choosing a rotary Attenuator for RCA pre

[NewYorkDave]

Rafael, here's a handy derivation of the capacitive reactance formula that can help to answer the question of "how big should the cap be?"

f = 159155 / C *R

where f=low frequency rolloff point in Hz, C=capacitance in microfarads and R=resistance in ohms.

In the case of a 0.22uF cap and a 470k resistor, f = 1.5Hz
So 0.22uF is plenty big enough!

 

[cuelist]

[quote author="NewYorkDave"]When you put a volume pot in between two stages that are enclosed within a feedback loop, you're varying A between zero and the full open-loop value, rendering the feedback useless at best.[/quote]

Yes and that was the point I was trying to make, though I certainly failed at that. Thanks for putting it (and me) straight).

[quote author="NewTorkDave"]Ignoring the possible instability at high frequencies caused by phase shift from inserting series resistance into the signal path, it has been demonstrated that small values of feedback can cause generation of annoying higher-order distortion products--so the sound of the amp, as well as its gain, will change significantly as the value of A changes with B remaining fixed.[/quote]

I think its generally agreed that a moderate amount of local feedback is a good thing while a large amount of global feedback (encompassing several gain stages) is not such a great idea.

 

[Guest]

[quote author="NewYorkDave"][quote author="cuelist"]Sorry, but that's incorrect.

G = A / 1 + AB (in the case of a noninverting amplifier)
where G=closed-loop gain, A=open-loop gain and B=feedback as a [/quote]
Sorry, parsing error.
Bode equation is this

G=A/(1+AB)

xvlk

 

[NewYorkDave]

parsing error.
Bode equation is this

G=A/(1+AB)

Correct. I should have included the parantheses.

think its generally agreed that a moderate amount of local feedback is a good thing while a large amount of global feedback (encompassing several gain stages) is not such a great idea.

"Generally agreed?" I don't know if anyone's ever taken a poll to confirm that :wink:
Local feedback is easier to implement than global, but that doesn't necessarily make it "better" in the absolute sense. I don't have any religiously-held convictions on this issue, myself. I will say that a fairly large amount of global feedback seems better when the gain of a circuit must be held to a particular, predictable value--for instance, in instrumentation, or when you want the front-panel gain indications on your mic preamp to always correspond directly to reality, instead of just being an approximation.

You may have missed it, but I posted a link to a paper by P.J. Baxandall that offers food for thought on the issue of "how much" negative feedback is desirable:
700kB zipped PDF file.

 

[Guest]

[quote author="NewYorkDave"]

parsing error.
"how much" negative feedback is desirable:
700kB zipped PDF file.

How much" negative feedback is desirable? in tranfsormer input
mic preamp. Clasical (output) paralel (input) series feedfback
serves here as input transformer with turns ratio reversed to mic
tranhsformer.
You can ommitt input transformer and negative feedback and noise
figure is still the same ?
Serves transformer hera as distortionless prepreamp ?
xvlk