Preamp volume control - why on the input side of the circuit

[rif]

 Almost every schematic I've seen ( which admittedly isn't that many)  for preamps, takes the input signal,  attenuates  it, then amplifies it,  then outputs.  Why throw away signal then amplify?  Wouldn't you want as high a signal to go through first?   And attenuating introduces some noise that than goes on to get amplified too.

(I was just working on my quickie and couldn't get this out of my head)

I'm going to take a SWAG and say impedance.

 thanks
-david

[Paul Birkeland]

Hello David,

This is a great question, I'm glad you asked!

Let's make some assumptions about an imaginary system.  Let's say you have a DAC with a 5V output and a power amp with 10K input impedance.

Now, let's say you have a tube preamp with 1.5V of bias, high input impedance (100K) and low-ish output impedance (1000 Ohms)

If we move the volume control to the output, here's what happens to the preamp.

1.  The 1.5V of bias means that the 5V signal from the DAC will outright clip the input, resulting in a lot of distortion regardless of how the pot is set at the output.

2.  At the output, we must have the pot.  If we want to drive a fair amount of cable and the 10K load of your power amp, that pot shouldn't be any larger in value than 1K.  If we want to keep reasonable gain/distortion numbers in terms of tube operation, we don't want any more than a 10K load at the output of the preamp, so certainly not a 1K pot! (The 1K pot will burn off half the available output signal, decrease the actual gain of the circuit, and increase the distortion generated during operation)

3.  Say we spit the difference in #2 and go with a 5K pot, then we must have a coupling cap appropriately sized for this.  This is a mandatory 5uF, and again we are not prevented from getting a nicer quality 1uF cap to feed our Paramount, Stereomour, or Sex.

4.  The output impedance of this preamp will be variable.  With the volume nearly all the way down, it may be well under 500 Ohms, but once you crank it up, it will approach 5K.  This will lead to variable performance depending on the position of the volume control

Despite all of this, you'll still see this done, and it can work if you're mindful of the gear in your system.

-PB

[Paul Joppa]

PB has covered the downside well. The upside is that a higher signal level inside the preamp (assuming it does not induce clipping) gives it a better signal-to-noise ratio when the following device (power amp) is very sensitive.

At one time, long ago, there were general-purpose amps built for studio applications that had two volume controls ganged together - one at the input and another at an intermediate point. Thus there was no control at the output (to maintain the low output impedance, yet the input was not so strongly attenuated. If the old Foreplay were not direct coupled, this would be a good way to control the volume. Of course, that would add a capacitor in the signal path.

[rif]

Thanks guys,  that makes lots of sense.   As an aside,  thanks in a broader sense for all your posts - a  short while ago none of this would have made sense to me.  I learn a lot from you ... and I'm getting there.

I actually own a preamp (apt  holman) that controls  volume by varying  the circuits gain.  I know that plays with the circuits intrinsic noise,  but I guess there are always tradeoffs.

- david

 edit

[Paul Joppa]

Thanks for the reminder, I had forgotten the apt holman. A true classic and a truly stunning piece of engineering!

[adamct]

The AMB M^3 can be built to do the same thing. In fact, I'm building one now...