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

snipped-for-privacy@panix.com (Scott Dorsey) wrote in news:ea57a4$ecd$ snipped-for-privacy@panix2.panix.com:

So its an issue of the wiper arm lacking the proper tension to keep it from bouncing?

When you turn the pot, it bounces up and down across the surface. There is mechanical noise, and a corresponding change in resistance that causes electrical noise.

When you stop turning the pot, the noise goes away. As a result, these are great for trimmer pots and things that get occasionally adjusted, but unacceptable for controls you'll be riding all day long like channel gains.

--scott

test

difference

That will just result in a tiny amplitude modulation of whatever sound level is there at that instant, and the sound level is changing all the time anyway, especially so as you're turning the pot on purpose already. That won't be very, whether the pot is cermet, plastic, or wirewound.

The difference with DC is that you can't hear DC, but you can hear the small variations in DC - real noise - that a bad pot can create.

OK, new audiophool product: volume controls based on variable capacitance.

John

Right, I'm quoting the Planck *rolloff*. It's normalized to 1 at low frequency.

Cheers,

Phil Hobbs

Sometimes that rolloff matters, even in circuits. I'm building electrical detectors (not photodiodes or photoresistors) that detect 200 THz signals by rectification. (Current efficiency level is about 4%, but they've only been working for a month or so, and the next best reported number is about 0.1%.)

Cheers,

Phil Hobbs

Please explain how a change in resistance creates voltage.

John

Be that as it may, he didn't support his identification of the excess noise as shot noise with any kind of meaningful argument - he just talked about a "noise profile" which is just a tad unspecific.

It does not create voltage, it creates a change in the drop in voltage from the signal that is already across the pot. If there is no signal and no DC offset, there will be no noise either. But if there is no signal and no DC, there is no need for a pot.

--scott

--- Old pro audio product: volume control based on a photoFET:

-- John Fields Professional Circuit Designer

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

command=text&at

Thanks for the memories, Of candlelight and wine, castles on the Rhine

and old Fairchild compressor/limiters and McCurdy broadcast consoles.

Well, pots don't "jump". And all your math indicates is that the signal gain changed, not that any new "noise" signal was generated anywhere. And the signal gain changed because, well, we weren't happy with the volume and we decided to get off our butts and turn the knob.

I don't think anybody can hear the difference between properly working cermet and conductive plastic pots. It is interesting that high-end golden-ear folks tend to prefer stepped rotary-switch based attenuators over pots. And stepped attenuators sure do have massive gain jumps, orders of magnitude more than the case you analyzed.

John

test

difference

noise.

Which is noise ! I'm talking about sharp step changes, not the 'modulation' intentionally caused which is vlf.

Practice shows otherwise.

Those variations may not be so 'small' as you think.

Besides, you should already know that it doesn't matter if the input is ac or dc. Try doing it with lf ac and you'll hear just as much noise as with dc. Obviously higher frequencies are more audible and tend to mask the noise but it's still there.

Check out 'zipper noise' for example.

Graham

Change in Vout / Vin obviously.

Graham

--- Sure they do, since they've got "stiction" in there, somewhere, and when you go to crank in that last little bit of resistance you need to get to 1.234 volts it skips over the spot you needed and now you're at 1.256 volts and you've got to go back the other way...

---

--- That's not true. Depending on the amplitude of the step and how abruptly it changed, spectral components not present in the program material (noise, IOW) will be present in the signal leaving the pot and propagating through the rest of the system.

---

--- So what?

---

--- That wasn't the point. The point was that pots can add noise to a system.

---

--- Sure, but that's a whole 'nother issue.

-- John Fields Professional Circuit Designer

CdS type photocells have also been used in this application btw.

Graham

Well, they do actually, or they wouldn't make any noise when being rotated ! Of course the effect varies widely with construction and materials not to mention age ( cycles of use ).

A step change in amplitude is indistinguishable from noise. Remember that those step changes will be in both directions as the pot as rotated in one direction.

That's because you're not familiar with their use in this respect it seems.

But no 'noise' though. I'm no fan of stepped attenuators btw. The cheaper ones btw are simply pots with mechanical detents !

Graham

test

difference

noise.

I wasn't considering bad pots; they could cause gross dropouts that you'd certainly hear.

dc. Try

higher

Well then, I guess you should go to stepped attenuators to avoid the, err, gain steps.

John

Some wise person once said, "A step change in amplitude is indistinguishable from noise."

John