Paper in oil capacitors

[cotdt]

Quote:

Originally Posted by Crowbar /img/forum/go_quote.gif I'd love to see that test data. It doesn't make much sense, since oil has a very high dielectric constant.

the dielectric absorption has been tested to be pretty low. for teflon it is about 2.1, and paper immersed in oil is very close to that.

there have been many tests that show paper in oil tests better than other film caps. here is a test of D-E curves:

http://members.aol.com/sbench102/caps.html

The paper in oil was more linear than other film caps. So it has lower distortion especially for coupling.

 

[deuginthesky]

Quote:

Originally Posted by cotdt /img/forum/go_quote.gif They would be useful in an amp, but a DAC? what will be its purpose? the analog output stage will only need small value caps in the picofarads, and for those you can go all-out and use air plate capacitors (the kind you see in textbooks). paper in oil in general is pretty good, except for the limited shelf life since the oil dries out eventually. they have very low soakage and distortion, and test better than all other film caps except teflon and liquid teflon. don't buy vintage oil capacitors because the shelf life is limited.

As an example, my DAC is a Zhaolu 2.5, and values on the analog output are 470pF to 100µF caps, with "many" 0,01µF and 0,1µF caps.

For values up to 2200pF, I have my silver mica caps, but for higher values I am considering to use PIO instead of the green WIMA film caps to improve sound.
I don't want to waiste my time (an money BTW), this is why I am looking for advices and tech explanation about avantages or disavantages of PIO in such an application.

Here are the PIOs, those I have are smaller because 200V rated instead of 1000V on the picture.
http://cgi.ebay.com/0-01uF-1000V-PIO...ayphotohosting

 

[cotdt]

I said that because a better circuit for the DAC won't require so many large capacitors.

Those are definately good caps, but I also have a Zhaolu 2.5 and if you click at the link in my sig you can see what I did to mine. The analog stage sucked so bad that I threw it away. Now my new analog output stage only has a single capacitor in the audio signal, which is a 470pf Silver Mica, instead of a series of large low quality capacitors in an inferior design. It is discrete and with this far better circuit I get far better sound. So basically if I wanted better sound I can replace the 470pf Silver Micas with teflons or air plates. Changing the output stage is the best way to go as a first step, IMHO.

I would use those oil caps as bypass capacitors in an amp, or speaker crossovers. .01uF is also a pretty common value in general.

 

[deuginthesky]

Quote:

Originally Posted by cotdt /img/forum/go_quote.gif I said that because a better circuit for the DAC won't require so many large capacitors. Those are definately good caps, but I also have a Zhaolu 2.5 and if you click at the link in my sig you can see what I did to mine. The analog stage sucked so bad that I threw it away. Now my new analog output stage only has a single capacitor in the audio signal, which is a 470pf Silver Mica, instead of a series of large low quality capacitors in an inferior design. It is discrete and with this far better circuit I get far better sound. So basically if I wanted better sound I can replace the 470pf Silver Micas with teflons or air plates. Changing the output stage is the best way to go as a first step, IMHO. I would use those oil caps as bypass capacitors in an amp, or speaker crossovers. .01uF is also a pretty common value in general.

I read quite a lot about the zapfilter which seems to be a very nice product, but I don't want to spend so much.

I am waiting for a group buying of burson audio discret OP amps which are 3 times less enpensive that a zapfilter. I know people using them on hi end products, they really enjoy them.
http://www.bursonaudio.com/Burson_HDAM_Module.htm

Would the K40Y-9 PIO caps be an improvement on the standard analog output ?

 

[cotdt]

Quote:

Originally Posted by deuginthesky /img/forum/go_quote.gif Would the K40Y-9 PIO caps be an improvement on the standard analog output ?

I'm sure that if you replaced all the capacitors in the signal path, you'll get an improvement but doing that with paper/oil would require some humongous capacitors. I've also measured some of the values and the actual capacitors used don't match the descriptions on the PCB. They are bigger values usually. You would also want to replace all the resistors with better ones.

 

[deuginthesky]

I will buy soon a "multimetrix" device, which is able to mesure capacitance among many other things.
As I got 100 0,01µF PIO caps and soon 100 0,1µF PIO, I hope some will match.

What kind of resitors may I use ?

 

[Crowbar]

I recommend the Meterman 37XR. The price is right, it's very accurate, and does inductance as well as capacitance.

 

[deuginthesky]

Thanks, and what about resistors ? Many prentend to be audio grade at many different prices. Which one(s) should I look for ?

 

[Crowbar]

Bulk metal foil (instead of metal film). You can buy them at say Partsconnexion in the Vishay section, $6 to $8 a piece

Vishay have a white paper explaining current noise and why it's lower in metal foil compared to metal film (not likely to make an audible difference unless you have very, very high gain).

I wouldn't waste money on Audio Note tantalums...

For less critical resistors I'd go for Roederstein MK3, but they're hard to find nowaday as they've been out of production for a long time. They're cheap too, like 25 or 50 cents a piece.

In metal film resistors, try to look for low temperature and voltage coefficient (all resistors vary somewhat their resistance with temperature and voltage applied, so they're slightly nonlinear). Sometimes it takes a bit of digging to find these numbers, maybe even an email to the maufacturer if they don't have the info in the datasheet or white papers on their site. Most such resistors would be quite good, the specific brands mentioned above likely being overkill.


How to decide which resistor is critical and will most affect the signal is not always too obvious. Follow the AC signal current loop through the amplifier. In an amp with high PSRR, class A loaded by constant current source, power supply resistors would be pretty well isolated from the signal, and the signal ripple superimposed on their current is very small. Resistors in constant current sources etc. would also have limited impact, as they mostly see just a DC current. Resistors that are often ignored but matter a lot are those parallel to the signal, shunting to ground or some fixed voltage. Guess what, just because it's in parallel and not in series with the signal doesn't mean it doesn't have an effect. For example, parasitic capacitance in such resistors can definitely have an impact on the signal, though on the other hand nonlinearity may not be important, such as in input impedance setting resistors from input to ground. If you're unsure, use a simulator to see the current flowing through resistors and think of how noise, nonlinearity, and parasitics of that resistor will affect the signal.

Is this likely overkill? As long as you're not using the super-noisy, likely-to-drift-with-age carbon composition resistors, it likely would be extremely difficult to have a difference (in blind testing) unless you have a super-high resolution system.

 

[jarthel]

Quote:

Originally Posted by Crowbar /img/forum/go_quote.gif Bulk metal foil (instead of metal film). You can buy them at say Partsconnexion in the Vishay section, $6 to $8 a piece

I believe that these are not the highly regarded S102 which are even more expensive (2x to 4x per piece) than those sold by PCX!!!