Boonton 4500B Peak Power Meter - Manual (Page 326)

Boonton 4500B RF Peak Power Analyzer
Application Notes 6-16
6.5 Measurement Accuracy
The Model 4500B includes a precision, internal, 1 GHz RF reference calibrator that is traceable to the National Institute for
Standards and Technology (NIST). When the instrument is maintained according to the factory recommended one year
calibration cycle, the calibrator enables you to make highly precise measurements of CW and modulated signals. The error
analyses in this chapter assumes that the power analyzer is being maintained correctly and is within its valid calibration
period.
Measurement uncertainties are attributable to the instrument, calibrator, sensor, and impedance mismatch between the sensor
and the device under test (DUT). Individual independent contributions from each of these sources are combined
mathematically to quantify the upper error bound and probable error. The probable error is obtained by combining the linear
(percent) sources on a root-sum-of-squares (RSS) basis. RSS uncertainty calculations also take into account the statistical
shape of the expected error distribution.
Note that uncertainty figures for individual components may be provided given in either percent or dB. The following
formulas may be used to convert between the two units:
U
%
= (10 (UdB/10) - 1) × 100 and U
dB
= 10 × Log
10
(1 + (U
%
/ 100))
Section 6.5.1 outlines all the parameters that contribute to the power measurement uncertainty followed by a discussion on
the method and calculations used to express the uncertainty.
Section 6.5.2 continues discussing each of the uncertainty terms in more detail while presenting some of their values.
Section 6.5.3 provides Power Measurement Uncertainty calculation examples for Peak Power sensors with complete
Uncertainty Budgets.
6.5.1 Uncertainty Contributions. The total measurement uncertainty is calculated by combining the following terms:
Uncertainty Source Distribution Shape K
1. Instrument Uncertainty Normal 0.500
2. Calibrator Level Uncertainty Rectangular 0.577
3. Calibrator Mismatch Uncertainty U-shaped 0.707
4. Source Mismatch Uncertainty U-shaped 0.707
5. Sensor Shaping Error Rectangular 0.577
6. Sensor Temperature Coefficient Rectangular 0.577
7. Sensor Noise Normal 0.500
8. Sensor Zero Drift Rectangular 0.577
9. Sensor Calibration Factor Uncertainty Normal 0.500
The formula for worst-case measurement uncertainty is:
U
WorstCase
= U
1
+ U
2
+ U
3
+ U
4
+ ... U
N
where U
1
through U
N
represent each of the worst-case uncertainty terms.