Push-pull amplifier designs can run their active driving devices in class-A, -B, or -C service, but your choice of class affects the impedances presented to those devices.
Please consider a transformer driven load resistance in the following configuration, where “1T” and “2T” denote turns ratio, and the impedance presented to current source I1 is the 50 Ω of the load divided by the square of the turns ratio (Figure 1). This comes as no surprise to be 12.5 Ω.
Figure 1 This configuration of a transformer driven load resistance has one current source, I1.
Now consider a transformer driven load resistance where a second driving current source, I2, is added to the circuit as follows (Figure 2).
Figure 2 Configuration 2 of the transformer driven load resistance has an additional driving current source.
The effect of the second current source is to double the impedance being presented to the first current source. Since the burden of delivering power to that 50 Ω load is now shared by two active sources, the current demanded for the total effort is shared between the two sources. The impedance transformation via the transformer is no longer dependent only on the transformer turns ratio. As a result, the 12.5 Ω originally seen by I1 becomes 25 Ω.
That effect can be seen in the simulation of a class-A push-pull amplifier in SPICE as follows.
Figure 3 This class-A push-pull amplifier SPICE simulation shows the effect of two current sources.
In class-B push-pull service, where the opposing transistor or MOSFET would go to cut-off, the presented impedance to the active device would still be 12.5 Ω per the turns ratio. However, this does not hold true for class-A service. Instead, we note that the slope of the Q1 load line for class-A push-pull service is 25 Ω, as predicted by the circuit model seen in Figure 2.
This principle was brought to bear in a Class-A linear amplifier that I designed for power amplifier service, which was part of US Navy CVA V.A.S.T., the building block BB21 signal generator that delivered +30 dBm into 50 Ω from 10 kHz to 40 MHz.
This article was originally published on EDN.
John Dunn is an electronics consultant, and a graduate of The Polytechnic Institute of Brooklyn (BSEE) and of New York University (MSEE).
Related articles: