>
dearth of transformerless discrete mic preamps
Hey, you didn't say it had to be balanced, any overload spec....
"Discrete" versus "IC"....
Of course nearly all current individual transistors are made on IC fabs. The electrons (and holes) don't know the difference.
OTOH, when you have to make all the transistors in one batch, you can't get very-different transistors for different functions.
To my mind the difference is "commercial". No audio product (until recently) sold in enough quantity to justify IC design. All our low-noise amps have to sell into instrumentation and other markets to get the sales volume needed to justify mask-making. And even inside Audio, not everybody agrees what is best. High input bias current is often part of low voltage noise. But high bias current upsets designers who hope to use high-value input resistors. And frankly there are more mediocre boxes (and chips) sold than really low-noise boxes. Also, once a suitable design begins to gel, the IC production engineers jump all over it. Once in production, profit is about how many chips fit on a wafer. They reduce feature size. In audio, in some places, this is good. In many places, a smaller transistor is bad for sound even if it meets spec. Also ICs have modest power handling ability so there is a strong incentive to run them Class AB. But it isn't all that easy to fix the Class AB idle current. If it is too high, it runs hot and also can't easily be told (on test) from a chip that is just leaky all over. IC production engineers like low power currents so leakers can easily be kicked-out. But if Class AB idle current is low, even a little low, it looks good on the scope but grates the ear.
If you slice all the possible planar transistors off the wafer and box them individually, a clever designer can mix-n-match and test and trim for the results s/he wants in a specific product, instead of some canned solution that maximizes total sales in many markets (industrial, military, instrumentation, telecom, etc).
But when you go far beyond my 1-transistor preamp into differential balanced controlled-gain low-THD, and want it to be stable against temperature and supply and parts variations, it gets to be a tough chore. Especially at the very high gains we like in dynamic-mike preamps. Using 2 or 3 high-gain low-offset push-pull-output chips really makes things easier, and unless you have brilliance or infinite time may give a perfectly acceptable (and salable) box better than you could design from pieces.
I happen to have an all-discrete balanced mike preamp design simmering. I may present it here in coming days. But it has "serious" flaws. It is a bit sloppy at gain=+60dB, it has a very odd unbalance quirk at low gain, and it runs hot as heck (like 6 Watts). I think it is useful in enough common situations that some folks here would find it useful, especially since it is brutally simple and could be visually elegant (pairs of fat heatsinks and resistors). But I would not present it as an all-purpose mike amp for all users. To do that would require many more parts and more thinking than I care to do for fun.
