[sdiy] "Boutique" capacitors

[Magnus Danielson]

From: "Batz Goodfortune" <batzman-nr at all-electric.com>
Subject: Re: [sdiy] "Boutique" capacitors
Date: Sat, 05 Feb 2005 12:05:01 +1030
Message-ID: <5.1.0.14.2.20050205113322.03aeb070 at mail.all-electric.com>

> Y-ellow the person formerly known as S.
>          Just quickly. If you're talking about Electrolytic caps then, in 
> general, Vintage = bad.
> 
> Electros have a shelf life. The better ones have a longer shelf life. There 
> is much written about this and I'm not up for engating my brain just now 
> but I'm sure you'll find all you need to know at one of these fine web sites.
> 
> http://www.faradnet.com/
> 
> http://www.tpub.com/electronics.htm
> 
> And I'm sure there are many more that people here could chime in with. If 
> memory serves, Harry wrote a particularly good run down on caps. Am I 
> correct here?
> 
> Newer caps are almost always going to be better. Better technology and 
> being newer, have a longer useful life. There are exceptions however, and 
> this is where the sound-card thing may come in...
> 
> A while back there was a story about how some big Taiwanese cap maker 
> thought they hit the mother-load when they "Acquired" the top secret 
> formula for the dielectric used in some "Top Brand" low ESR Japanese 
> Electros. They went into production with them, offered them cheaper to 
> mother board manufacturers etc, and several months later, all these high 
> end computers were mysteriously dying. And some of these boards were the 
> REALLY expensive ones.
> 
> Turns out that the formula they had "acquired" was a fake. Put there for 
> the specific purpose of catching out industrial spys and the folk who pay 
> them. You see, it's all in that formula and each fab has their own recipe. 
> And that's where the different lies. Just because it says they're low ESR 
> doesn't mean they actually are. If crappy caps had worked their way into 
> the sound card market at some stage, (which is entirely likely) then 
> chances are replacing them with just about anything else would have made a 
> huge improvement.

Just one out of many listings on this story is here:
http://www.burtonsys.com/bad_BP6/story2.html

I've seen it in IEEE Spectrum too, a few years back.

> And I won't go into the greed of Motherboard Fabs who must have thought 
> they were getting the bargain of a life time.
> 
> Likewise, replacing old, really old caps with new ones stand a good chance 
> of making an improvement. Even if they're not some hot recommended brand by 
> a bunch of audiophiles. Who generally have no clue about electronics.

I agree. Electrolytic caps age, they dry out. Modern caps has a much slower
process as far as I've understood it.

> "I don't know much about the art of electronics but I know what I like."
> 
> And I'll restrain myself here, from listing a litany of great jokes told by 
> audiophiles as truths.

One doesn't know if one should laught, cry or listen carefully. Every once in a
blue moon, there might be something to learn, but it may not be on what you
initially expect.

> Generally speaking though, the general consensus is that Electrolytics are 
> evil nasty things that should have been drowned at birth. But 
> unfortunately, quite often a necessity.

I agree.

> Hope this helps.
> 
> Be absolutely Icebox.
> 
> PS: I only mentioned "vintage" because I don't think I've seen a NEW 
> Rubycon cap for maybe 10 years.

Rubycon naturally has a website:
http://www.rubycon.co.jp/en/

Now, low ESR can bite you in the chair-interface part of your pour soul. It may
or may not be what you want. Also, people spend too much time on the big caps
but not as much on the smaller ones. The trick is to combine the caps such that
the total impedance stay low over an interesting range of the frequency
spectrum. In the low end, the caps will support the output impedance and
additional wiring inductance from the PSU regulation at frequencies where it
fails to track the deviations fast enought. Infact, even when the cap is acting
as an inductor (beyond the self-resonance) it is usefull since the impedance is
low enought. It's a bit aquard in the beginning, but it makes sense in the end.

Cheers,
Magnus