Preamp design - gain vs volume

[marcelomd]

Hi,

I was looking into preamp designs (the ones that go with the power amps, not pedals) and noticed they tend to fall into two camps regarding gain control:

- Active device set for a large gain followed by an attenuator. All tube amps I know are like this. To my surprise, lots of opamp circuits too;
- Active device with variable gain (feedback loop, or source resistor).

What are the pros and cons of each approach?

[MrStab]

Hi Marcelo,

In terms of op-amps, at least:

Gain > Attenuate will limit your available headroom at all times, whereas true Variable Gain will let you adjust as needed.

But! Variable gain often requires interrupting a feedback loop with a pot, which has a small risk of intermittency (especially with age) and may force an op-amp to go full open-loop gain.

Why not both? (fig 3)
https://sound-au.com/project01.htm

[marcelomd]

Hi,
I actually got thinking about this after seeing ESP's bass guitar project ((https://sound-au.com/project152-1.htm)):

Gain, attenuate, gain attenuate... Interesting he always uses even numbers of opamps per section. Maybe to fully utilize ICs?

Anyway, my first thought was "why waste energy amplifying a signal just to knock it down?". Something with noise that I don't understand? Old practices inherited from tubes or transistors?

[MrStab]

I'd never thought of the even-numbered op-amps as deliberate, but it could be! ESP is great.

I think the convention of amplify > attenuate is partly related to human expectation. On the surface, it goes all the way up to +10dB (or whatever), unity, then all the way down to -infinity. That's really handy and convenient for the user, even if it's not electronically ideal. It's a bit like a boost/cut pot on an EQ.

[amptramp]

One bit of wisdom I learned from playing with a synthesizer is never allow a control position to let the gain go to zero or you will have a dead effects chain that you will not be able to diagnose with the kind of speed you need when playing live.

The use of two gain pots means that if the controls were linear, the halfway position on the pot results in one-quarter of the maximum output.  This may approach an audio taper with enough pot sections.

Varying a stage gain may affect bandwidth and stability whereas a pot at the output doesn't, but it does vary the output impedance.  But this can be kept low enough that it doesn't change much of the effect on the following items in the effects chain.

[marcelomd]

OK,

How about this approach? I've seen it a few times, but not at the front end.

[r080]

I was trying to decide between an MXR microamp circuit vs a gain then attenuation booster for a pedal and noticed there are some tradeoffs. With the microamp, it looks like you have a constant output impedance, but your bandwidth changes slightly with gain. With my other design, the output impedance was acceptable for the whole range, but it did vary. The bandwidth stayed the same.

It might take analysis of different options to see all the tradeoffs, and decide what works best in your situation.