Characterization of a Printed Circuit Board Via

Transcription

1 Characterization of a Printed Circuit Board Via Brock J. LaMeres Thesis Defense May 25, 2000 Department of Electrical and Computer Engineering University of Colorado Colorado Springs, CO

2 Objective To Develop an Equivalent Circuit Model for a Printed Circuit Board Via

3 Purpose To Characterize the Discontinuity Caused by a Via To Understand How the Physical Dimensions of a Via Contribute to its Electrical Response

4 Characterization Approach Develop an Equivalent Circuit 3D Electromagnetic Field Simulations SPICE Simulations Time Domain Reflectometry Network Analysis

5 Printed Circuit Board Via (Cross Section)

6 Printed Circuit Board Via (Actual Via)

7 Equivalent Circuit (Coupled Model)

8 Equivalent Circuit (Distributed Model)

9 Equivalent Circuit (Lumped Model)

10 Equivalent Circuit Response Lumped Eq. Circuit Perfect Step Voltage Out (V) Distributed Eq. Circuit Coupled Eq. Circuit Ideal Step Input Time (ns) ps Guassian Step Voltage Out (V) Lumped Eq. Circuit Distributed Eq. Circuit Coupled Eq. Circuit Guassian Step Input Time (ns)

11 3D EM Field Simulations (Varying Pad Radius) Capacitance vs. Pad Radius Via Capacitance (ff) Rcyl =.009", Rgnd =.023" Rcyl =.006", Rgnd =.018" Rcyl =.004", Rgnd =.014" Pad Radius (.001") Rcyl =.004", Rgnd =.014" Inductance vs. Pad Radius Via Inductance (ph) Rcyl =.009", Rgnd =.023" Rcyl =.006", Rgnd =.018" Pad Radius (.001") 50.0 Zo vs. Pad Radius Characteristic Impedance (Ohms) Rcyl =.004", Rgnd =.014" Rcyl =.006", Rgnd =.018" Rcyl =.009", Rgnd =.023" Pad Radius (.001")

12 3D EM Field Simulations (Varying Cylinder Radius) 500 Via Capacitance (ff) Rpad = +.005", Rgnd = +.010" Rpad = +.006", Rgnd = +.012" Rpad = +.007", Rgnd = +.014" Capacitance vs. Cylinder Radius Cylinder Radius (.001") 550 Via Inductance (ph) Rpad = +.005", Rgnd = +.010" Rpad = +.006", Rgnd = +.012" Rpad = +.007", Rgnd = +.014" Inductance vs. Cylinder Radius Cylinder Radius (.001") 65 Characteristic Impedance (Ohms) Rpad = +.005", Rgnd = +.010" Rpad = +.006", Rgnd = +.012" Rpad = +.007", Rgnd = +.014" Zo vs. Cylinder Radius Cylinder Radius (.001")

13 3D EM Field Simulations (Varying Ground Clearance Radius) Rpad =.016", Rcyl =.009" Capacitance vs. Ground Clearance Radius Via Capacitance (ff) 250 Rpad =.012", Rcyl =.006" 200 Rpad =.009", Rcyl =.004" Ground Backoff Radius (.001") 450 Rpad =.009", Rcyl =.004" Inductance vs. Ground Clearance Radius Via Inductance (ph) 400 Rpad =.012", Rcyl =.006" 350 Rpad =.016", Rcyl =.009" Ground Backoff Radius (.001") 60 Zo vs. Ground Clearance Radius Characteristic Impedance (Ohms) Rpad =.009", Rcyl =.004" Rpad =.012", Rcyl =.006" Rpad =.016", Rcyl =.009" Ground Backoff Radius (.001")

14 Time Domain Reflectometry (Experimental Setup) Circuit of Experimental Setup Test Printed Circuit Board Actual Laboratory Setup

15 Time Domain Reflectometry (Empirical Results) TDR varying Pad Radius (Zo decreases) Vr/Vi (V) Rpad =.010" Rpad =.015" Rpad =.020" Rpad =.025" Rpad =.030" Time (ns) TDR varying Cylinder Radius (Zo decreases) Vr/Vi (V) Rcyl =.004" Rcyl =.005" Rcyl =.006" Rcyl =.008" Rcyl =.009" Time (ns) TDR varying Ground Clearance Radius (Zo increases) Vr/Vi (V) Rgnd =.036" Rgnd =.031" Rgnd =.026" Rgnd =.021" Rgnd =.016" Time (ns)

16 Network Analysis (Experimental Setup) Circuit of Experimental Setup Actual Laboratory Setup

17 Network Analysis (Empirical Data) S11 (db) (reflected) varying cylinder radius S11 (db) Rcyl =.009" Rcyl =.006" Rcyl =.004" Frequency (GHz) S21 (db) (transmitted) varying cylinder radius S21 (db) Rcyl =.004" Rcyl =.009" Frequency (GHz)

18 Simulation vs. Empirical Data (Time Domain Reflectometry) 0.25 Pad Radius Example Vr/Vi (V) TDR Simulation Time (ns) TDR Sim ulation Cylinder Radius Example Vr/Vi (V) Time (ns) TDR Simulation Ground Clearance Radius Example Vr/Vi (V) Time (ns)

19 Simulation vs. Empirical Data (Network Analysis) 0-10 S11 (db) S11 (db) Empirical Simulation (w /o connector model) Simulation (w / connector model) Frequency (GHz) 4 0 S21 (db) S21 (db) Empirical Simulation (w/o connector model) Simulation (w/ connector model) Frequency (GHz)

20 Effect of Via on S-Parameters (including trace and pcb connectors) 0 S11 (db) Simulation with via Simulation without via S11 (db) (with and without via) Frequency (GHz) 4 S21 (db) Simulation with via Simulation without via S21 (db) (with and without via) Frequency (GHz)

21 Design Guidelines (for minimizing the via discontinuity) [1] Use the minimum size drill bit for creating the via cylinder. [2] Use the minimum size via pad radius. [3] Do not use the minimum size ground clearance radius. [4] Use the thinnest printed circuit board possible. [5] Place ground vias around the signal via when passing through multiple ground planes.

22 Conclusion An Equivalent Circuit was presented that can accurately characterize a printed circuit board via. When used with 3D EM Field Simulation, the actual response of the via can be predicted.

23 Characterization of a Printed Circuit Board Via Brock J. LaMeres Questions or Comments? Department of Electrical and Computer Engineering University of Colorado Colorado Springs, CO