Agilent Technologies ENA Series User Manual page 190

Calibration
Changing the Calibration Kit Definition
Loss
C0, C1, C2, C3
L0, L1, L2, L3
In most existing calibration kits, THRU standards are defined as "zero-length THRU," i.e.,
the delay and loss are both "0". Such a THRU standard does not exist, however.
Calibration must be done with two test ports interconnected directly.
NOTE The measurement accuracy depends on the conformity of the calibration standard to its
definition. If the calibration standard has been damaged or worn out, the accuracy will
decrease.
184
determined through measurement or by dividing the exact physical
length of the standard by the velocity coefficient.
This is used to determine the energy loss caused by the skin effect
along the length (one-way) of the coaxial cable. Loss is defined using
the unit of Ω/s at 1 GHz. In many applications, using the value 0 for
the loss should not result in significant error. The loss of a standard is
determined by measuring the delay (sec.) and the loss at 1 GHz and
then substituting them in the formula below.
( ) Z ×
loss dB
Ω
⎛ ⎞
--- -
Loss = -------------------------------------------------------- -
⎝ ⎠
( ) × delay s ( )
s
4.3429 dB
It is extremely rare for an OPEN standard to have perfect reflection
characteristics at high frequencies. This is because the fringe
capacitance of the standard causes a phase shift that varies along with
the frequency. For internal calculation of the analyzer, an OPEN
capacitance model is used. This model is described as a function of
frequency, which is a polynomial of the third degree. Coefficients in
the polynomial may be defined by the user. The formula for the
capacitance model is shown below.
( ) ( × ) (
C = C0 + C1 F + C2 F
F: measurement frequency
C0 unit: (Farads) (constant in the polynomial)
C1 unit: (Farads/Hz)
C2 unit: (Farads/Hz2)
C3 unit: (Farads/Hz3)
It is extremely rare for a SHORT standard to have perfect reflection
characteristics at high frequencies. This is because the residual
inductance of the standard causes a phase shift that varies along with
the frequency. It is not possible to eliminate this effect. For internal
calculation of the analyzer, a short-circuit inductance model is used.
This model is described as a function of frequency, which is a
polynomial of the third degree. Coefficients in the polynomial may be
defined by the user. The formula for the inductance model is shown
below.
( ) ( × ) (
L = L0 + L1 F + L2 F
F: Measurement frequency
L0 unit: [Farads] (the constant in the polynomial)
L1 unit: [Farads/Hz]
L2 unit: [Farads/Hz2]
L3 unit: [Farads/Hz3]
( ) Ω
0
2 3
× ) ( × )
+ C3 F
2 3
× ) ( × )
+ L3 F
Chapter 4