Has anyone actually MEASURED anything that shows the superiority of paper-in-oil capacitors over polypropylene?
superiority in mass? use a scale! dimensions? use a caliper! so what "superiority" are you talking about? 😛
Many people have done such measurements (including myself): among the non-Ecaps, PIO is actually inferior in all respects, compared to any other dielectric except plain paper.
Even the crappiest metallized mylar easily outperforms it on DF, distorsion, etc
Pretty hard to tell much difference in those scope photos; except, of course, the ceramic looks terrible. Perhaps a more quantitative analysis would be in order.
Well, I was hoping for some defense of their outrageous price ($17.95 each) in terms of dissipation factor, distortion, stability, or double-blind testing of their sound superiority.
My friend who bought some to put in his lap steel guitar thinks they are WAY prettier than orange drops.🙄
The location in a circuit is said to mean something. Some knowledgeable people say that any decent cap is good at the input, for instance, but I'm not an expert.
FWIW, I agree with Elvee's comments wrt PIO capacitors and their relative merit.
There is another potential issue with leakage current increasing as they age.
Subjectively in HIFI I thought in most cases they basically sucked. In an MI application the criteria are different.
There is another potential issue with leakage current increasing as they age.
Subjectively in HIFI I thought in most cases they basically sucked. In an MI application the criteria are different.
Here are a few measurement examples:
Attachments
Last edited:
I see 5 thumbnails in post13.
All open to show the details written beside each capacitor.
What are the % values indicating?
Edit:
former posts 14 & 15 have disappeared and took my post16 along with them ?
All open to show the details written beside each capacitor.
What are the % values indicating?
Edit:
former posts 14 & 15 have disappeared and took my post16 along with them ?
Just cleaning up the redundant posts.
Percentages and ppm's are the loss factor (tan δ)
Losses by themselves aren't non-linear, but they are a "vehicle" for non-linearities: although possible theoretically, I have practically never seen a very low-loss cap presenting a significant non-linearity.
This doesn't imply that all lossy caps are non-linear, but it is normally one of the preconditions.
Losses are also an important indicator of other "quality" parameters: for example, DA (which BTW is a linear phenomenon) is incorporated in the loss figure, and deviation from the intrinsic dielectric value: for example, PET taken in isolation has around 0.3~ 0.35%.
If you find a PET cap having losses much larger than 0.4~0.5%, this means that its construction is supoptimal or that it has aged in a less than graceful way.
A very useful diagnostic tool: this can be seen on some paper caps, which are known to have a problematic ageing, and have gone through the roof
Paper is a fairly useless dielectric. Made sense before plastic was invented, but not now. Most people routinely remove paper caps from old equipment because they are (or will become) leaky.
It sounds like you're assuming the PIO caps are superior. That may not actually be the case.
I compared a $0.50 Panasonic EF-series polypropylene cap to a $2 Solen polypropylene to a $20 Silver-in-Oil Mundorf. I measured the impedance of the cap with an HP4194A impedance analyzer and worked out the ESR and ESL with the "equivalent circuits" function of the instrument. The Mundorf lost on all accounts (lower SRF -> higher ESL; higher ESR). The Panasonic won (lowest ESR and highest SRF). The Solen was slightly worse than the Panasonic but not by much. I suspect the difference was caused by the difference in lead length as the Solen was an axial type and the Panasonic a radial type.
All caps were 2.2 uF. A common capacitance for a coupling cap.
I recognize that the ESR/DF and ESL only tell part of the story. Dielectric absorption will play into the sound quality as well and is not captured in a simple impedance sweep. Dielectric absorption should impact the THD, though, so with a distortion analyzer one could measure it... Douglas Self did this and reported the results in his Small Signal Audio Design book.
I reached the same conclusion. In an audio circuit, you're dealing with errors in the ppm range. You won't be able to pick that out on an oscilloscope.
The ceramic cap looks horrid, but the type of ceramic is not specified. A quick search on Digikey or any of the capacitor manufacturer's websites reveals that there's no such thing as a "one size fits all" when it comes to ceramic dielectrics. The voltage coefficient of the dielectric depends on a wide array of parameters, including the physical size of the cap, so you can't even recommend a specific dielectric as trouble-free. The only exception would be NP0/C0G. Even then I'd be hesitant to use those as coupling caps without characterizing their performance first.
Similarly, a recommendation such as "all electrolytics are bad" does not hold water either. Forum member twest820 measured the THD of a bunch of Nichicon Muse ES series capacitors in a coupling cap application using my Audio Precision APx525. He and I both concluded that these caps did not cause any measurable distortion in this application. It follows that they'd be sonically transparent as well.
Tom
- Status
- Not open for further replies.