Power Amplifier Matching

The purpose of the Power Amplifier (PA) matching is to achieve the amplifier’s best performance. An ideal PA is physically not possible, and side effects must be considered during the PA matching. The matching components of the PA affect the following listed parameters:

Monitor these three parameters during the matching process. The goal is to achieve the maximum output power with the lowest possible current consumption and lowest possible harmonics. The simplified circuit diagram of the power amplifier used in the ATA8510 is shown in the following figure.

Figure 1. Simplified Circuit Diagram Power Amplifier


The simplified PA is a current source with a certain internal resistance. Additional parasitics from the internal circuit, pad and bond wire must be considered for the matching. The current can be configured via the FEPAC register. For more details on the register, refer to the user manual of the product. As the current and the resistance value change with each FEPAC value, every output power setting requires a separate matching network.

The PA matching differs from the method applied for the receiver site as the PA is a source and generates a signal. In principle, the PA behaves the same as NWA on only one single frequency without considering the reflection. The goal is to transfer all the energy through the matching network in combination with the maximum efficiency of the PA. According to the power transfer theorem, match the load impedance with the complex conjugate of the impedance that is visible at the pin of the device. As the required matching cannot be measured, it must be estimated using the subsequent listed assumptions:

Figure 2. Voltage Inside of the PA
Ueff=(VSpa2Vsat)2

The following figure shows the typical matching layout used for the PA of the ATA8510.

Figure 3. Layout Example for PA Matching


For the ATA8510, the serial component is an inductor and the shunt component is a capacitor. This is mandatory when an RF antenna switch is used as it is biased with a DC from the PA.

With the given parameter and the listed assumptions, the internal resistance for an assumed output power of +14 dBm can be calculated with the following equation.

Figure 4. Internal Resistor
Ri=Ueff2P[mW]

The internal resistor value is in the range of 39Ω and, with the assumed parasitic of 3 pF, a complex impedance of ZS=(39j122)Ω can be calculated. The external matching, including the bond wire, must transfer the Z0 = 50Ω to the PA impedance's complex conjugate value. The target point for the Smith Chart transformation when assuming Z0 is the load impedance, is ZS*=(39+j122)Ω . This results in the situation shown in the following figure.

Figure 5. PA Matching Circuit with Impedance’s


The required matching components can be calculated with simulation tools. For the +14 dBm example, the calculated values to transform the 50Ω impedance to the complex value are L = 51 nH and C = 4.3 pF. The calculated values must be corrected by the known parasitic influences from the PCB and the bond wire. The outcome is a start value for the matching process with respect to the current consumption and the harmonics. The values of the matching elements must be adapted until the required values for the three parameters are reached.

Notes: