I have seen in some books, in the transformer-coupled Class A amplifiers
If use the emitter follower configuration, we can get a high input swing and for the collector connected version, we can obtain a voltage gain.
Now, my question is what is the best practice or the correct way?
Putting the transformer in the emitter (ie CC mode) is rare and has the disadvantage of requiring more drive power.
It is extremely unusual and not really a good example for a teaching book to use.
In the vast majority of cases, a push-pull configuration would be used in a power amplifier to improve efficiency and cancel the magnetic field created in the transformer as a result of the collector current. In a common-emitter configuration, the two collectors are at different AC potentials and require insulation from each other and the heat-sink. If the emitter follower arrangement is used they are at a common potential and so can be mounted directly to a common heat-sink thus reducing the thermal resistance of the system.
A big disadvantage is a requirement of a high-voltage drive to the base - the base will need to swing from slightly above the positive rail to far below the negative rail (assuming NPN devices). This complicates the driver circuit considerably. The only practical way to do this is to use a transformer from the driver to the output stage.
This arrangement was rare back in the early years of transistor power amplifiers - I would not expect any modern design to be lke this. One reason is that the relative cost of transformers and active devices has radically changed. Now transformers are expensive, transistors are virtually free. Whereas in the past transistors were the expensive item.
