Using an additional, linear final amp permits post-saturation placement of time fx, as follows:
Preamp - somewhat important
Saturating power amp - most important
Linear final amp - least important
The use of the terms "good" and "bad" in this classification system follows from the dominant importance of the type of saturating power amp, compared to the lesser importance of the type of preamp.
For example, "ST" designates solid-state preamp, tube power amp.
For example, "TTS" designates tube preamp, tube power amp, and solid-state final amp.
It is not very likely that a *solid-state* power amp driving a dummy load would be of great interest; we can ignore those 4 permutations. Thus we can say that a 3-stage amp-rig is this general rig architecture:
preamp (tube or solid-state)
tube power amp driving a load
final amp (tube or solid-state)
I consider the TTS configuration to be the classic 3-stage amp configuration as far as type of processing stages (tube vs. solid-state):
Tube preamp
Tube power amp
Solid-state final amp
This configuration provides the best combination of Tone, cost, weight, and reliability. Where distortion occurs, tubes are used. Where linear amplification is needed, solid-state is used. Because the power tubes are not being used for final, loud amplication, a single small power tube can be used, and this tube can even be a so-called "preamp tube" type such as 12AX7. So the tube complement for a TTS rig could be as little as two small, inexpensive preamp tubes. For further cost reduction, the preamp stage could be a well-voiced solid-state preamp: a great-sounding STS rig could use as little as one so-called "preamp" tube, such as a 6SN7 or 12AX7. It only has one small tube, but that tube is placed in the most key position: in the saturating tube power amp, rather than in the preamp or in the final amp.
Amptone.com ultra gear-search page