I had reminded us in yesterday's post the three elements of a preamplifier:
- Input selector
- Volume/balance control
- Gain stage
We know volume controls are intrusive to sound quality. Some, more than others.
I had come up with the idea of eliminating the volume control by designing a gain-variable amplifier stage. A great idea with only one problem. How to do it?
The gain of an amplifier is typically set by a pair of resistors. These can control a single stage or an entire gain block.
Here's a typical application:
The gain of this simple amplifier is set by the two resistors I spoke of, Rg and Rf. Divide one resistor by the other to arrive at the gain. So, imagine Rf is a 10kΩ resistor and the other is 1kΩ. You get a gain of 10 (20dB).
So if that's true, would it not be easy enough to replace one of the resistors with a pot? Then, turning the pot would vary the gain of the amplifier. Voila!
Unfortunately, that doesn't work for a number of reasons—some very technical but let's see what we can explain simply.
The first problem is the idea of adding a pot in this amplifier. Turns out it doesn't matter where you place the pot, it will always add its sonic signature once in the circuit. So we've not managed to make an improvement if we implement it.
The second problem is even worse. This type of circuit uses feedback and it is the feedback ration the two resistors are changing. When you change negative feedback the sound of the amp changes too (as well as stability). Each different level would sound slightly different: lower levels would have greater amounts of negative feedback, louder would have less.
The third problem is perhaps the biggest. You can't easily have much less than a gain of one. Which means we can't have a volume control that goes from zero output to full output. At least not easily, practically in any useful way. Just ain't gonna work.
There has to be a better way.
And, there is. One we use today in the BHK preamplifier.
