Chapter 2 31
Measurement Concepts
Calibration/Normalization
Measurement Concepts
Calibration/Normalization
Types of Measurement Errors
There are three basic sources of measurement error: systematic,
random and drift.
• Systematic Errors
— are due to imperfections in the analyzer and the test setup
— are assumed to be time invariant (predictable)
— are characterized by the calibration, and can be removed during
measurements
•Random Errors
— vary with time in a random fashion (unpredictable)
— are mainly a result of instrument noise (source phase noise,
analyzer sensitivity)
— cannot be removed
• Drift Errors
— are due to instrument or test-system performance changes after a
calibration has been done (unpredictable)
— are primarily caused by temperature variations
— can be removed by re-doing the calibration
So doing a calibration before making the measurement improves the
measurement accuracy. Any time you change the frequency, power
settings, or cables within a measurement, you must repeat the
calibration routine.
Open/Short Calibration
An open/short calibration is used for reflection measurements and only
corrects for system tracking errors (source match and reflection
tracking). This type of calibration is essentially a normalized
measurement where a reference trace is stored in memory. This
memory trace data is then subtracted from later measurement data.
Press
Source, Open/Short Cal to access this functionality.
The reference trace is created by terminating the directional coupler or
bridge in a short and also in an open. These two sets of data are then
averaged together. The resulting trace data is used as the reference
trace to correct future measurements. This calibration data must be
To Next Page
Loading...
Need help?
General
