Silver Mica caps VS Polystyrene caps

[KevinLee]

Hello All,

Is there an advantage to using one type over the other in feed back circuits or in a CD player output stage? What differences can one expect sonically?

I know that it is easier to get polystyrene in larger values and that silver mica tends to cost a bit more.

I have the choice of these two caps when doing upgrades and have next to no experience using these two types of caps.

Any suggestions or experiences appreciated.

Cheers

KevinLee

[Romy]

I am a person who is VERY far from screwing with electronics but “accidentally” I have explored exactly this qestion within one of my phonostages with a full RIAA in feedback. I used the different types of polystyrene, different types of polysterol, the different type of micas, two type of air caps and so on, so on and so on… Ironically the cheapest 12c genetic no-name mica killed absolutely everything with VERY strong margin….

So…
The Cat

[mcp]

IMO polystyrene is neutral whereas silver mica is very revealing. It can be difficult to work with sometimes as it brings out the harshness in some designs. But when you get it right, the music sings. It has that sparkle, especially the highs.

In my amplifiers, I use silver micas exclusively for feedback and miller comp. For hf limit on input, polystyrene or silver mica.

[KevinLee]

Thanks for the input so far,

I just swapped about 8 caps in my integrated amp to mica from polystyrene and look forward to hearing the difference.

Do silver mica caps have a long break-in period?

Thanks

KevinLee

[haldor]

Silver mica caps biggest claim to fame is how stable they are (ie their characterists don't change with age/temp etc) . If anything is not going to require "break in" it will be silver mica caps.

Phil

[sam9]

What exactly is the function of the cap? That makes some difference. In any case, audio signal cannot pass through any siver mica cap I'm aware of. The highest value is something like .01uF or less and that essentially blocks any audio signal. Generally to pass full audio bandwidth you need a value of 1uF or higher.

The point of all this id that siver mica or polystyrene can have an effect of the quality of the audio signal, but indirectly. Perhaps they are used as power supply bypass or possible compensation caps for an opamp.

So it's a least in part a question of "what function do they perform in the specific circuit?" and not merely "what flavor of cap do you like?"

[andy_c]

Originally posted by sam9
(...) In any case, audio signal cannot pass through any siver mica cap I'm aware of. (...)

Ahh, but you didn't consider frequency compensation and the Miller effect. It's not uncommon for a power amp VAS to have a low-frequency voltage gain of -10,000 (if it's cascoded and has a high-impedance load such as a MOSFET driver). When the mica cap is connected from input to output of such a stage in its compensation role, the capacitance seen by the input stage is multiplied by 1 minus this gain (Miller's theorem), resulting in a very large effective capacitance value. It's not uncommon for the open-loop -3 dB frequency of the amp to be not only within the audio range, but even below 20 Hz for some designs. So thanks to Mr. Miller, the mica capacitor is very much in the picture.

[sam9]

True enough, I was thinking mostly about possible functions in a CD player per the orginal post. Unless it's something exotic that uses discrete opamps, I don't think Miller capacitance is likely to be of concern.

In the case you mention, it remains unclear to me if factors other than the capacitive value and the bandwidth of the cap are really that important. Despite being unclear, I use silver micas for this purpose as I don't think any of the "polys" have enough bandwidth to be sure c/l gain goes negative before the accumulated phase shift reaches 180 deg. In principle, an NPO cermaic would do as well but since there are only a couple needed per amp why not take the "Caddillac" so to speak?

[AKSA]

Silver mica is definitely the go.

I've tried all sorts of caps in this role, and it has profound effect.

Cheers,

Hugh

[Circlotron]

If the voltage across the cap *varies* as in an active filter, (with a coupling cap, the voltage across it should not vary if it is big enough, so "cap effects" are a bit suspicious ) then you want to use a cap that doesn't change it's value with applied voltage otherwise it will introduce nonlinearities. With ceramic caps, NPO = pretty good, X7R = awful, Z5U = tragic. I did an experiment once with a Z5U, and normally the energy in a cap increases as the square of the voltage (double the applied voltage = 4x the stored energy) but these Z5U's decreased their capacitance at such a rate that the stored energy was *directly* proportional, not a square law thing.

Anyway, one of silvered mica's claims to fame is very low losses at high charge/discharge frequencies. e.g I have many times in the course of work seen a Cornell Dubilier cap 20x20x8mm handling 2000 VA's @ ~450kHz wile only getting slightly above skin temperature. That kind of treatment sends most other capacitors to cap heaven VERY quicky.

Whether micas has any useful audio properties I don't know. One things for sure though, if you wanted a cap that would still be useable 10,000 years from now, mica would be a good bet. Very stable material.