Vishay talk bollocks - was re: Teflon Film Tin Foil capacitors

Eeyore wrote in news: snipped-for-privacy@REMOVETHIS.hotmail.com:

Like I guessed. Nonetheless, I've heard good reviews from people that have used them. I was mostly on Roederstein MK3s, and in critical positions like feedback, Caddock MK132, but I may switch to these, as the Roedersteins are long out of production and getting hard to source, while the Caddocks are just as expensive as these metal foils Vishays.

They also say '...a compliment of commercial parts...' - they really meant complement unless they are praising the extra noise ;)

Cheers

PeteS

This is pretty bad. The "trained ear" thing, and the testimonial as a result of a non-blind hearing test, are typical sonicbabble.

Their claim that crystal grain boundaries are the source of resistance is simply absurd. Single metallic crystals, like turbine blades, are not superconductors.

It is true that metallic resistors have much less shot noise (which they strangely call "current noise") than carbon or oxide resistors, which might matter in audio applications but seldom does. In something without DC bias, like an attenuator, it simply won't matter.

They rate thickfilms as "unacceptable" at -10 dB, which is -130 dB shot noise with their scale (uV/V) properly adjusted. If they had honestly shown the resistor table as going from -160 to -95 dB, it would have been obvious that this issue is mostly bogus. Where are you going to get a signal source that has a 120 dB s/n ratio?

Vishay foil resistors are good in some hyper-precision applications, but are insane for audio. Because of the way they're trimmed, their series inductance varies from part to part, and their heatsinked parts have ghastly eddy-current hooks, 2000 ppm or more transient errors, which is why we make our own current shunts.

But mainline companies like Vishay and Belden write golden-ear nonsense to play to the audiophool niche market.

John

The Caddock stuff is thickfilm, which has classic shot noise. This doesn't matter, you know.

John

some

You can't even spell its correctly. I find it amusing that someone who can't spell a common three letter word complaining about someone else's spelling of a foreign name... /dry humor off

John Larkin wrote in news: snipped-for-privacy@4ax.com:

Maybe Vishay hired that id10t Kevin Aylward, who called me all sorts of terrible things when I told him that resistors have shot noise... and said a lot of other remarkably stupid things as well.

some

An annoying rare example of a typo for me.

Graham

some

Really? I remember one just a few days ago ;-)

...Jim Thompson

John Larkin wrote in news: snipped-for-privacy@4ax.com:

But Kevin claimed that he's a professional engineer, and has worked designing analog monolithic circuits for Texas Instruments.

Pretty much the same here,

Since the audio types can't measure what's happening, truth becomes a matter of hormones and rhetoric.

Of course, Graham deliberately kicked the beehive when he posted something like this to both audio and electronics design groups, natural enemies if ever there were such.

I *do* work at 120 dB s/n ratios over DC-50 KHz bandwidths and I do use surface-mount thickfilm resistors.

John

snipped-for-privacy@ieee.org wrote in news: snipped-for-privacy@s13g2000cwa.googlegroups.com:

Granted. but there are good resistors and bad resistors. and a bad thin film deposited on a contaminated substrate will exhibit an excess noise, which when analysed is different from the excess noise of a cruddy carbon resistor, and is in fact, truly, shot noise.

Yes, and there are several mechanisms by which this excess noise may be generated. Some is due to the mechanical stress and strain imposed on the resistance element. Some is due to electron tunneling through the non-conductive portions of the device.

The fact that you lump it all into "excess noise" just means you haven't analysed its origins.

Not natural enemies IMHO. Certainly shouldn't be. I believe that science holds the relevant answers if only correctly applied.

There do seem to be any number of clowns in the 'esoteric' audio biz who simply don't believe in science though.

Your point about the uV/V graph being displaced by 120dB from normal is highly relevant and a great way of being disingenuous by Vishay.

Having said all of that, I have come across cheap 'n nasty metal film parts that are significantly noisier than those from the 'better' manufacturers btw. I can't however see any rationale for requiring anything better than good quality metal film in audio.

Graham

Good resistors don't have shot noise - which is what you see when your current is flowing as single non-interacting electrons or holes, so that the currnt is genuinely quantised.

Good resistors do have Johnson noise, which simply reflects the electron cloud carrying the current is a room temperature so that the individual electroncs are doing a bit of Brownian motion on top of the drift due to the elecric field.

Crap resistors have "excess noise" which is to say, more than Johnson noise.

Kevin Alyward understands this sort of stuff, and you clearly don't.

On Sat, 22 Jul 2006 11:48:12 +0100, Eeyore Gave us:

Bullshit. Still widely used in the power supply industry. Bulk mediums are required for arc suppression, and are also good as current limiters in HV applications.

On Sat, 22 Jul 2006 11:48:12 +0100, Eeyore Gave us:

You're a brainless wonder. Of course you cannot abate all noise. The goal is to make one as noise free as possible.

On Sat, 22 Jul 2006 11:48:12 +0100, Eeyore Gave us:

They mentioned four resistor manufacturing methods.

On Sat, 22 Jul 2006 11:48:12 +0100, Eeyore Gave us:

The misspelling of Boltzmann merely means that a non electronics trained person write the PDF. The data, given to that person was written as given, and that person made no embellishments whatsoever.

It was a misspelling. Period.

On Sat, 22 Jul 2006 08:09:58 -0700, John Larkin Gave us:

A modulated laser, and the output from the laser detector.

John Larkin wrote in news: snipped-for-privacy@4ax.com:

and

amongst other things, image sensor front ends for telescopes.