Usefulness of buffer

Frequent asked about building blocks: gain stages, buffers, clipping configurations, ...
Post Reply
User avatar
shicky
Information
Posts: 12
Joined: 16 Mar 2008, 16:11
Location: Toronto

Post by shicky »

I have a question.
I realized some pedals don't have buffers.
One of the biggest purpose of buffer is to provide large input impedence and isolate the circuit inside.
However wouldn't taking buffer out and putting a large resistor achieve same result?
I heard that it has something to do with noise and curret gain.
Wouldn't taking out buffer results in more "transparent" sound?
Well.. I don't think a buffer would make a noticeable difference.. but still..

User avatar
R.G.
Resistor Ronker
Information
Posts: 337
Joined: 22 Sep 2007, 02:24
Been thanked: 39 times

Post by R.G. »

shicky wrote:One of the biggest purpose of buffer is to provide large input impedence and isolate the circuit inside.
However wouldn't taking buffer out and putting a large resistor achieve same result?
How would a large resistor do the same thing? If you mean a large parallel resistor, then the resistor might as well not be there, the circuit will still suck treble if it is so inclined.

If you mean a large *series* resistor, it does indeed give large input impedance and isolation of the circuit inside. But it also cuts the signal down hugely, in the ratio of the resistor and the circuit it's feeding. The bigger the series resistor, the smaller the signal that gets through.

And, by the way, the more thermal noise the series resistor generates.
shicky wrote:I heard that it has something to do with noise and curret gain.
I get a little nuts when people hear things on the internet, like the factoids drifted in on the wind like pollen. :lol:

It's got a lot to do with noise. And a resistor has no current gain. Buffers do. So a buffer can take the input voltage, multiply its current, and not load down the signal source.
shicky wrote: Wouldn't taking out buffer results in more "transparent" sound?
How, exactly?

And while we're at it, can you define exactly what "transparent" means?
shicky wrote: Well.. I don't think a buffer would make a noticeable difference.. but still..
Well designed buffers are inaudible in terms of noise and distortion. They are VERY audible in terms of keeping treble from being sucked out by loading.

User avatar
shicky
Information
Posts: 12
Joined: 16 Mar 2008, 16:11
Location: Toronto

Post by shicky »

How would a large resistor do the same thing? If you mean a large parallel resistor, then the resistor might as well not be there, the circuit will still suck treble if it is so inclined.
Sorry for being unclear. For an opamp based overdrive with soft clipping, you can put large parallel resistor before clipping stage. Timmy circuit for example. But now I think of it, it would affect the gain of the clipping stage... Sorry for asking a stupid question.
It's got a lot to do with noise. And a resistor has no current gain. Buffers do. So a buffer can take the input voltage, multiply its current, and not load down the signal source.
Well, of course resistor can't have current gain.
I am an electrical engineer student :roll: but a noob :D
And while we're at it, can you define exactly what "transparent" means?
Ya, I know that its kind of vague term.. or wrong term.
I used it to describe something that is not muddy. (losing treble perhaps)
In fact, some electronic components for audio applications are advertised or known as "transparent".
Well designed buffers are inaudible in terms of noise and distortion. They are VERY audible in terms of keeping treble from being sucked out by loading.
When I was playing with circuits, I noticed that buffer does matter.
I just wanted to know some disadvantages of having a buffer.
In fact, transistor buffers output is only about 30% of an input signal and opamp buffer output is about 96% of the input.
( depending on the type of transistors and opamps )
By having 10 or so buffers on the signal path, it might alter the signal.

Well, I am not really a tone-freak :) so I wouldn't know for sure.
Thank you for clearing things out by the way.

User avatar
Torchy
Information

Post by Torchy »

you can put large parallel resistor before clipping stage. Timmy circuit for example.
Where is this resistor in the Timmy schem ? I see only a 510k to Vref biassing resistor.

User avatar
R.G.
Resistor Ronker
Information
Posts: 337
Joined: 22 Sep 2007, 02:24
Been thanked: 39 times

Post by R.G. »

shicky wrote:Well, of course resistor can't have current gain.
I am an electrical engineer student :roll: but a noob :D
S'OK. We were all there.
Ya, I know that its kind of vague term.. or wrong term.
I used it to describe something that is not muddy. (losing treble perhaps)
In fact, some electronic components for audio applications are advertised or known as "transparent".
OK, I'll accept that. That particular term sets me off. People who wouldn't know a phone jack from a pillbox get really picky about how transparent one heavy metal distortion compared to another.

You pass.
I just wanted to know some disadvantages of having a buffer.
(a) Cost (b) smallish added noise (c )removing artifacts you might have liked, like a buffer before a fuzz face is not always good because it's a sound you didn't expect.
In fact, transistor buffers output is only about 30% of an input signal and opamp buffer output is about 96% of the input.
You do need to go look that one up again. Transistor emitter followers can have voltage gains of 0.96 to 0.99 depending. Opamp buffer gain depends on frequency, but at low audio, gains of 0.999... for followers are common. I'm particularly worried about who told you transistor followers only "followed" to 30%. Whomever it was either had a set of special conditions they didn't tell you, or did not themselves know. Even the perennial follower loser, the triode, is about 85-90% depending on tube.
By having 10 or so buffers on the signal path, it might alter the signal.
That is a real, valid concern. Noise build up is the major problem with opamp buffers. If you're using JFET source followers or tube cathode followers, you may get a build up of lowered gain, as 0.9*0.9*0.9*0.9... gets small after a while.

What what your profs tell you. Some of them really do know. The others? Sigh...

Post Reply