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

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

You haven't used one as an offset trimmer in a Hi-Z musical instrument amplifier. (Where did the signal above 3 KHz go?)

Please don't top post.

John

That at least is measurable. I'm quite sure the effect, if present, is of no significance in the audio range.

I sort of wish Graham hand't cross-posted to audio and electronic design groups; this subjective nonsense doesn't belong here.

John

No but the science bit is useful.

I'm quite intruiged by the noise mechanisms at work.

Graham

The usual reason for preferring plastic pots for audio is that they don't crackle when you turn the knob, because the surface is smooth and soft compared with cermet.

Cheers,

Phil Hobbs

Capacitance to what ?

Err...... Not really. The input impedance of the following stage is not a big issue. It's normal ( and good practice ) to load a pot with ~ > 5x the track resistance.

Ummm... No. They're used precsiely because they wear very *slowly* ! Widely used in industrial controls for that reason.

The ones that wear quickly are the common carbon track types found in almost everything audio except really high end. CP's are inbetween but much better than carbon.

I'm not keen on step changes myself.

The thought had crossed my mind.

That's my suspicion too but I'd like to see it proven !

Amusingly yes, the supposedly fabulous tubes have the worst microphony issues. It may please the listener though ! It probably adds some low level reverb actually !

Graham

Yes indeed but you only hear this difference when adjusting the control. Better than carbon pots too btw.

You see, it's not clear if that's what was meant by 'better' in this instance, although I thought he claimed that the cermet 'sounded better'.

Graham

That is easily testable; the noise would increase as a power of applied voltage, whereas fractional shot noise decreases.

John

Worn pots make spitty/static-y noises in guitar amps all the time.

Cleaning them doesn't help. The amount of vibration in a Twin Reverb being played balls-to-the-wall is *horrendous*. Interestingly, sometimes pots which make these razzy noises are completely quiet when rotated. The test for this problem is to put a low-freq signal through the amp (50 Hz is fine) and run it up to about a quarter of its rated output. Tap the knobs gently with a soft-face hammer. If the wiper is jittery, the 50 Hz tone will spit when the pot is whacked. Replacing them is the only thing that works.

This is true. (Not the part about adding reverb, though. The distances involved in the feedback path are too short to produce reverberation.)

Microphonics is a form of positive feedback - regeneration, as it were. Many of the tubes with "legendary" reputations - in guitar amps, anyway - have poor performance in the microphonics department, especially Telefunkens and Mullards. (Larger-plate tubes do it worse.) In some cases, the frequencies thus reinforced are pleasing to the guitarist - and in some, not. This is why you see mountains of utter horseshit on the Net about how certain tubes sound certain ways; without taking the local conditions such as circuit topology, loading, applied voltages, etc., as well as ancillary equipment, playing style, input signal strength, control settings and such into account, it's nearly impossible to make tonal judgements without actually plugging the glass in and earballing it yourself. General statements/predictions can be made, of course, based on previous experiences with the same or similar gear, but I've seen writeups on tubes that had more adjectives in 'em than a thesaurus.

Lord Valve Expert

Gotta hand it to you there. Why would I do a thing like that?

But why would capacitance matter in an offset trimmer?

Interestingly, I've done some recent experiments (resulting in some products) to see how high a frequency system can have its gain cleanly tweaked with a surface-mount trimpot. I'm doing it now at 200 MHz, and

John

In article , Phat Bytestard wrote: [....]

The truth is that they emit wide band electromagnetic noise. The spectrum goes from DC to daylight. The bit that happens to be conducted out on the leads we call electrical noise. There is IR and microwave and radio noise there too.

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

I think that because I've spent the last 30 years studying the physics of electronic components, both active and passive, in order to generate reliability assessments, failure analyses, and other such administratrivia. The summation of a shot noise profile and a Johnson noise profile mirrors the measured results most closely.

Why do you think that?

Do thickfilms have shot noise?

John

The Johnson noise spectrum doesn't quite make it to daylight, but peters out in the thz region. It couldn't be true white noise, because then the power would be infinite.

John

Shot noise and Johnson noise are both "white" which is to say, frequency independent.

What kind of noise profile lets you separate the two?

If you are saying that you are seeing more white noise than you'd expect from the resistance and temperature of your device, you can't claim that the excess is shot noise - there are other sources of white noise.

I don't think that you have made your case.

Please take this in the spirit of childish pedantry it's intended but it makes me cringe when presumably educated people use the "word" irregardless. A simple, "regardless" will do. Thank you, s.

snipped-for-privacy@ieee.org wrote in news: snipped-for-privacy@p79g2000cwp.googlegroups.com:

Voltage vs Time. given a resistor in thermal equiibrium, apply a step increase to the current through the resistor. some noise will follow the change in current instantly. additional noise will track the Johnson noise expected from the change in temperature generated by the change in current, which is not instantaneous, but dependent on the ability of the substrate to sink the additional heat generated by the current

Yes, there are additional possibilities. Analysis of these sources doesn't account for all observed behavious, therefore some noise is shot noise.

That could also be explained by hypothesising a range of thermal paths to the substrate within the resistor - the less effectively heat-sunk resistance paths would heat up faster and further.

You can't explain it any other way, so it has to be shot noise. Not all that convincing.