My Gainclone is too Loud

[PaulHilgeman]

Hey, I made an inverted gainclone with this schematic:

An externally hosted image should be here but it was not working when we last tested it.

It is WAY Too loud. I am using speakers that are about 89dB sensative. I have a stepped attenuator on the input, all of the way in one direction, there is 107kOhms of resistance, that is the quietest it gets, and it is too loud to comforitably talk. What are my best options for lowering the volume output?

-Paul Hilgeman

 

[Steve Eddy]

It is WAY Too loud. I am using speakers that are about 89dB sensative. I have a stepped attenuator on the input, all of the way in one direction, there is 107kOhms of resistance, that is the quietest it gets, and it is too loud to comforitably talk. What are my best options for lowering the volume output?

Well, a simple first step would be to replace the 220k resistors with 110k (or just wire another 220k in parallel). That'll drop the voltage gain of the amplifier down by about 6dB.

se

 

[Peter Daniel]

If you are using preamp, the first step would be getting rid of it.

With just the source connect it straight to the amp, it can't be too loud, unless you wired your stepped attenuator incorrectly.

 

[Pedja]

Hi Paul,

If your stepped attenuator is linear (and not logarithmic), it explains the situation, i.e. my experience is the same. Try logarithmic attenuator, or at least decrease values of the few resistors closest to ground. Or try linear pot, you’ll get somewhat rougher control at the lowest levels than at the rest of the motion, but it will be better than it is now. Or use some good (good old) logarithmic “audiophile” pot.

Lowering Rfb (and overall gain) is not recommended here, because with 100k volume control it will give overall gain (worst case) of 110/(10+50/2)=3.14. National in LM3875’s datasheet claims minimum stable gain of 10.

Pedja

 

[EchoWars]

If your stepped attenuator is linear (and not logarithmic), it explains the situation, i.e. my experience is the same. Try logarithmic attenuator, or at least decrease values of the few resistors closest to ground. Or try linear pot, you’ll get somewhat rougher control at the lowest levels than at the rest of the motion, but it will be better than it is now. Or use some good (good old) logarithmic “audiophile” pot.

If you have a linear pot and need a log pot, read this. Might save a buck or two.

 

[Pedja]

If you have a linear pot and need a log pot, read this. Might save a buck or two.

With the amp in inverted mode linear pot behaves as logarithmic… but not at the lowest levels, as you can see in the article you suggested (figure 2). Hence the problem with linear stepped attenuator.

Pedja