board, open and short measurements will need to be completed first to extract the L and C
values of the test board without the connector. Next, the measurements are taken to obtain
the total L and C values of the connector plus the test board. The two results are subtracted,
resulting in the L and C values of the connector. This typically requires two separate test
structures, one to extract the board parasitics and one to extract the total board and DUT
parasitics. These structures should be placed in close proximity to each other to minimize
differences across the board due to manufacturing variations.
Figure 11.28: Simple example of a socket test board used to measure the parasitics.
One-Port Impedance Measurement.
Impedance measurements can be cal-culated based on inductance and capacitance data
obtained per unit length using the self-value procedures described. An alternative method is
to use the VNA in TDR mode. Most VNAs have time-domain capability using the inverse
FFT (fast Fourier transform) to translate the frequency-domain data to a time-domain
response. When this is done, the time-domain response will be as discussed earlier in the
chapter for TDR. Therefore, measurements should be extracted in a similar manner. To
achieve the highest accuracy, the VNA output impedance should be calibrated to the
nominal design impedance. Another nice feature of using the VNA in the time domain is the
ability to set the system bandwidth and thereby tune the edge rate to match the design.
Transmission Line (R, L, G, C).
Transmission line parameters can be derived from both open and short S
11
measurements.
General common method for model derivation or validation is to utilize modeling software to
extract the parasitics from the measured data. The measured data, in terms of magnitude
and phase versus frequency, can be input into the modeling software to curve fit the
frequency-domain response and thereby yield the RLCG values used to accurately model
lossy, frequency-dependent models.
11.8.4. Two-Port Measurements (T
d
, Attenuation, Crosstalk)
Two-port measurements such as S
21
and S
12
are used to measure the forward and reverse
insertion loss. This is used to measure mutual terms, attenuation, and propagation delay.
Mutual Capacitance and Inductance of an Electrically Short Structure.
The techniques presented here are very useful for measuring the mutual capacitance and
inductance between two electrically short structures, such as connector or socket pins.
Again, the structure is considered short if the electrical length (delay) is much smaller than
/4.