[NJARC] non-Polarized Capacitors

[John Ruccolo]

Hi Ray/Dave/Folks,

I agree -- use non-polarized film or oil/paper/film caps (*if* they're still good) in crossover applications. My only gripe is that they are rather *expensive* new. Two 10UF/450V electrolytics back to back cost less than 2 bucks. But a single 1UF/630V film cap (in quantities of 10) from Mouser is $2.25.

But if it's an audiophile-grade speaker system whose crossover(s) you're rebuilding, I agree with Ray and Dave -- ditch the elctrolytics.

I never think about ESR and I'll have to do some reading up on that.

JR

--- On Mon, 10/19/09, Ray Chase <raydio862 at verizon.net> wrote:

> From: Ray Chase <raydio862 at verizon.net>
> Subject: Re: [NJARC] non-Polarized Capacitors
> To: "davesica" <davesica at juno.com>, njarc at mailman.qth.net
> Date: Monday, October 19, 2009, 11:46 AM
> Just remember 
> Reply = Poster
> Reply All = Everyone
> 
> _________________________________________________________
> You bring up a very valid point about the ESR of
> electrolytics, an often 
> overlooked parameter of capacitors that varies widely in
> electrolytics. 
> Wonder how the ESR of electrolytics would vary over wide
> audio ranges?  I 
> say leave the electrolytics to the job they do best; i.e.
> filtering and 
> bypass at low frequencies.
> ----- Original Message ----- 
> From: "davesica" <davesica at juno.com>
> To: <njarc at mailman.qth.net>
> Sent: Monday, October 19, 2009 10:56 AM
> Subject: Re: [NJARC] non-Polarized Capacitors
> 
> 
> > Just remember
> > Reply = Poster
> > Reply All = Everyone
> >
> >
> _________________________________________________________
> > I've always been told that wiring two electrolytics
> back-to-back makes a
> > nonpolarized cap. I did a little reading on this
> subject and it seems
> > that manufactured nonpolarized electrolytics ARE just
> back-to-back
> > electrolytics in one can. As far as I can tell, these
> are technically
> > what is known as "Bi-polar" electrolytic caps. True
> nonpolarized caps are
> > all of non-electrolytic construction (using Mylar,
> polystyrene, etc. as
> > the dielectric.)
> >
> > For some applications, using bi-polar 'lytics should
> work fine. But since
> > we're talking about use in crossovers here, the issue
> of ESR (effective
> > series resistance) seems to come into play. I'm as
> skeptical as anyone
> > else here about the "magical" properties of certain
> audiophile
> > components, especially some of those highly worshipped
> tubes, but it
> > seems that using "nonpolarized" (or "bi-polarized")
> electrolytic caps in
> > a speaker crossover network *could* have a noticeable
> effect on the
> > performance of the system. The numbers I've seen are
> all over the place,
> > and no one seems willing or able to quote hard and
> fast actual or typical
> > ESR values, but if I'm interpreting this correctly,
> "typical" or "high
> > ESR" electrolytic caps of values commonly used in
> crossover networks
> > *could* have and ESR as high as 30 ohms or more, and
> might easily have an
> > ESR of several ohms. "Low ESR" caps (the lowest ESR
> values are found in
> > non-electrolyic caps, although "low ESR" electrolytic
> caps are made)
> > might typically measure a fraction of an ohm.
> >
> > Used in series with a 4-ohm speaker, the difference
> between a low ESR cap
> > (say, "0.2 ohms) and a standard one (at perhaps 8
> ohms) would have to
> > have a noticeable effect on the frequency response of
> the speaker system.
> > If you're unlucky enough to choose a cap with an ESR
> of 30 ohms, putting
> > that in series with the speaker is likely going to
> have a big effect.
> >
> >>From this, it seems that for best results,
> crossover caps should be
> > genuine nonpolarized film construction rather than
> bi-polarized
> > electrolytics. (Of course, if the original equipment
> used high ESR
> > bi-polarized electrolytics, then using too "good" a
> replacement might
> > also adversely affect the frequency response by making
> the tweeter too
> > "bright!")
> >
> > That's what I've been able to deduce. Any "real"
> engineers out there care
> > to corroborate my conclusions (or show me where I'm
> wrong?)
> >
> > --Dave Sica
> >
> >
> >
> >
> >
> 
> ______________________________________________________________
> NJARC mailing list
> Home: http://mailman.qth.net/mailman/listinfo/njarc
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:NJARC at mailman.qth.net
> 
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
>