Polarised or not?

[Neil Purling]

I have been reading an account of the restoration of a Leak TL12 Point One.
It mentions that the smoothing block was not polarised originally and the HT power supply had been changed by persons unknown by using some polarised electrolytics.
What is the virtue of a non-polarised capacitor used to smooth ripple on a HT line?

[G6Tanuki]

I'm not familiar with the particular design but non-polarised smoothing capacitors are likely to be paper-in-oil types whereas the polarised ones are electrolytics.

PiO are invariably bulkier (in terms of capacitance-per-unit-volume) and more expensive (Farads-per-pound) than electrolytics of the same working-voltage; this is the reason electrolytics are so popular. The downside of electrolytics is that they can lose capacitance with age, and they may 'leak' electrically (a problem with older types if the equipment was used infrequently - not an issue with modern ones).

Assuming the parts used are reputable, unless you're an appearance-over-performance type I'd not fret about the replacement of an old non-polarised smoothing cap with modern electrolytics.

[Silicon]

The non-polarised 'film' capacitors usually have a lower ESR than electrolytic capacitors and therefore, in theory, better capacitors for reducing ripple.

[G8HQP Dave]

The lower ESR is likely to have more theoretical value than practical value in this application.

[Radio Wrangler]

Back in the day, oil filled metal cans with paper-foil capacitors were expected to be longer lived than electrolytic capacitors, and they were right (mostly!)

High voltages and moderate power levels meant that paper-oil parts could give enough microfarads to do perfectly adequate reservoir and smoothing functions.

Then along came modern trends in loudspeaker design, making them much less efficient. Along came trends to music with much higher listening levels. Much much more power was wanted, and thus more microfarads in the power supplies. PIO would have been too big and too costly, so electrolytics got used. Over the years, very good electrolytics have been developed.

Theoretically, the lower ESR of a good PIO part (they're not all good) has advantages over the higher ESR of a typical electrolytic part.

BUT there's another theory (also true) that the ESR acts to damp high frequency resonances and instabilitities. Replacing electrolytics with super duper ultra low ESR capacitors has been known to make some circuits misbehave.

This is audio stuff we're talking about and the whole audio field is notorious for people having fads and clutching at any theoretical justification real or imagined... it's all out there. What is needed to navigate all the crap is a sense of proportion, of knowing which effects, properties and theories are of a significant magnitude and which aren't.

There are some perfectly satisfactory amplifier designs using electrolytic capacitors. The art is to use good parts in circuits designed around them to carefully use their stress capabilities and to handle those factors related to their wear-out mechanisms.

David

[Neil Purling]

I was wondering after trawling on-line and finding some Russian K73P-2 block capacitors.
Whether they are PIO or not I don't know. There appears to be some question over the nature of the dielectric in these.