Many audio circuits use PIO capacitors. What are the electrical Characteristics of PIO caps that make them special (i.e. justify the cost)? Which parts of an amplifier circuit benefit most from the use of PIO caps?

  • 12
    \$\begingroup\$ They form a special magical reaction with the oxygen free cables, the special wooden box the amplifier is in, and various other superstitions only audiophools can see. These things aren't meant for ordinary muggles who have to do electronics by applying physics and math. \$\endgroup\$ – Olin Lathrop Feb 1 '13 at 13:17
  • \$\begingroup\$ Didn't we have a question like this few years back... If I remember correctly, it had great answers. \$\endgroup\$ – AndrejaKo Feb 1 '13 at 13:36

Oil submersion has been commonly used for high voltage applications such as power engineering, for capacitors as well as circuit breakers / relays up until the late 70s, and is still used today, albeit rarely, apparently.The benefits cited by the previous generation of engineers include:

  • Suppression of arcing to a higher voltage than air
  • Improved cooling of local heat spots due to circulation of the fluid
  • Improved insulation between the contact electrodes, having a higher dielectric constant than air
  • Elimination of bubbles within the paper or film, or between electrodes, thus avoiding sporadic arcing and variation in characteristics
  • In specific high or low pressure deployments, oils were preferred due to being less compressible (almost not at all) compared to air. Compression or rarefaction would unpredictably change the distance between, and thus capacitance between, the electrodes.

Vintage audio amplifiers were vacuum-tube based, and often operated at high voltages, thus justifying the use of such high voltage friendly components. However, modern semiconductor based amplifiers work at (typically) much lower voltages, thus ought not have any strong engineering justification for continuing to use oil-filled capacitors. There's no accounting for the spiritual beliefs of audiophiles, though.

Footnote: Oil was later largely replaced in circuit breakers by SF6 gas, which is not flammable, hence eliminating the safety hazard of oil. Vacuum based circuit breakers are also now in use in high voltage deployments. These two options do not appear to be prevalent in capacitors, though.

  • \$\begingroup\$ I like the point you make about difference in power supply voltage with tube based and semiconductor based amps. \$\endgroup\$ – jippie Feb 2 '13 at 10:55
  • \$\begingroup\$ Interesting point. However the people using oil immersion switchgear would regard vacuum tube circuits as "low voltage". The EC "Low Voltage Directive" applies up to 1500V DC. Oil in oil/paper capacitors increases the permittivity, reducing the size of the capacitor. It doesn't circulate to any extent and doesn't play any role in cooling or arc suppression. However its viscosity will control acoustic vibration to an extent; a modern film capacitor driven hard can emit faint sound, like an inefficient electrostatic speaker; if its plates are moving its capacitance must presumably vary too. \$\endgroup\$ – Brian Drummond Feb 2 '13 at 11:59

As a professional engineer and amateur musician, I was intrigued by this question. All I could find on the subject can be summed-up from here ...

Many electronic experts will tell you that there is no scientific explanation of why paper in oil capacitors will give you better Tone when used in a guitar circuit. But it is a well known fact in the guitar community that paper in oil will be warmer, smoother and have more "Sparkle" than ceramic disc, mylar or polypropylene capacitors. The original Bumble Bees and Black Beauties were paper in oil and thought of by many to be the "Holy Grail" of Tone as far as capacitors go.

And from here ...

Paper in oil (or PIO) caps are just one variety of capacitor. Others are made out of mylar, ceramic, etc. PIO caps were used in the Les Pauls of the 50's and many appreciate their tonal qualities over more modern capacitors.

The effect of any capacitor on frequency response & transient response will depend on other impedances in the circuit in which they are fitted so a capacitor can have no 'sound' as such.

But we are straying into highly subjective (and some might say religious) territory here and fortunately, since I am not a guitarist I am quite unqualified to comment further ...


It's the same reason people will build a hi-fi amplifier on an 80 pound slab of granite, avoid op-amps in their designs, or use only 00 gauge pure gold speaker cables: they aren't proper engineers, and they are prone to taking a slice of truth, taking it to excess, and ignoring everything else.

