[sdiy] lowpass gate etc (apologies to Kurt Cobain)

[harrybissell]

HiAnthony....

at the risk of having people say that I have balls the size of Texas...
(or perhaps saying they are the size of BBs... :^)

I'll start by saying that probably most people will not be able to tell the
difference between most filter capacitors.  Woah.

You may be able to tell the difference where the capacitor has VERY
significant 'parasitic' components...  for instance Electrolytic caps
have quite a bit of ESR (equivalent series resistance) and a substantial
self-inductance, large dielectric absorption, and quite a variation of
value with temperature.  OTOH most people avoid electrolytics for filter
circuits
because they are usually too large in capacitance... and because their
parasitics make them poor candidates for the application.

You mentioned Polystyrene... which is probably the best overall cap for
filter applications.  Sliver Mica would be a close second... equal in almost
all parameters except is has larger dielectric absorption.  I have not done
a lot of research on this (and people are welcome to jump me on this) but
probably the dielectric absorption will not be TOO bad at audio (AC)
frequencies.
By the time dielectric absorption would be an error source, the filter input
will
probably have reversed polarity. At very low frequencies it may begin to be a
factor...
(I"m guessing that would be sub-audio anyway)

You are absolutely correct in your intuition about the NP0 ceramics. They are
the
best, most linear dielectric in the ceramic family.  Other dielectrics like X7R
(fair) and Z5U (poor) will vary capacitance with applied voltage ie. distort.
They have fair to poor temperature stability...   Most people wisely avoid the
higher Q dielectrics. Use them only for decoupling caps...

Can you hear the distortion ???   I'm not sure I can.  Audiophiles will swear
they
can (otoh they are Audiophiles, what would you expect from them :^)

I think the biggest thing that you can hear is capacitor tolerance. Capacitors
with
a tight tolerance (or matched caps) will usually result in higher "Q" values...
they
will probably resonate better and over a wider range of frequencies. Cutoff
slopes
can be better as well, and I think that you WILL hear those effects.

If I was looking for different sounding filters, I'd probably just choose
different
filter topologies (didn't Kurt Cobain sing "what else could I be... all
topologies..." ?)
A State Variable will sound different than a Sallen-Key for sure. I think the
component
variations might be too subtle to make a lot of difference

H^) harry



anthony wrote:

> I'm in the middle of making a dual lowpass gate like the Buchla 292C. I
> remember reading someone saying they'd made some each with different types
> of capacitors for different sounding filters. Before I finish them up and
> can listen for myself I was wondering just what different capacitors can be
> expected to sound like in different audio circuits. I think the impact of
> each different type might be different for the different uses for capacitors
> in audio circuits - like feedback networks or coupling caps or caps in a
> filter. In my circuits I'm using all polystyrene in one and a combo of
> polystyrene and silver/mica in the other. I had thought about using NP0's
> because I had this funny idea that they probably sound better than most any
> other ceramic. But then I decided that this was probably silliness and
> decided to not use any ceramics at all. Do ceramics really sound all that
> bad? I've never actually taking the time to compare the sounds of variosu
> capacitors. What would be the best kinds of circuits to use to test how
> capacitors sound? I've been wanting to make a headphone amp out of some
> medium-mu twin-triodes - maybe that'd be good.