Capacitors - Which type to use

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.

I'm currently considering building a phono preamp, I have some design ideas but I'm unsure of the type of capacitors to use in the RIAA equalisation. I expect to be using some caps in the 10nf to 150nf range in the feedback networks, but what type? Polyester, polystyrene, polypropylene, ceramic?

I'm assuming stability is the most important criterion?

Any ideas?

Polystyrene is excellent, and is available in values up to around 100nF as +-2% parts. Larger values are available, but tend to get expensive quickly, and are generally +-5 or 10%.
Don't forget that you can parallel caps (or series them, for that matter) to get other values.
Polypropylene is nearly as good, and is available in larger values. The problem here is that you can't always find the smaller values without having to go to the expensive boutique parts. Tight tolerances are somewhat harder to find.
Polyester is sliding somewhat down the scale, and ceramic...well, if you were making a stereo for your pet hamster, it might do in a pinch, but you'd have to hate the hamster to treat him so badly...

My first choice would be polystyrene. However, polystyrene caps are getting rare, especially in values needed for lower impedance RIAA networks. Paralleling works well, though, and you can play games with statistics to get good accuracy for the network. My next preference is film-foil (as opposed to metalized) polypropelene.

Make sure you match values in left and right channels. To me, this is more important than getting 0.0005% accuracy on the absolute value of the cap.

I'll leave the discussion of whose cap is best to the folks that have played with more types than me :) Some years ago, I had the good fortune to have access to a network analyzer and found a stash of some polysterene caps, some film-foil polypropelene caps and a few teflon caps that acted very much like an ideal capacitors out past 500KHz. I can't remember the mfr names for these parts, but it wasn't any of the exotics. Anyway, these have worked for me for years.

I'm thinking about a new RIAA preamp myself. I want a discrete this time. Anybody remember the old Leach JFET based design?

Good luck!
Ha! I have a Leach JFET preamp sitting in my project box since
forever. I ended up buying a Hafler SE100 all-JFET preamp at
a good price and never looked back.

One of these days I might finish it. His designs ran at+/- 24
volts and the Jung regulator can be modified to operate up to
that voltage. That's the power supply I'd use nowadays.

I remember the Leach preamp with fond memories. I built the original bipolar-input design shortly after Leach's article came out in Audio and remember later changing all the resistors from carbon comp to metal film one Christmas day. It was replaced in my house a couple years later by the Marsh design from Audio Amateur, then others. Why do you ask? Thinking of building it? I just went to Leach's website to take a look at it for nostalgia's sake. I wonder why Leach never used a fully-complementary output stage.

It'd be easy and cheap enough to build that I'd like to build it again just to see what I think of it today, that is if I wasn't already swamped with 2 big amp projects and a long-neglected list of mods I want to try on my present phono preamp.

I share your preferences for RIAA caps.

Marlow, good luck with your preamp. And like Grey Rollins said, avoid ceramics for RIAA and for frequency compensation. They are truly terrible in these applications and can exhibit piezoelectric effects, especially when hit with a hard transient.

Some years ago, Ben Duncan designed a preamp for Hi Fi news magazine, he used Eurofarad caps (don't know what type) that had the opposite tempco to the holco resistor he specified, that way the time constants didn't wander about.

Something to think about.

Consider dielectric absorption

Hello Marlow,

you should consider the effect of dielectric absorption. This is, some electrons enter the dielectric and come out again delayed, leading to some regain of voltage after taking out charge. Although a rather slow process, this may lead to signal errors, especially in amps without overall feedback.
Polypropylene, polystyrene and teflon are good, polyester and NPO ceramics are a little worse, and standard ceramics or electrolytics are very bad.
Besides of that, standard ceramics (X7R) are indeed piezoelectric, which will lead to further distortion. You should never use them in an audio circuit.

For more information, see:,1570,28,00.html

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.