**1 Introduction**
In the design of RF amplifying circuits, impedance matching between the input and output is a critical aspect. The design of the matching network is essential for achieving optimal performance. When the required impedance is known, RF circuit design software like RFSim99 can automatically generate a matching network, making the process much more efficient. In cases where the impedance matching requirements are not strict or other performance parameters are of greater concern, the input and output impedances of the device can often be approximated. However, this approach may require considering the parameter dispersion of the device itself, as long as the design error remains within acceptable limits.
In the case of RF power amplifier design, particularly in the driver stage and power output stage, it is crucial to maximize power gain and efficiency. Therefore, accurately knowing the input and output impedances of these stages is essential. Typically, the input and output impedances at specific frequencies and power levels are provided in the manufacturer's data sheet for power transistors, offering engineers a reference point. However, due to variations in operating conditions—such as frequency, temperature, bias, supply voltage, input power, and output power—the actual values may differ significantly from those listed in the manual. To reduce power consumption and improve efficiency, it is sometimes necessary to measure the input and output impedance of the power transistor under real working conditions. While a network analyzer is the preferred tool for such measurements, it is not always available. In such cases, ordinary instruments like oscilloscopes and impedance meters can also be used. This paper outlines a practical method for measuring the input and output impedance of an RF power transistor using common instruments.
**2 General Method of Impedance Measurement**
Impedance measurement typically involves three main methods: the bridge method, the resonance method, and the voltammetry method. The bridge method is known for its high accuracy and is widely used for precise measurements. However, when dealing with active nonlinear devices like RF power transistors, especially under real working conditions, the bridge method becomes challenging due to the complexity of the signals involved. Similarly, the resonance method is also difficult to apply in such scenarios, mainly because the waveforms produced by nonlinear large signals are not sine waves.
The voltammetry method, on the other hand, is one of the most traditional approaches. It relies on Ohm’s law, where impedance ZX is calculated as ZX = UXejθ / IX, with UX being the effective voltage across the impedance, IX the current through it, and θ the phase difference between them. However, the base and collector voltages and currents of an RF power transistor are not sinusoidal, making it difficult to accurately measure the fundamental components and phase angle. As a result, this method has significant limitations in this context.
Given these challenges, alternative methods are needed. The following section describes an indirect measurement technique that addresses issues like harmonic filtering and the need to test the power transistor under real operating conditions. This method has been proven to be simple and effective.
**3 Indirect Impedance Measurement Using the Transfer Function Method**
In Figure 1, the networks HA, HB, and ZX form a test setup, while HC represents its equivalent network. HA and HB are passive linear two-port networks designed to provide matching, isolation, and filtering, ensuring a clean sine wave is observed at bb'. The transfer function of HC can be expressed as:
$$
H_C = \frac{U_{aa'}}{U_{bb'}} e^{j\theta}
$$
Where $ U_{aa'} $ and $ U_{bb'} $ are the effective voltages at points aa' and bb', respectively, and $ \theta $ is the phase difference between them. Once these values are measured, the transfer function $ H_C $ can be determined. Since HA and HB are known linear networks, the unknown impedance $ Z_X $ can be calculated accordingly.
**4 Principles of Test Network Design**
First, the design of HA and HB should be as simple as possible based on the application requirements. A more complex network increases both computational effort and potential errors.
Second, the selection of components for HA and HB should prioritize resistors, capacitors, and inductors that closely resemble ideal models. Inductors should be used sparingly due to their limited Q factor and complex real-world behavior. Before use, all component values must be precisely measured with a precision impedance meter, and care should be taken to minimize parasitic effects during circuit assembly.
Third, the power transistor must operate normally during testing, and the network should be either resonant or slightly off-resonance (with a low-Q tank). The measured parameters are only meaningful under specific operating conditions and frequencies.
Finally, the probe capacitance at bb' should be minimized, and the input resistance of the probe should be as high as possible. Only the probe’s capacitance needs to be considered in calculations, and its size must be measured before testing.
**5 Example of Measuring Input and Output Impedance of an RF Power Transistor**
The application manual of an RF power transistor usually provides its input and output impedances under certain operating conditions. When designing an RF power amplifier, if the transistor operates in the typical state described in the manual, the provided impedance values can be directly used. Although there may be some parameter variation, the error is generally small. However, when the operating conditions change—especially the frequency—the values in the manual may no longer be accurate and should only be used as a reference.
For example, the input and output impedance data for the VSC band RF power transistor 2SC2630 from Mitsubishi Electric is: Zin = 0.8 + j1.2 Ω, Zout = 1.5 - j0.6 Ω, @Po = 60 W, VCC = 12.5 V, f = 175 MHz. Another example is the VHF band RF power transistor 2SC1971, which has Zin = 0.8 + j3.2 Ω, Zout = 6.2 - j3 Ω, @Po = 6 W, VCC = 13.5 V, f = 175 MHz. These values serve as useful references but should be verified under actual operating conditions for precise design.
Jiangsu Qilong Electronic Technology Co., Ltd. , https://www.qilongtouch.com