When you read:

be warmer, smoother and have more "Sparkle"


thought of by many to be the "Holy Grail"

that really means:

some people think they are neat, for no particular reason at all.

  • 8
    \$\begingroup\$ As an aside... Many audiophiles tend to gain additional enjoyment knowing that they have invested in equipment that plays sound. The main problems with "warmth" is that they are recreating old distortion effects tainted by memory. Very accurately reproduced music is often discarded as cold, harsh, or mechanical, while muddying the peaks (as tube amps do) and other slight effects of harmonic distortion produce the desired sound. It's not about fidelity, it's about nostalgic fidelity. That's why the older, expensive because they are no longer relevant, parts are in demand. \$\endgroup\$ – Edwin Buck Feb 1 '13 at 15:12
  • 2
    \$\begingroup\$ @Edwin: I would like to see some double-blind studies done with regard to such phenomena; the person arguing that vintage or expensive (whatevers) have a different sound from newer/cheaper ones would be allowed to select the audio source material and listening environment. They would be given ten buttons, five of which would be randomly selected to use a "new" component, and five of which would use a vintage one. If their favored components have a different sound, they should be able to determine which five buttons are different from the other five. Which component sounds "better"... \$\endgroup\$ – supercat Feb 1 '13 at 17:12
  • 2
    \$\begingroup\$ ...is purely a subjective judgment; if someone finds that some quirky harmonic distortion makes a recording more aesthetically pleasing to them, I cannot possibly argue with that, provided that there exists at least some combination of source material and listening environment where they can actually tell the difference. If they can't distinguish between their favored component and a more conventional one, however, any expressed preference would be groundless. \$\endgroup\$ – supercat Feb 1 '13 at 17:17

If you are restoring - or recreating - a historically important amplifier, such as a 1940's Leak TL12 hi-fidelity amplifier or a Vox AC-30 guitar amplifier, it's important to use the correct parts or the closest modern alternatives you can.

At one time, the best and most reliable high voltage capacitors were paper-in-oil. So the original designers used them because there was no good alternative. (I find they tend to go leaky after 40 years or so)

  • \$\begingroup\$ Perhaps, the difference then against the modern types of capacitors is that paper-in-oil capacitors have proven long service-life reliability in the field? When did we have the electrolytic caps in their current format, probably around the 70's? It will be like comparing artwork in oil medium (long history, known aging characteristics) versus the newer acrylics. \$\endgroup\$ – shimofuri Feb 1 '13 at 13:15
  • 2
    \$\begingroup\$ I have some electrolytics from the 1930s, but they tend to be lower voltage, and - even when new - it was expected that they wouldn't work properly until they had reformed on powering up the circuit. "The Manual of modern radio" says "in some of the older types the leakage current is very heavy and it might take ten minutes before the condenser is healed" (from the 1934 edition!) \$\endgroup\$ – Brian Drummond Feb 1 '13 at 14:01

This guy actually took measurements and put up some Lissajous, so at least we can see some data. I see a bunch of straight lines, but he says the paper/oil was "cleanest", but I see a bunch of straight lines (with some notable exceptions).

  • 1
    \$\begingroup\$ I think those curves show that polyester, polystyrene, polycarbonate, polypropylene, mica and paper-oil will be equivalent in low voltage, low bias applications. Guitar tone caps are certainly low voltage (fractions of a volt) and no bias. \$\endgroup\$ – David Johnston Mar 14 '16 at 20:21

There might be some odd acoustic behavior as the foils of the capacitor will experience some force from the voltage across them, with different resulting effects for paper, mylar, etc.

Way back when, I had some oil filled glass body capacitors from a high voltage unit of an old oscilloscope. They were rated for 40 KV or some such. But when I put 60Hz ac across them (probably straight from the power line. It was a long time ago), they blew up. Go figure. I imagine they could not take the 60 Hz vibration.

  • \$\begingroup\$ If a capacitor has high capacitance and high ESR then putting it across an AC supply will dump a lot of energy into the capacitor. That energy has to go somewhere..... \$\endgroup\$ – Peter Green Oct 20 '16 at 1:33

Not the answer you're looking for? Browse other questions tagged or ask your own question.