 1/30/09 All 1/30/09 System Level 2/13/09 Amanda/Joel 2/10/09 Design/Integration System Simulations 2/13/09 Amanda 2/9/09 Vertical Launch 2/13/09 Anaren/Michael 2/9/09 Manufacturing Cleanup 2/20/09 Laying out on CAD software Michael/Joel/Amanda In progress Manufacturing 4/24/09 Anaren Test and Verification 5/15/09 All System A 5/15/09 All System B 5/15/09 All 40