Path: utzoo!utgpu!news-server.csri.toronto.edu!bonnie.concordia.ca!thunder.mcrcim.mcgill.edu!snorkelwacker.mit.edu!usc!wuarchive!zaphod.mps.ohio-state.edu!samsung!munnari.oz.au!yoyo.aarnet.edu.au!sirius.ucs.adelaide.edu.au!hydra!francis
From: francis@cs.ua.oz.au (Francis Vaughan)
Newsgroups: sci.electronics
Subject: Re: How do I choose a capacitor type ?
Keywords: distortion capacitors
Message-ID: <2348@sirius.ucs.adelaide.edu.au>
Date: 6 Feb 91 01:07:51 GMT
References: <64@owl.ukc.ac.uk> <1991Feb5.185727.14775@news.larc.nasa.gov> <4193@eastapps.East.Sun.COM> <1991Feb5.211414.16831@news.larc.nasa.gov>
Sender: news@ucs.adelaide.edu.au
Reply-To: francis@cs.adelaide.edu.au
Organization: Adelaide Univerity, Computer Science
Lines: 95
Nntp-Posting-Host: hydra.ua.oz.au

In article <1991Feb5.211414.16831@news.larc.nasa.gov>,
kludge@grissom.larc.nasa.gov ( Scott Dorsey) writes:
|> In article <4193@eastapps.East.Sun.COM> gsteckel@east.sun.com (Geoff
Steckel - Sun BOS Hardware CONTRACTOR) writes:
|> I wrote:
|> >> For coupling capacitors in audio circuits (DC blocking caps), use
polyester
|> >> or polypropylene caps.
|> >
|> >The `true believers' in audio wonderfulness would never use a polyester cap
|> >anywhere.  In fact, they do have a large `dissipation factor' and
apparently
|> >are somewhat nonlinear, as are `high-K' ceramics (with Z5U or Y5V tempco
|> >markings).
|> >I would recommend polystyrene or polypropylene film caps for
wonderful coupling
|> >caps, or low-K (NPO) ceramic or mica caps for small values ( < 100 pF).
|> >They will be much larger physically and cost much more than the
polyester caps.
|> 
|> That's right.  Make sure you buy only Wondercaps or some expensive foreign
|> brand.  Only polypropylene will do, and the harder they are to get,
the better
|> they are.
|> 
|> Anyway, polyesters are usually quite fine capacitors for coupling, although
|> at high voltages (in in working with tube gear), the polypropylenes have a
|> good upper hand.  Ceramics are bad news.  Electrolytics
|> in the audio path are bad news.
|> 

There is a fairly well established precedence rule in selection of capacitors.
This is based on minimising a number of factors. And is very much horses for
courses. There are quite well established standards for defining these 
charateristics (including some MIL Spec ones). 

You can actaully measure the distortion produced by a capacitor in  a simple 
test rig. With some dieletrics the results are not good. (> 1%). The
distortion 
is often related to the input voltage (low volts, worse distortion) and shows 
some correlation with DA.

All capacitors have a self resonant frequency, and capacitance tends to drop
around here. Many designs seek to get this frequency as high as possible by
lowering the inductance with suitable mechanical design. 

For audio work the critical parameters are distortion and dielectric 
absorbtion (DA).

The best capacitors are PTFE dielectric. These are almost unobtainable in 
anything but very small values. The dieletric film is quite hard to make 
(it is hard to make without imperfections). PTFE has almost vanishing DA.

The next favorite dieletric is Polystyrene. These are available in values 
up to the low nF without much trouble. DA is almost as good as PTFE. I have 
had no problems in replacing ceramic caps in audio circuits with styrocaps 
even though I was a little worried about stability in feedback loops.

The next dielectric down is Polypropylene. Again nearly as good as 
Polystyrene. Available is quite large vaules (10mF) but big and not cheap. 
Most people I know use Roedenstein (SP?) caps here, Wondercaps are many 
times the price for no proven electrical advantage. Polyprops are available 
in 1% tolerances and are always the first choice for active cross overs and 
RIAA networks. They typicly have very good thermal charateristics too. 
(Low delta-C for delta-T)

Next in line is Polycarbonate, bringing up the rear of the best dielectric 
group. Available in reasonable values, they are about 1/2 to 1/3 the size 
of Polypropylene for the same value and rating. They exhibit noticable greater
DA than the others but are still quite good. 

Then comes Polyester. These capacitors exhibit something like 4 times the DA
of the previous group. They are at least cheaper. Thermal charateristics are 
nothing wonderfull and they are mostly 10% or worse tolerances.

Bad capacitors are high  K ceramics (many thousands or time worse DA and 
distortion) electrolytics, and really bad are tantalums. Low K (NPO) ceramics 
are better but still much worse than any of the plastics. However sometimes 
there is no option if you need to tame instability. Tantalums really have no
place anywhere near an audio signal. Leave them in power supply
bypassing where 
they are really good. There are a few intersesting ways to improve
electrolytics 
(and unfortunatly an electro is often the only way to get the Farads). 
The higher the voltage rating the better is one theory. Bypassing with
plastics 
is a must. Bipolar (back to back electros) seem to be a bit better, and
if they 
are biased with something approaching their working voltage things improve as 
well. (Hard to do, often the whole purpose is to get rid of a voltage offset, 
and the bias voltage must be very clean, or you ruin everything). Impeadances
can be hard to work out too.


						Francis Vaughan