> Gain, of course, controls your overall output level/voltage.
No!
Don't confuse gain and level. You can have low gain with high levels, or high gain and still not have high levels.
In a mike-amp, your output level is fixed (roughly) by the sensitivity of the next stage (tape-deck input, etc). The gain sets the input level needed to get that desired output level. The whole point of the gain knob is so you get +4dBm whether the gig is ribbons on harpsichord or 414s on Fender amp. A mike amp is semi-fixed output, highly variable input.
> when high gains are required (like in mic preamps where you're doing 50-70dB) slew rates are ABSOLUTELY critical... the higher the amp gain, and the higher the frequency you want to reproduce the faster the slew rate must be.
In audio systems, slew rate is almost always about output stage levels. It does not matter how you got to that level: strong source and low gain, or weak source and large gain. (In a fixed-compensation amplifer, that is.)
Jung's old tests suggest you need at least 1V/μS per Volt of output signal level (i.e. you need to be able to slew in less than a μS) for "unimpaired" audio. Hence a 741 is not a bad amp if your signal levels never exceed a volt. The 5532 and TL071 series amps with >5V/μS have given generally excellent performance with 1.23V nominal (5-10V peak) level structures.
We mostly-only use fixed-compensation unity-gain amps. You can't put a 30pFd cap on a front panel or gang it with a pot. However if a fixed-gain stage will -only- be used at higher gains, you can choose to use a decompensated amp. A 741 is a 301 with a 30pFd cap. If you put a 3pFd cap on a 301, it is unstable for gains of less than 10, but is also 10 times faster, making a 301 a respectable audio amp for slewing signals approaching voltage-clipping, whereas a 741 has to be held to much less than clipping level or it will cream cymbals.
There is one variable-compensation topology. Fashion now calls it "current feedback", a bad name. But long before chip makers discovered it, microphone amp designers were doing it. The classic transformerless mike preamp with a variable resistor between two emitters is "current feedback". Frequency response, stability, and output slew rate change little over a wide range of gain.
I don't understand your obsession with RF filtering at inputs and outputs. If you live under a radio tower, move!. Sometimes you can't, and sometimes you have to filter. Commercial gear usually does because they can't know what their customers have to do. But RF filtering always degrades audio. Maybe insignificantly, but sometimes it bites. Be sure you need it, and be sure filtering is the best solution. FWIW: while I always consider the issue, I rarely do much about it except be sure the gain response is smoothly declining well into the MHz.
> some variable input impedances. I have to admit I really love that part of my VIPRE pre-amp.
Simple variable impedance can't be wonderful. The Vipre must be doing something more. Anyway, if the Vipre is good, why waste time doing something else? It is about the music, not the gear.
No!
Don't confuse gain and level. You can have low gain with high levels, or high gain and still not have high levels.
In a mike-amp, your output level is fixed (roughly) by the sensitivity of the next stage (tape-deck input, etc). The gain sets the input level needed to get that desired output level. The whole point of the gain knob is so you get +4dBm whether the gig is ribbons on harpsichord or 414s on Fender amp. A mike amp is semi-fixed output, highly variable input.
> when high gains are required (like in mic preamps where you're doing 50-70dB) slew rates are ABSOLUTELY critical... the higher the amp gain, and the higher the frequency you want to reproduce the faster the slew rate must be.
In audio systems, slew rate is almost always about output stage levels. It does not matter how you got to that level: strong source and low gain, or weak source and large gain. (In a fixed-compensation amplifer, that is.)
Jung's old tests suggest you need at least 1V/μS per Volt of output signal level (i.e. you need to be able to slew in less than a μS) for "unimpaired" audio. Hence a 741 is not a bad amp if your signal levels never exceed a volt. The 5532 and TL071 series amps with >5V/μS have given generally excellent performance with 1.23V nominal (5-10V peak) level structures.
We mostly-only use fixed-compensation unity-gain amps. You can't put a 30pFd cap on a front panel or gang it with a pot. However if a fixed-gain stage will -only- be used at higher gains, you can choose to use a decompensated amp. A 741 is a 301 with a 30pFd cap. If you put a 3pFd cap on a 301, it is unstable for gains of less than 10, but is also 10 times faster, making a 301 a respectable audio amp for slewing signals approaching voltage-clipping, whereas a 741 has to be held to much less than clipping level or it will cream cymbals.
There is one variable-compensation topology. Fashion now calls it "current feedback", a bad name. But long before chip makers discovered it, microphone amp designers were doing it. The classic transformerless mike preamp with a variable resistor between two emitters is "current feedback". Frequency response, stability, and output slew rate change little over a wide range of gain.
I don't understand your obsession with RF filtering at inputs and outputs. If you live under a radio tower, move!. Sometimes you can't, and sometimes you have to filter. Commercial gear usually does because they can't know what their customers have to do. But RF filtering always degrades audio. Maybe insignificantly, but sometimes it bites. Be sure you need it, and be sure filtering is the best solution. FWIW: while I always consider the issue, I rarely do much about it except be sure the gain response is smoothly declining well into the MHz.
> some variable input impedances. I have to admit I really love that part of my VIPRE pre-amp.
Simple variable impedance can't be wonderful. The Vipre must be doing something more. Anyway, if the Vipre is good, why waste time doing something else? It is about the music, not the gear.
