Op-Amp Biasing

[GGBB]

In the two op-amp buffers below from the AMZ buffers page, is there an advantage to biasing one way or the other (other than parts count)?

[R.G.]

The first way is better, for noise reasons, because of the parts that are not shown.

There is presumably a big, fat capacitor to ground over wherever that "Vr" is generated. That shunts noise from the divider string to ground. It's a bit quieter for conditions where the bias resistor equals the parallel resistance of the separate string.

Moreover, you typically have more than one opamp to bias, so you generate the voltage once and use it many times.

[GGBB]

Thanks R.G.  I would be running two op-amps, but both would share the same input.  Does that nullify the noise benefits of the proper Vr setup?  Something like this but only two-way and there would be separate adjustable gains via the feedback loop on each: http://www.guitarpcb.com/apps/photos/photo?photoid=130355998.  I would also be implementing power supply filtering.

EDIT: Actually, now that I've thought about it and drawn the schematic in my head, with two op-amps sharing the same input, the two R1s from Vr would be in parallel.  So I guess there's no point in doing it that way with a shared input.

[GibsonGM]

How about the AMZ Splitter?  This thing works awesome!  And low parts count, very expandable, and so on.  If you need different output levels, you can add a volume pot to the outputs.   I only introduce this as some feel that opamps are a little 'sterile sounding' than JFETs.   I've built this one, and really like it.    

http://www.muzique.com/lab/splitter.htm

[GGBB]

How about the AMZ Splitter?  This thing works awesome!  And low parts count, very expandable, and so on.  If you need different output levels, you can add a volume pot to the outputs.   I only introduce this as some feel that opamps are a little 'sterile sounding' than JFETs.   I've built this one, and really like it.    

http://www.muzique.com/lab/splitter.htm

Thanks, but I'm not actually building a splitter - just modding an existing circuit that uses two op-amps.

[R.G.]

How about the AMZ Splitter?  This thing works awesome!  And low parts count, very expandable, and so on.  If you need different output levels, you can add a volume pot to the outputs.   I only introduce this as some feel that opamps are a little 'sterile sounding' than JFETs.   I've built this one, and really like it.    

Using a single JFET, or for that matter a tube, introduces a bit of asymmetrical distortion. It's pleasant-sounding, but not accurate. As an analogy to beverages, some people like a little sugar in their tea (... one lump, or two?) or coffee. Some people like a lot. Some people think that wine that's not "dry" enough is a barbarism. But very few people like to sip pure syrup.

The supposed "sterility" of opamps is usually a complaint that they do not color the sound pleasantly like non-feedback stages do.  If the opamp is not driven to clipping, then it's usually very accurate, and eliminates treble loss, so the sound is *very* accurate. If what you wanted was a little more sugar added to your sound, then by comparison it's "sterile". Or "clean". Or whatever your adjective of the day is.

It's also worth noting that a JFET has three terminals. So does an opamp, leaving aside the matter of the power supply pins for the moment - we'll come back to that. If you count resistors and caps in the splitter circuit, you get one *more* resistor than an opamp follower would need - the source resistor. So a single opamp follower has the same number of pins used as a JFET, and successive parallel followers save you a resistor per section. That makes up for the power supply pins in some ways.

If you like sweetening, one JFET makes it sweeter, and using opamps to preserve the sound keeps it that way in the face of cable runs.

If you're wondering what I have against JFETs, nothing. Except that they have *zero* power supply rejection, so any power or ground noise comes right through. Opamps in general have significant power supply noise rejection, albeit some more than others.

And for splitters: the term "splitter" is a bit misused. An audio splitter only needs to keep one output from loading another down. This can be done with resistors as well. Real splitting for hum-reduction purposes need more complex circuits to reject the hum that can be induced from driving different load grounds.

Sorry - I read this as I had my first cup of coffee. 

[Digital Larry]

The supposed "sterility" of opamps is usually a complaint that they do not color the sound pleasantly like non-feedback stages do.  If the opamp is not driven to clipping, then it's usually very accurate, and eliminates treble loss, so the sound is *very* accurate. If what you wanted was a little more sugar added to your sound, then by comparison it's "sterile". Or "clean". Or whatever your adjective of the day is.

About 20 years ago I worked very briefly for Velodyne Acoustics, who make servo-controlled subwoofers for home theaters.  Their approach drastically reduces distortion in the low-mid frequency range and gives a more accurate reproduction of the recorded material.  They told me "a lot of people don't like this sound".  They were in the process of developing some lower end (price wise) products at that time which were NOT servo controlled and were probably more successful as a result (of both factors).

[amptramp]

Using a single JFET, or for that matter a tube, introduces a bit of asymmetrical distortion. It's pleasant-sounding, but not accurate. As an analogy to beverages, some people like a little sugar in their tea (... one lump, or two?) or coffee. Some people like a lot. Some people think that wine that's not "dry" enough is a barbarism. But very few people like to sip pure syrup.

We had one member at the London Vintage Radio Club give a talk on his SET (single-ended triode) amplifiers at one meeting.  This was a stereo amplifier with power supply and the huge transformers you need for single-ended output.  He played some music through it that I was familiar with and it sounded pleasant, but you could hear the difference from the original sound as produced by a more "accurate" amplifier.  One of the methods of linearizing the output is to drive a closed speaker enclosure is to match the high side of the output with the compression side of the speaker so as the excess output drives the speaker, it pulls back from the listener, compressing the contents of the enclosure more.

[R.G.]

In a way, single ended stages with a reputation for "tons of tone" are the musical analog to pancake makeup. It may look fine if you don't look too closely, and might or might not be covering something best unseen.

[PRR]

> They told me "a lot of people don't like this sound". 

We all grew up over-driving small cheap speakers. It isn't loud without 3%-15% IM distortion.

Hmmm... their approach doesn't change the intrinsic IM caused by two tones from a large-excursion radiator, only the driver electromechanical (and air-suspension) nonlinarity. Maybe acoustic IM without driver IM is un-natural? Or just not what we grew up with?