volume controll in the feedback loop

[sss]

hello guys

lets say i got a simple NI amp , the gain is set by 2 resistors :

Av=1+R2/R1

what if i will controll the volume of an amp by changing the gain?

whats wrong with that?

i think that way the SNR will be better but there must be something wrong here because nobody is doing that

[dkemppai]

Hi,

It's usually done for stability reasons. An amplifier may not be stable over the entire range of gains you wish to run it at.

Increasing the gain (or keeping it higher), can allow for a reduction in compensation feedback required to keep the amplifier stable. (Note, some amplifiers are stable only if used above some minimum gain value, say |g|=5, or |g|=10...)

-Dan

[Tony]

i have tried this, but you can not zero out the signal, you still need a voltage devider refferenced to ground.....

[sss]

Originally posted by dkemppai
It's usually done for stability reasons. An amplifier may not be stable over the entire range of gains you wish to run it at.

what if i'll do i in a pre amp , i think most op amps will be stable from unity gain to ~ 27db

[markp]

Originally posted by sss


what if i'll do i in a pre amp , i think most op amps will be stable from unity gain to ~ 27db

Not all opamps are stable at gain=1 so check the data sheets. Also, what about turning the volume all the way off where gain is less than 1? I would not advise you to use this method, do it the tried and true way.

[anatech]

Fosgate amps used to do it this way. The control was in the feedback loop of an op amp. Minimum gain was 1. If the control opens, there is no feedback, that includes dirty controls to max gain. Very ugly.
I wouldn't recommend this either.

[rayban68]

Use a pot that has a switch built in, so that when you turn it all the way down it clicks off. For inverting op-amp, put the pot in series with a resistor that sets your gain at about 1 when the pot is all the way off. This way, you turn the pot to the point where it is almost ready to click off, and the amp is at unity gain and still stable (and also very quiet) turn the pot a little more and the volume cuts off. For an non-inverting amp I guess you don't need the extra series resistor.

[sam9]

Even if stable, distortion varies with gain. I suspect the origin of this idea came from the use of the feedback loop foor tone controls. That may make some sense there as the gain is frequency dependant and probably does not change over such a wide range as would be needed for a volume control.

If you implement the idea with an opamp in an inverting topology, then I think negative gain (attenuation) is possible, but you still run up against the two concerns stability over the entire range and distortion increasing with gain.

A pot at the input is so much more simple.

[sss]

lets say u got a power amp with a voltage gain of 30

to controll the volume most of the time there is a potentiometer at the input that will change the overall gain of the amp to less then 30 .
what i'm sayin is why not connecting a potentiometer insted of a feedback resistor , no matter if its inverting/non inverting amp.
that way the signal to noise ratio will be better i guess
and at lower volume level there will be more feedback so ->less distortion

[anatech]

But, real life problems may wish you had not done it this way later. Depending on the pot, you may inject noise into the circuit off the wiper of the pot. AC hum would be the most common.
What happens if the wiper gets grounded to the chassis? Conductive cleaner or insulator leakage could give you a headache here. DC instability may result if the pot is rotated quickly also.
Build it & see. If it doesn't work well, substitute a fixed resistor and go back to a control on the input. Nothing is lost then.
-Chris