Characterizing Passive Components

1
Characterizing Passive Components

• Ahmad Fallah (CIENA CSD) (completed while at
  PIC-JDSTG).
• Lincoln Davidson (Analog Design, LLC).
• Debashis Banerjee,
• Robert Nelson, David Rogers
• (North Dakota State University).

2
Outline

• Test Fixture and Setup
• Test Uncertainties
• Error Correction Methods
• TRL/TRM Method
• The Measurements
• Setup Verification
• S-to-Z Conversions
• The Really-Short Line
• Notes on Calculations for C and L Values
• Summary
• Future Work

3
Test Fixture and Setup

• The 50-Ohm line Test Jig

4
Test Uncertainties

• Shift in the measurement plane
• Connector and line mismatches
• Length of the test fixture

5
Error Correction Methods

• Mathematical method
• ABCDD ABCD-1 ABCD m ABCD
• TRL Method
• Use common standards and available transmission
  lines to calibrate out the unwanted effects.

6
TRL/TRM Method

• Thru (a short piece of transmission line),
• Reflection (an open),
• Isolation,
• Match

7
Measurements

• For the 50-ohm matched test fixture, it should be
  verified that the cursor remains at the center of
  the Smith chart (e.g., VSWR 1)
• A Shorter line (fixture was used) to avoid these
  problems specially at Higher frequencies.

8
Setup VerificationS21 plot for the 2-cm test
fixture from 100 MHz to 1000 MHz
9
S-to-Z Conversions

• For a Shunt Element
• ZSH (Z0S21)/(2(1-S21)) R j (?L
  1/?C)
• For a Series Element
• Zs 2Z0(1-S21)/S21
• R Real Z
• ?L 1/?C Im Z

10
The Really-Short Line

• A very-short piece of the line was cut of a test
  fixture, such that the SMA connectors butted
  against each other.
• Not much more improvements in the accuracy of
  the measurements was achieved.
• The Point
• It is easier to solder components on a larger
  test fixture.

11
100-nF capacitor parametersApprox. Resonance
Freq.21.1 MHz, C138.69 nF, L2.02 nH
12
10-nF capacitor parameters Note there is a typo
in the paper
13
2.2-nF capacitor parametersApprox. Resonance
Freq.134.65 MHz, C2.94 nF, L1.15 nH
14
Notes on Calculations for C and L Values

• To calculate the values of C and L, equation 5
  was solved at the first two data points.
• The measured results agree with the published
  (nominal) value for the capacitance value for
  some of the components, but not for all
  measurements.
• calculations result in different values if
  performed at different frequencies.
• to obtain improved results, better numerical
  techniques need to be employed

15
470-pF capacitor parametersApprox. Res.
Freq.334.56 MHz, C0.50 nF, L6.50 nH
16
Ferrite Bead Parametersat f 100 MHzZ624.565
ohms, R58.6578 ohms, X621.804 ohms
17
Ferrite-Capacitor ComboModeled as a shunt
element
18
Series impedance parameters for the L-C filter
19
Summary

• It is possible to use a simple and inexpensive
  device for characterizing passives.
• Calibrate out the uncertainties using the TRL/TLM
  method.
• Method is most effective in determining the dip
  vs. frequency.

20
Future Work

• Better numerical techniques need to be employed.
• Work in conjunction with our other work. I.e.,
  verify the results of the components measurement
  on live boards.
• Improve the frequency and impedance value
  accuracy and resolution by acquiring a higher
  number of data points.
• Include measurements taken with an expensive
  measurement system.
• Compare components of different manufacturers to
  each other.