Interesting reading this. I have .1uf FT-3 caps bypassing 2.2uf Sonicaps on my Eminent Tech. Planar tweeters and it sounds grainy almost MP3 like on high hats and cymbals. Started happening recently. I wonder if it is the caps. I have been trying to figure out the problem with equipment and cable changes but keep coming back to the speakers. Maybe there is something here!
I'd never recommend paper capacitors as they will degrade with time. If not hermetically sealed they will absorb moisture and change value, become lossy and leaky. And paper will oxidize over time anyway. Polypropylene will do nicely, doesn't need to be hermetically sealed either, will not go leaky, and most are self-healing even.
Any suspect old capacitors should be tested for leakage at the voltage they will be used at.
Any suspect old capacitors should be tested for leakage at the voltage they will be used at.
First time to quote myself. Ha! Have to take back what I said about the FT-3 caps. They sounded like MP3s on the highs because my Moode software was set to run Tidal via MP3. Duhh. Homer Simpson move on my part. Now set back to stream Flac everything sounds excellent.
In absolute terms, correct.
However, in practice they last (or should) about a couple of hundred years.
After all we have books made of primitive paper that are a few hundred years old, and still in good condition!
The only capacitors that I ever seen fail - have been powersupply caps, or caps where they work in high power and in hot conditions.
Paper in oil caps in Audio, are almost always used for by-passing where there is negligible power involved.
For me ( In a valve/tube amp) a K40Y sounds better than any cap. ten times the price and on a par with the best of the exotic ones.
I use them exclusively and none have failed under constant few hundred voltgaes, nor faded or changed capacity.
Old books don't have to have low leakage and low dielectric loss do they? Nor do they have to have a constant low moisture content, nor do they get subjected to MV/m electric fields.
We don't use paper as an insulator any more these days for good reason. For instance impregnated paper used to be used in transformers and motor windings, now we use nomex (a relative of kevlar). Nomex "paper" is pore-free, tolerates high temperatures without degrading, is a very good insulator under high electrical stress for extended periods.
Similarly in capacitors we moved on from paper to other dielectrics which are superior in every respect (except perhaps recyclability).
We don't use paper as an insulator any more these days for good reason. For instance impregnated paper used to be used in transformers and motor windings, now we use nomex (a relative of kevlar). Nomex "paper" is pore-free, tolerates high temperatures without degrading, is a very good insulator under high electrical stress for extended periods.
Similarly in capacitors we moved on from paper to other dielectrics which are superior in every respect (except perhaps recyclability).
hate to drag this any more than necessary.
As I had indicated , in absolute terms, you are correct -
however in practice , the Soviet PIO capacitors in question, do not go out of spec. in a few tens of years.
The are glass sealed, well made and measure perfectly according to what they were designed for, after 50 years.
By no means did I compare them to new technologies or new exotic materials nor for use in anything other than audio amplifiers.
After all - this is "DIY-Audio".
Last edited:
I have only tried such a thing in powersupplies, never in direct coupling or decoupling.
My theory is this:
Different capacitors have slightly different characteristics, usually their characteristics vary gently and smoothly with frequency, this is acceptable to the ear. If I parallel two different capacitors, then overall change in characteristic of the combined capacitors may not be smooth against frequency.
I usually stick to one.
In powersupply smoothing, to filter out RF I may use an additional plastic capacitor in parallel with the Electrolytic ones, it is a compromise.
I've rarely seen any deterioration on sealed paper caps (K40, Sic-Safco, Gudeman, Vitamin-Q, Westcap, Hopkins... ) IF unabused (overvoltage, overheating) , and yes , they are VERY, VERY nice on crossovers
All the sealed Russian caps test perfect in my experiance, zero leakage, even beyond their rated voltage. The MBM's are not sealed, and are almost always junk so ignore those.
None of the Russian caps have a direction marked (some people claim they do), but they dont. I always test them for myself on an oscilloscope to determine the outer foil.
None of the Russian caps have a direction marked (some people claim they do), but they dont. I always test them for myself on an oscilloscope to determine the outer foil.
Correct.
I do not recall them having any direction, nor any of the leads are connected to the casing.
I even used a 200V one in a 300V situation (by mistake ) for years, without any issues.
They are magic, best sounding signal caps for the price (and many times over).
