the tone cap myth...

[Joe Gwinn]

Can you measure the dissipation of the oiled-paper and mylar capacitors you are comparing? Most LCR meters can make this measurement.

More generally, to get a valid tone comparison, you must match the capacitance values to within a percent or two using a meter, or the difference in RC filtering effects will mask any effect from the dielectric.

[RedHouse]

I had similar experiences, but not "all" PIO caps had delightful results. I had some actually dulled the entire range of the tone pot and made all setting un-usable. It's hit-n-miss really.

Eventually I went back to polypropylene in my personal Lester Polfus

[freefrog]

Can you measure the dissipation of the oiled-paper and mylar capacitors you are comparing? Most LCR meters can make this measurement.

No, I can't measure it but your answer suggests that the "X factor" is this question of dissipation. Thx for the info!

More generally, to get a valid tone comparison, you must match the capacitance values to within a percent or two using a meter, or the difference in RC filtering effects will mask any effect from the dielectric.

I understand and agree but I wasn't expecting two 30n caps to deliver a brighter tone than two 22n... :-)

[Mike Sulzer]

-at 10/10, the sound is actually brighter and more open than before.

At 10/10 you have a 500K resistor in series with the capacitor, which, if functioning, has a much lower impedance than 500K at the higher frequencies.

A useful test to get an idea of how much difference a capacitor can make at 10/10 is to install a known good cap, and then put a switch across it so you can easily short it out. Then do a double blind listening test.

You can also do this test with a known bad cap, an open circuit. That will tell you how much difference the resistive load of the pot makes.

[Joe Gwinn]

No, I can't measure it but your answer suggests that the "X factor" is this question of dissipation. Thx for the info!

This is one theory, as yet unproven. The alternate theory, also unproven, is that the type of dielectric makes no difference whatsoever, so long as we are not talking about electrolytics.

I understand and agree but I wasn't expecting two 30n caps to deliver a brighter tone than two 22n... :-)

How do you know that a capacitor claimed to be 30 nF is in fact 30 nF? Capacitors can easily have +/- 20% or +/- 50% tolerance, and 30/22= 1.364 to one, well within the overlap between nominal values.

If you don't have a good LCR meter, you're basically wasting your time.

[freefrog]

This is one theory, as yet unproven. The alternate theory, also unproven, is that the type of dielectric makes no difference whatsoever, so long as we are not talking about electrolytics.



How do you know that a capacitor claimed to be 30 nF is in fact 30 nF? Capacitors can easily have +/- 20% or +/- 50% tolerance, and 30/22= 1.364 to one, well within the overlap between nominal values.

If you don't have a good LCR meter, you're basically wasting your time.

I have a good LRC meter. :-) Sorry if my previous answers suggest that it's not the case - English is not my mother tongue and I'm not electronician, so, misunderstanding is something to expect with me !-))

anyway, thx to waste your own time to increase my limited knowledge of hobbyist.

[FunkyKikuchiyo]

freefrog did say in his initial post that he measured the caps to get true readings.

I agree with what your ears are telling you, actually. A really good cap (especially PIO) often sounds like it has more highs, even when capacitance readings would imply otherwise. My personal explanation is that THEORETICAL capacitors will pass certain freqencies and block other frequencies. REAL capacitors have leakage currents, losses to heat, resonances and microphonics in addition to their basic capacitance duties. All of these things, regardless of whether it is wired in line or wired to ground, effectively create a sonic foot print and this is how we associate a certain "sound" with dielectric types. If you ask an electrical engineer about this who has never dealt with audio and thus immune to hi-fi/guitar mag hypnosis, he'll tell you that it is perfectly valid. After all, this is why all of these capacitor types exist. It isn't just to humor audio dorks.

Red House makes a good point. PIO caps are probably so expensive because they are so easy to screw up. They have a high failure rate, and I have some that sound TERRIBLE. I don't know what it is exactly, but the low notes have these beats in them, like when you have tuning issues or when the pickups are too close. Microphonics maybe? I'm not sure. I've had russian surplus PTFE caps do this, too.

[Joe Gwinn]

Mention of some really bad-sounding capacitors perked my ears up - we should collect and test these. If we can figure out why they are bad, it may be the clue as to what to look for in less dramatic cases


War story: In the 1970s, I worked for a very small company that made equipment for pharmacological research. Their main product measured how much rats or mice ran around in their cage, which as it turns out is useful to know.

One of our best customers returned a unit to us, saying that it was not getting the same control results as all the others he had been using for years, and the question was why. We were mystified, because the circuit et al had not changed.

But this customer was a very careful experimentalist, and we believed his not-same result. The sensors were optical in nature, triggering if the rate of change of illumination exceeded a threshold. Initial tests showed that the old and new sensors differed in the required rate of change, the new sensors being slower. But the circuit was the same, so a component must have changed despite being nominally the same.

So I took the slowest new sensor and the fastest old sensor I could find, and started swapping components, to see which ones mattered. By a process of elimination, I found that the gain of the first transistor of the newer sensors was much higher than that in the old sensors. Bingo. (Miller effect causes higher gain to be slower.) So we replaced all the first transistors (about 40 units in total), and uniformity was restored.

[FunkyKikuchiyo]

Why couldn't you have adjusted the bias/drain current on the transistors? That probably is a stupid question... I know precious little about transistors.

I still have the mediocre caps. If there is something I can test with an Extech or a DMM I'm more than happy to do so and share the data. I have six Russian military surplus teflon caps that have a weird beating sound, and some sort of paper in oil cap that came from Allparts that does the same thing, although not as noticeably.

For some reason at one point I was attributing it to the size/substance of the conductor, like a metallized film vs. foil sort of thing - I don't remember my logic on that though and I may have just pulled it out of my fanny...

[Steve Conner]

What Freefrog is describing sounds like a cap with really high ESR. The sound being brighter on "0" is the clue, the way I see it at least.

I used to be on a crusade against all this capacitor stuff, until I read the series of articles by Cyril Bateman. Now I don't know what to think.

If you find some funky old capacitor that does it for you, great, use it and enjoy it. I still believe that it would have sounded like any other capacitor when it was new, and the extra mojo is actually degradation. Hopefully it will be stable and not change its tone any further.

[Joe Gwinn]

Why couldn't you have adjusted the bias/drain current on the transistors? That probably is a stupid question... I know precious little about transistors.

One could, but it would not eliminate the Miller effect. Miller effect - Wikipedia, the free encyclopedia

I still have the mediocre caps. If there is something I can test with an Extech or a DMM I'm more than happy to do so and share the data. I have six Russian military surplus teflon caps that have a weird beating sound, and some sort of paper in oil cap that came from Allparts that does the same thing, although not as noticeably.

Step one would be to use the Extech to measure capacitance and dissipation of the bad-boy caps, at 120 Hz and 1 KHz.

For some reason at one point I was attributing it to the size/substance of the conductor, like a metallized film vs. foil sort of thing - I don't remember my logic on that though and I may have just pulled it out of my fanny...

I suppose that the series resistance might matter, if extreme.

[jmaf]

Lester Polfus

I had to look that up on Wikipedia ... I should be ashamed of myself.

[LtKojak]

i should be ashamed of myself.

yes, you should!