Capacitors for specific uses...

[mutterd]

Hey guys,

Im working on some shelving filters for a preamp design and I got to thinking…

Speaker crossovers often use film caps for band pass duties, power circuits use electrolytics for filtering, I have been using ceramics for HF grounding….

so what sort of caps are best for what sort of circuits?

Im not really taking about what brand, and what model of caps sound best…

but for instance - the amp cards I am using are pretty dense with Tantalums, which I dont see in use nearly as much these days - what are Tantalums good for?

For my shelving circuits - everything i need is readily available both in both films and tantalums - is there any reason the films are better…

For my high pass filter, there are also electrolytics in the range - I see them used all the time in the tone stacks of guitar amplifiers, in the signal path and supporting it - how would these stand up in filters?

so what about ceramics?

what about polystyrene caps?

I know this stuff may be covered in a bunch of different places - maybe we could consolidate the ideas here?

thanks so much guys,
Timothy

 

[joaquins]

Warning going there, different opinions on all this...

First, take care speaking about ceramics, there are pretty good ones which are as stable and ideal as you can get a cap for many applications, or as crappy as they can get.

Tantalums are meeean! well, not really, they are much more reliable than electrolytics (and more expensive) and because of the reliability is common for broadcasting specs takes all electros out and use tantalums. Also there are other things on this, but this is one example.

In general terms film caps are not as stable as bests ceramic but they could be bigger in value and very linear, much lower leakage compared to electrolytics, and we like them to be everywhere the signal path is being directly passing through. The difference between different poly materials are smaller but still sometimes matter.

As I said, this is a sensitive topic and I don't want to make much statements on this, as an example THD is measurable in some caps in filters for example, you could probably do it, electrolytics and cheap ceramics will show this for sure. There is an article that may came helpful for this, I can't leave it here cos it's too big, but look for "picking capacitors" from Jung and Marsh.

JS

 

[gyraf]

Tantalums where you don't want changes in capacitance over time, but prefer to live with a full short at eol.

E.g. the timing capacitors in the ssl bus compressor.

Also they show lower leakage than most electrolytics (in the 90'es), so you occasionally see them as dc-blocking/pot protection capacitors in large mixers.

Only stupidity puts them in power supply decoupling after 1985

+1 for  "picking capacitors"

Jakob E.

 

[ruffrecords]

Capacitors can introduce distortion and different types introduce different amounts. In all case, the amount of distortion introduced depends on the ac voltage across the capacitor. This has led to a lot of pointless uninformed debate amongst audiophools about the best caps to use for coupling between stages.

The bottom line is it is the ac voltage across the capacitor that is important. as a rule, film caps produce much less distortion than electrolytics for a given ac voltage across them.

In a coupling cap between audio stages, the idea is to pass the ac signal with as little loss as possible. To do this, you generally choose a value that has a very low ac reactance at the lowest frequency of interest. This means that at all frequencies of interest there is very little ac voltage across the capacitor so the distortion it introduces is incredibly small. This means you want to choose a large value capacitor which means you probably need to use an electrolytic or tantalum. Although these produce more distortion than film types, the ac voltage across them is so small it does not matter.

The only place where capacitor type really makes a difference is in EQ. Here we deliberately want the reactance of the cap to be significant at the frequencies of interest so by definition there is a significant ac voltage across it. This means the cap can introduce distortion. If you want your EQ to sound clean then use low distortion types like film or polystyrene. You don't usually see electrolytic or tantalum types in EQ. Ceramic are good for decoupling op amps  because of their high frequency performance. They are not normally used in EQ as they can be microphonic.

Doug Self has a good section on this topic in chapter 2 of  his book Small Signal Audio Design.

Cheers

Ian

 

[JohnRoberts]

To expand upon Ian's post and answer more of your question the optimal cap depends on the application.

Power supply reservoir caps are all about how many farads fit in a given volume, ripple current, and perhaps ESR making electrolytic preferable. . Power supply decoupling caps are often about HF impedance making ceramics effective. Where the rubber meets the road, for audio filters film capacitors with low voltage coefficient are useful. DC blocking is generally about pushing the pole frequency below the actual passband so they do not experience terminal voltage change so electrolytic are popular for the way low poles, with one film cap establishing the dominant HPF. .

Of course this is just broad strokes and some very good articles have been written on the subject, and a few exceptions, along with some absolute rubbish.  The larger the differences claimed by the audiophools, the less you should listen. (IMO)

JR

 

[CJ]

surplus stores can have good deals on caps,

plug and play, good ear training,
have a friend connect them so you can do blindfold tests,

Phillips has a good rep for Neve lytics,

those flat Wima types pick up noise, but can sound good in an eq,

some crazy people use oil filled caps for pwr amp supplies,

don't care for mica caps that much,

Vitamin Q's can be nice but mat contain PCB's

that red ERO cap in the bottom left is my favorite for coupling caps in git amps,

do not hear anything amazing about expensive Hoveland Musicaps,

anything with epoxy end caps and polarity lines will sound nice, 

 

[mutterd]

Wow, thanks guys,

Thats really great.

I had found the Jung article when I was digging around on this subject - good info there for sure, its what got me questioning tantalums in the first place…

Gyraf - that makes total sense - it also makes sense why they would be in these cards - word is that designer was pretty hard line about his circuits not drifting over time...

The only place where capacitor type really makes a difference is in EQ. Here we deliberately want the reactance of the cap to be significant at the frequencies of interest so by definition there is a significant ac voltage across it. This means the cap can introduce distortion. If you want your EQ to sound clean then use low distortion types like film or polystyrene. You don't usually see electrolytic or tantalum types in EQ. Ceramic are good for decoupling op amps  because of their high frequency performance. They are not normally used in EQ as they can be microphonic.

Doug Self has a good section on this topic in chapter 2 of  his book Small Signal Audio Design.

Cheers

Ian

and for my immediate purposes, this is exactly what I was looking for - thanks so much Ian.

JR - aside from the 2 articles mentioned here - any other readying you could recommend?


and for reference - here is the Jung article referenced above:
http://waltjung.org/PDFs/Picking_Capacitors_1.pdf

off to re'read the capacitor section of Small Signal Design now…
thanks fellas.
Timothy

 

[JohnRoberts]

I won't speak ill of any specific articles, but the one I've heard good reports about are the Cyrus Bateman series in Electronics World (that I didn't actually read completely). 

I wrote about capacitor behaviors briefly in my 1980's audio mythology column. There was some real capacitor nonsense going on back in the '70s-80s (and still a little now), but modern caps are better.

JR