Alternative output level attenuation via gain?
Mar 23, 2006 at 9:49 PM Thread Starter Post #1 of 6

firefox360

500+ Head-Fier
Joined
May 23, 2005
Posts
765
Likes
10
I just pretty much got this idea from the PS Audio GCHA Gaincell technology, and was wondering about. Is it possible, for example in a CMOY circuit, to replace the gain resistors with a volume pot, and use that system to adjust the output volume rather than placing the pot in the input signal path?

Seems like it would be simple enough in my head. A volume pot is sort of like a variable resistor from what I know. Would this design yield better audio quality (ofcourse it would be using a nice good pot)?
 
Mar 23, 2006 at 11:25 PM Post #2 of 6

DaKi][er

500+ Head-Fier
Joined
Jul 28, 2003
Posts
845
Likes
11
In the cmoy's non-inverting configuration you cant get a gain less than 1, which will be the equivalent of plugging your cans into your source at full volume
Also comes issues of stability with some opamps with changing feedback resistors

You don’t get it out of the signal path, the feedback loop is very much still an important part of it, and in the end an overall very bad idea
 
Mar 24, 2006 at 2:13 AM Post #4 of 6

rickcr42

Are YOU talkin' to me?
Joined
Jun 21, 2001
Posts
13,874
Likes
14
Can be done I think but not as simply as replacing Rf with a pot.


I have the schematic written down somewhere in my chamber of horrors but from memory it is done with three resistors and a pot instead of resistor/pot.

Rg and Rf hooked are up normally but with the side of Rf that normally goes to the opamp output now going to the wiper of a pot so one end of this now becoming the new "Rf" connection to the output and the other end attached to a new resistor that has the end opposite the pot going to circuit ground.

Two resistor ratio profiles instead of one in effect

From memory the ratio of this resistor is 1:10 of the value of the pot and the pot 1:1 the value of Rf with Rg being the same as rf making the NEW resistor the actual gain set resistor (x10 or +20dB).

because messing with the feedback loop can make it unstable for DC it is a good idea to both AC couple the output and to de-couple the loop at Rg with a 100-220 uf electrolytic to gnd (between rg and gnd)

At least that is how I remember it
tongue.gif
 
Mar 24, 2006 at 12:33 PM Post #5 of 6

kin0kin

Headphoneus Supremus
Joined
Jan 4, 2005
Posts
2,804
Likes
12
if you change the gain/resistors, the opamp characteristic will also change. designer will now have do deal with a solution that'd solve problems in multiple models instead of just one. the signal response will also change if the resistor values are not properly selected (overshoot, skewed..etc), and rickcr42 also mentioned problems with dc offset. and with that implementation, we will have to deal with ppl talking about which gain sounds best
biggrin.gif
 
Mar 24, 2006 at 9:08 PM Post #6 of 6

firefox360

500+ Head-Fier
Joined
May 23, 2005
Posts
765
Likes
10
Hrm... seems like it wouldn't be a very viable design for a non-inverting amplifier design. Even on an inverting amplifier, it doesn't seem to be worth the work required. Guess I'll just be sticking to the standard ole' Pot-in-the-input-signal method.
 

Users who are viewing this thread

Top