I find that data is very useful in making decisions. How startling a statement to make on a ham radio blog. For instance, when I post subjects about how to build some small building blocks for a radio project that gets more views than something like fabricating a complete rig or a linear amplifier like yesterday. So back to the building blocks.
But 1st the elephant in the room. I am less of a purist than other homebrew hams. I tend to look at the utility provided by circuits and not so much as the pedigree of the topology. I am OK if I cannot physically see every component in a build (discrete parts versus IC's). My rational is performance, cost and time to build.
Many an argument has been made about building audio amplifiers using totally discrete components versus adding in IC's. I usually will test drive a particular topology and then decide what fits best with the desired result and my time available.
Yes, I have built discrete audio amplifiers, and a good example was The Let Build Something Project I was able to improve this design by adding in place of the 2N3904 and 2N3906 the TIP31C and TIP32C devices for more output with no other circuit changes.
Thus, today an audio amplifier stage using not one, but two IC's is presented for your dining and dancing pleasure. The pre-amp stage uses the NE5534 which is a low noise op-amp, and the output IC is the 14Pin LM-380. This can be built using Manhattan construction and I would suggest the use of W1REX MePads as that will put you on a better path to success. For me I have this design in my CNC computer, so I just punch the start button.
You will see this amp in many of my projects including the KWM-4, The Sudden Transceiver and The Simpleceiver.
So hopefully more views than a tube linear amplifier using high voltage.
BTW the Lets Build Something link (earlier in this post) if you look an the 2nd Generation LBS link there is another audio amp with a 2N3904 driving the LM386-3 IC
73's
Pete N6QW