buffering and other stuff question

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
G

Guest

Guest
what purpose do buffers serve in audio circuits usually? are they mostly used to isolate things like led display or other circuitry?

i have the ic opamp cookbook and was going to build the variable dead-band circuit. im building it for school and this book is so dense that i wont be able to get through the stuff i need to so that i can figure this out myself.

i wanted to build this circuit to run audio through it and use it as a distortion generator.

do i need to buffer the input and/or output? what will be necessary to make this circuit ready to operate in a line level configuration?

img003.jpg
 
Buffers usually refer to unity-gain stages with high input impedance and low output impedance. Thus you can take a signal from a source and not load it down by much, and feed it to a load unchanged in voltage, even though the load may be pulling substantial current.

To determine when you need them, look at what the source impedance is and what the load impedance is. Nowadays we generally like to have a source impedance much lower than the load impedance, at least in audio work. This is in contrast to high (~radio etc.) frequencies where, to transfer the maximum amount of energy we strive for what is called a conjugate match---another whole subject.

What is your source impedance for the circuit in question? Is it a lot less than 10k (which is the input impedance of the circuit shown)? If so, you don't need an input buffer.

What is your load presented to the circuit's output? How much current will it pull on peaks of output voltage? As shown the circuit has fairly low output impedance, particularly at low audio frequencies where the 741 has ample open loop gain. But that condition holds only as long as it can supply the needed current to the load and the feedback network, which I wouldn't want to push to more than a few milliamps. If your load is separated by a significant length of cable you will also need to see what the effects of that cable's capacitance are, and may want to place a few hundred ohm resistor in series with the cable feed to avoid circuit oscillations.

In general I don't like to put any more op amps in the signal path than are absolutely necessary, but there are different schools of thought on this.
 
the source for this circuit will most likely be my d/a converters. im not sure what their ouput impedance is, but im sure its about normal for modern line level devices. what is normal? ~1k ohms?

when you say that the circuit has fairly low output impedance, can you give me a rough number?

the output will be feeding back into a/d converters or some other line level device.
 
It depends on wheher the D/A outputs a raw current output or already has converted to a voltage via an op amp (probably the case). In that event 10k is plenty high probably---even if they put 1k in the output you aren't losing much level. OTOH I don't know what Walt's circuit exactly wants to see to get the effect you want. But if the voltage out of the D/A is too low then you would want to put some gain in line. You could also reduce the value of the input R a bit but I wouldn't take it much lower.

The op amp will probably look like a few ohms in the midband. You will likely want to put a few hundred in series so that will dominate the output R. Certainly plenty low to feed anyone's A/D except "naked" ones for video etc.
 
The op amp will probably look like a few ohms in the midband. You will likely want to put a few hundred in series so that will dominate the output R.

a few hundred in series with what?

i think that for now i will just build the circuit as it is shown in the book. it seems like it should at least work, if not work well.
 
it will be quite a short cable run, so i dont expect to run into problems with cable capacitance.

thanks for the help.
 
Back
Top