Another Buffer Question - Headroom...

[Phorhas]

I Jack's buffer article (in his Lab-Notebook sectiom - http://www.muzique.com/lab/buffers.htm )  he writes:

The value of R2 is too critical and may be any value from 3.3k to 10k without much change in the sound. I prefer to use lower values since this allows more drive on the negative portion of the audio cycle where the only pull-down is the source resistor

In another place (really don't remember where) I read that lowering the source ressitor gives a bigger headroom.

So my questions are:
1. what exactly does jack paragraph mean (sounds)?
2. Is the headroom thingy true?
3. are 1 & 2 also true to MOSFETs and BJT?

Thanx ahead,
Dan.

Oh, and Jack - True a great site you've got there, I love the lab notebook.
:wink:

[R.G.]

1. what exactly does jack paragraph mean (sounds)?

There's a typo there. I believe it should read "the value of R2 is not too critical. For emitter follower buffers, the exact value of the output resistor is not critical. As long as the resistor is low compared to whatever loading is put on it, then the circuit will function well and still fulfill our expectations of a buffer. I think that the comment about sound means that resistors of a few up to 10K will still drive cable capacitances without significant capacitance-caused treble losses, hence little noticeable change in the "sound of the circuit".

2. Is the headroom thingy true?

Not as stated. "Headroom" usually means how much signal above the necessary driving level can the circuit put out. The value of that resistor has nothing to do with how big a signal level a follower can put out. That's determined by the power supply and bias voltages. Lower values of emitter resistor do provide better pull-down on negative going signals, so the circuit can drive lower-impedance loads better, but that's not headroom as the term is usually used.

3. are 1 & 2 also true to MOSFETs and BJT?

Same comments.

A follower with a resistor for a load like shown has an active pull up (the transistor) and a passive pull down (the resistor). So it actually has an asymmetrical output impedance for large signals. The positive going signal is pulled up by the transistor's emitter/source resistance, which is in the units-of-ohms range, but pulled down by the resistor.  Jack refers to this, I believe. Followers with active loading, such as replacing R2 with a constant current source, have extremely low up and down impedances.

Neither of these conditions make much difference if the load driven by the buffer is significantly higher impedance than the emitter resistor. If you have a 10K emitter resistor driving a 100K or larger load, then changing the 10K resistor down to 1K or anything in between, or even going to active loading on the emitter will not change the frequencies transmitted to the load much.