[sdiy] Discrete Op Amps

[cheater00 cheater00]

On Sun, 9 Apr 2017 14:08 Mattias Rickardsson, <mr at analogue.org> wrote:

> On 9 April 2017 at 12:45, Gordonjcp <gordonjcp at gjcp.net> wrote:
> > That doesn't really happen though.  All opamps sound pretty much the
> same.  All distortion pedals sound pretty much the same too, and it's
> really only the frequency response - which is not determined by the
> distortion "shape" - that gives it a particular tone.
>
> Quite an oversimplification. Different distortion pedals typically
> have a pre-equalisation that differs, followed by a nonlinearity of
> many kinds, followed by a post-equalization that differs.
>
> The pre/post equalizations not only shape the "tone" but affect the
> waveshape and hence the result of the chosen nonlinearity, and also
> determine parts of the dynamic behaviour.
>
> The nonlinearity (distortion shape) can vary - I can think of at least
> 5 radically different distortion constructions used in commercial
> distortion pedals, and they do give different tones and frequency
> responses.
>
> Common for successful distortion pedals is that a good combination of
> these parts are chosen, since all of them are affecting the end
> result.
>
> > Building discrete opamps to chase a particular distortion shape is the
> sort of thinking that leads to folk buying specially-imported
> hundred-dollar rocks to damp out vibrations in their mains cables.
>
> True, locally produced hundred-dollar rocks do the trick just as well. ;-P
>
> When it comes to discrete op-amps and their usefulness today, I do
> think that they can be seen primarily as a distortion device - not as
> the hi-fi component solutions they once used to be. Using them would
> require that you know what particular distortion they give you, and
> why you want it. And they could very well be a valid choice, of
> course. But today there is no big need of replacing op-amps per se, so
> introducing another known type of distortion in the signal path might
> be a much more reasonable way than staring at clumsy discrete op-amps.
> :-)
>

Here's the thing though: your op amp is *always* a distortion device. Your
choice is whether the dostortion is nice or it isn't, but it's always
there, and it is in no way described by thd or gbw figures. This matters
much less on harmonically simple sounds without dynamics (eg a VCO), much
more on complex sounds. You cannot run away from distortion. Especially in
synths which have a huge dynamic range compared to mastered music it will
happen; and even if you don't hear it as distortion, minute amounts on
sporadic peaks will affect your perception of the sound as, basically, nice
or not nice.

You might want to check a few dozen op amps for what they will do to your
sound, and possibly pay several usd per device, or you could design your
own with exactly your desired effects in mind, and use a bunch of
sub-1-cent 0201 smd devices in a circuit that'll perhaps be smaller than
the expensive op amp plus its power supply, bypass, bandwidth limiting, and
layout which are all necessary for preventing sneaky parasitics. And also
you don't have to pay for your device to have several GHz bandwidth and
other features you'll never use.

>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://synth-diy.org/pipermail/synth-diy/attachments/20170409/44ad51ff/attachment.htm>