[kelly]

Please excuse my very naive newbie question.

Instead of putting a pot or stepped attenuator in the signal path, why couldn't a multi-step gain switch be made so that you simply have a different gain for each desired output volume?

 

[rickcr42]

it would work fine but only if........

A - it has from 24 to 30 positions for the steps.Otherwise you might find the adjustment rnge is too course if using headphones with different sensitvities . If you will always use the same headphones the curve could be set to match your personal listening levels

B - you do not have phase or rf compensation caps in the feedback loop (I do) . In that case the change of resistance would change the frequency of the pole kinda like a tone control and really screw with stability .

There is a thing called resistor noise and the larger the resistor the higher the noise level . So a 100K pot will always be noisier than a 10K pot . Problems arise when trying to drive a 10K pot from a source with a highish impedance so 100K is the most common value seen if at the input of an audio device . If tubed then 500K is seen a lot .

A passive attenuator on the other hand is always a comprimise of the preceeding stage "seeing" a high enough impedance as to not be loaded down or have a high frequency attenuation , but low enough that it is a factor of ten lower than the input impedance of the next stage-usually an amp .
So 10K is the norm .

Rick

 

[Ob3ron]

Quote:

but low enough that it is a factor of ten lower than the input impedance of the next stage

Also pardon my newbishness... wouldn't a factor of 10 mean that, by voltage divide, the load gets only 10/(10+1) * 100% = 90.9% of the source voltage? That seems low to me, but for all I know, it may very well be within tolerable limits - I am afterall only new to electronics

 

[rickcr42]

not a precise measuremant nor even a standard but a genaral rule of thimb for designing stages .
And the lower the input impedence of a stage the more drive current you will require to drive the combination of cable and input . That is where a lot of equipment falls down , drive current.

Just like an power amp stage can clip when it is stressed , so too can a line level stage .

for more information see the DIY links resource above , go to the Rane library and read up on equipment interfacing and drive requirements

 

[CaptBubba]

Also running the wires to the control would be much longer than normal. I would prefer the feedback loop to be as small as possible, so you don't pick up any RF noise. You are also limited to a minimum gain of unity, which could be too high.

 

[morsel]

Just an FYI since line level audio signal standard impedance is 47k a 50k pot is theoretically ideal if you can get them.