🚀 go-pugleaf

In the capacitor wars, I've got to agree that electolyic capacitors
are terrible.  Perhaps, it would be a good idea to start a Jihad
against them.  As mentioned, electrolytic units are appropriate for
brute force bulk regulation of power supplies.  In power supplies,
the continuous application of DC keeps the liquid dielctric
material formed and acting as an insulator.

One point to keep in mind is that capacitors integate current.
When a large brute force filter capacitor is present in a supply,
it will draw a very large current while it attempts to nearly
instantaneously reach operating voltage.  This property can result
in blown out fuses and/or rectifier diodes if percautions are not
used.  A number of methods are possible to slow down the initial
charging.  Often, a current limiting resistor is placed between the
rectifier and the filter capacitor.  A modest timing circiut can be
used to close a relay contact across the resistor.

One other thing is electrolyitic units is that there is finite
leakage current.  Sometimes, the leakage current is considerable.
I once saw a guy burn up a plate supply transformer in a tube
preamp by using a humungous filter cap bank.  It turned out that
the filter bank drew more leakage current than the preamp itself
drew to operate.

I try to avoid electrolytic capacitors in audio circuits.  I'm
amazed at how many times I've seen electrolytics used to directly
couple the audio path in Compact Disk players.  They'll almost
certainly fail after a few years because there isn't a DC bias
(usually) of sufficient magnitude to maintain the integrity of the
dielectric film.  Perhaps this is Japan getting even for WW II by
building in planned failure into their products?  Note that smaller
"working voltage" values are useful if you insist in using an
electrolytic cap in an audio circuit.  A 5 volt rated capacitor is
more likely to stay working than a 25 volt rated electrolyic
capacitor in a circuit that has a 3 volt DC bias.  It's difficult
to believe that with all the fancy electronics, optics and
mechanics in CD players that engineers can't concoct a stable
direct coupled audio output circuit.

For general non critical applications, I use stacked mylar (tm)
capacitors.  These are OK for most audio applications such as for
bypassing the Miller-effect feedback resistor in the emitter of a
transistor circuit.  Values are commonly available in the nF to 1
to 2 uF range.  Mylar has a quite modest dielectric absorption,
and is value stable at consumer temperature ranges.  A product
example is Panasonic V-series.  They are also pretty cost
effective.  On drawback is that tolerances are about +/- 5 to 20%.

For critical needs such as sample & hold capacitors and highly
stable audio oscillators, I use polystyrene capacitors.
Polystyrene is very temperature stable and has very low dielectric
absorption.  Mallory offers two lines of polystyrene units.  One
line has a leakage R of 100-1000 megohms.  The other series is up
over 1000 megohms.  The disadvantage of polystyrene is that it
results in physically large capacitors.  It is also modestly
expensive.  Values range from pF to nF typically.  Tolerances are
available as low as +/- 1%.

Polystyrene degrades in dielectric performance at RF frequencies.
For RF, mica capacitors are quite good. Silver/mica capacitors are
quite expensive relative to ceramic types.  Mica offers very good
temperature stabilty and quite good dielectric performance.  Mica
caps are moderate in size.  C vales are .5 to 1000 pF
approximately.  Tolerances are available to +/- 1% or better.  A
typical vendor for mica units is Cornell-Dublier Electronics (CDE).

Polycarbonate capacitors are better than polystyrene in performance
of the dielectric.  Polycarbonate units are also available from
several sources with very tight tolerence values.  Unfortunately,
polycarbs often cost several DOLLARS each (ouch!).  They're mostly
used for applications where high untrimmed accuracy and wide, wide
temperature range stability is needed, for example space based
radios.

Most electronics industrical suppliers such as Newark or Pioneer
should be able to burry you in a ton of capacitor data sheets
(well, almost) if you phone-up and request some.  Much of the
discussion present in magazines such as "Audio Amateur" seems more
opinion than fact-- especially in re capacitors.  I'll admit some
opinion that I'm a relatively demanding listener; I really can't
acknowledge any real noticable sonic differences in mylar,
polystyrene and similar type capacitors when used appropriately in
audio coupling circuits.  Electrolyic units do certainly add
noticable coloration to sound, but then electrolytic capacitors are
readily measurably inferior in performace on very inexpensive test
gear (such as a $14.95 pocket multimeter from Radio Shark).

I get a laugh when I see electrolytic capacitors used in coupling
applications in CD players.  In an apparent move of appeasement,
they often have a red plastic sleeve (as opposed to the more common
blue or orange) that is emblazoned "AUDIO GRADE".  I wonder just
who the manufacturers are trying to impress, since the same CD
players bear labels on the outside of the cabinet saying something
to the effect of "dont open-- danger of hideous government
punishment.." or some similar rot.  Seemingly, "AUDIO GRADE" ==
"inferior to industrial grade".

  --Bill