The procedure
- Select the frequency range and number of points appropriate to the load and its behavior.
- Perform a one-port calibration with a short, open and standard load at the connector where the load under test will be connected. This removes the effect of the cables and places the reference plane at that connector.
- Connect the load to be measured without moving the cable or bending it sharply.
- Read S11, which is the reflection coefficient at that port.
How to read the data
- Magnitude in decibels. It equals minus the return loss; the more negative, the better the match.
- Phase. It indicates the position of the reflection relative to the reference plane.
- Smith chart. It represents the coefficient on a plane whose center is a perfect match. From it the normalized impedance is read, which relates the load impedance to the reference impedance by the formula: impedance equals the reference impedance times (1 + Γ) divided by (1 − Γ).
- VSWR. Derived from the magnitude.
What limits the measurement
The smallest reflection that can be measured reliably is limited by the residual directivity of the system, which is determined by the quality of the calibration load and of the connectors. A very well-matched load is measured with high uncertainty if the system does not have enough directivity. Cable reflections, cable movement and dirty connectors also contribute error.
When the load has different connectors
If the load does not have the same connector type as the port, the calibration must be performed at the load connector, or an adapter must be used whose contribution is included in the calibration. An uncharacterized adapter adds as an extra reflection.
What to check first
If the results seem unstable, verify with a known standard, inspect the condition of the connectors and confirm that the calibration is still valid.