Which Capacitors for Audio?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Junction

Well-known member
Joined
Mar 23, 2007
Messages
427
Location
Sydney, Australia
??? There is a myriad of caps out there, I really don't know which ones are most suited for the various diy projects here, can someone help provide some idiot guidance on which are the better quality film type caps for audio use?

For example have been looking at Wima caps and there are various types, i.e MKS2, MKS4, MKP2, MKP4, MKP10 and probably more. Then there are the yellow MKT type and for tube circuits there are the yellow polyesters, orange drops etc.

Thanks

 
This is a massive subject. There are good articles on the web.

As a very vague guide, I'd tend to go for non-metallized rather than metallized when buying film caps.
 
And non-metalized basically means film/foil?
Could you maybe name several types?
 
Or you could go with PeterC's advice, "I shove in whatever I have because I'm not a capacitor snob."

I thought that to be very funny and he designs some great stuff.

WIMA stuff is more than fine.

CC
 
I had the same question when I got into all this stuff. I have discovered that the type of cap offers more variety than the brand of a cap. Such as putting a polypropylene cap inplace of a ceramic can give slight tonal differences. Good old Panasonic FC's and FM's are cheap very good electrolytics. Nichicons are also well regarded. No need for black gates!

I don't know the answer to the poly film question. I just use whatever. I have WIMA's all over one channel of my LA 4 (MKS I think), but then in my other channel I have some panasonic polyester films because I had them on hand. Don't notice any difference.

What I want to know is resistors. Does it really matter if I use the bags of 10 for $0.50 from the local store? Do Vishay dale/Xicon resistors make that big of a difference? I know that tolerances come into play here, but what about noise? 

 
conleycd said:
I had wondered the same thing about resistors.  I had purchased a ton of resistors from ebay - chinese ones I think.  Leads are little wimpy but - between JohnRoberts and PPR I could assume "don't worry about it."

Here's the thread:

http://www.groupdiy.com/index.php?topic=33304.0

CC

You should worry about everything but not obsessively.  :eek:  And I suspect there are bottom of the barrel modern parts that are worse that old school middle of the road parts, but you'd have to work to find the real cheap crap as a DIY in the west.

Re: Resistors, yes they are generally benign but some cheap small ones had enough voltage coefficient that you could see their distortion in high voltage swing applications (like feedback networks across a power amp.

With caps, one rule of thumb is how much terminal voltage do they see in normal used. Not how much voltage does one terminal see, but how much does the voltage difference between the two terminal change in use.

A capacitor in a DC blocking circuit will see very little terminal voltage in normal use, while caps that make a tone control or equalizer will be changing a lot, and changing in the middle of the audio band where our hearing is most sensitive. So invest in good caps for EQ, blocking applications not so much.. 

Of course this is a slight oversimplification but will help you major on what's most important.

JR

 
So invest in good caps for EQ, blocking applications not so much..  
But this is the problem, one wishes to invest in decent quality caps, but how does one know which ones are decent quality ...
have been looking at Wima caps and a few types I have found are MKS2, MKS4, MKP2, MKP4, MKP10 and then there are the yellow MKT type. So which ones would be most desirable??
 
Anyhow, I was asking which wimas are film/foil type, so it's evidently MKF.
Now, between MKS and MKP, go with MKP, since they are polyprop, although metallized.
As for sound properties, russian paper capacitors are great and relatively inexpensive - compared to boutique stuff, Jensen and such. Maybe not so much with blocking caps, but in speakers crossovers I can defenitely hear polypropilene caps, both cheap and expensive, they seem to give some edge to the sound, and paper caps sound smoother. For ultimate clarity, there's also russian teflon caps, although they are large and only come in small values, but again they cost almost nothing compared to brand name teflon caps with outrageous price tags. Notice, you have to have some reasonably hi-end monitoring to hear those differences. Also, if you trying to build something, just go ahead and order what you can see, like, panasonic or vishay caps are really no better or worse than wima. Overbuilding and boutique parts are art in itself, bordering on acute audiophilia, where capacitors are hand rolled from rice paper by virgin maids, and hookup wire is cryogenically treated at Telefunken factory in 1939....
;)
 
yup both are good, i think both are...smooth in an EQ, with the MKP
a little, umm...creamy sounding.

what do you want?  what  'we' like would vary from person to person.
something super clean?  maybe just a little texture?

look at the spacing on the leads, common sizes are easier to change in and out
to check out differences.

i wouldn't go too crazy on real esoteric stuff, but that's just me, but if i had the
bucks and the time...maybe a little.

i usually use wima mkp for no good reason other than legends of old and the sound
of stuff i see them in.    i see mkp in those places too.
i guess i like polyprop the most (among 'conventional' caps)
polystryne in the high bands can be pretty nice.  polyester is good too.
try a few.  that's the only way for you to really know.
all those specific models you listed are quite decent. even quite good.

me, i don't like those little shiny green kinda flat, kinda like chicklet things in EQ...
i dunno, mylar? ain't that just polyester?    but that's just me...
&  i'm just another guy buyin' donuts.
 
and now that the standard stuff is out of the way...

I recommend to see what the Soviet Union military has to offer. If you were building point to point or some new design and have no space constraints there are magic keywords in ebay like "russian PIO" or "PIO cap". They are mostly the same price, but often even cheaper than modern polypropylene caps. Without exception they sound better than film caps and will take nearly any abuse. Just beware, they are 2-3X bigger than polyprops of equal rating.

also beware of US resales and rebranding of this stuff. These guys know to charge 10X the price.
 
If working on solid state, i dunno ... With toobz , you should have a look at FKP series , from WIMA and ROEDERSTEIN : Theyre FKP1832 are famous !
Styroflex are very fine in EQ, and maybe the best ever made . But beware of LOW WORKING VOLTAGE and LOW TEMPERATURE soldering process  ;)
 
Junction said:
But this is the problem, one wishes to invest in decent quality caps, but how does one know which ones are decent quality ...

You really won't know till you hear them in a specific application. Your ears will tell you so much better than any online post.

I don't recommend stocking up on any one kind of cap if that's what you're thinking of doing. I've changed my capacitor preferences too many times over the past decade, and you will probably too.

In my experience, the difference is much more audible with electrolytic caps than film caps. Try building two identical channels of something with a lot of electrolytics in the signal path, like a green pre, and compare a couple popular brands.  It's pretty eye opening. I found some much raved about caps much less pleasing than others.

If you feel like getting obsessive, here are some good articles:

Picking Capacitors
Capacitor Sound?
 
Or you could go with PeterC's advice, "I shove in whatever I have because I'm not a capacitor snob."

Did I????? Dont remember.....

I generally go with Wima or any quality Polyester, Polyprop is bit brighter & polycarbs a bit warmer.

I would not stick in a Y or Z series ceramic except for PSU bypass, but I only buy NPO or COG ceramics anyway.
 
One thing that I have learned is to keep silver mica's out of the signal path. I used one once in front of the IC in a tube screamer. Talk about a cold sounding pedal. Another version I made had an NP0 ceramic disc in that position which I much preferred.

For ceramics, C0G/NP0 seems to be the only way to go. It makes me wonder though. What are the X and Z series ceramics good for? All I have read is to stay away from them. 

How about tantalums? I have seen several pieces of neve inspired gear that have a Tantalum mod offered. What is the advantage with tantalums over lytics?

Those russian PIO's are freakin huge! I saw the build pictures that you posted up kingston. Was it your MILA? Anyway, it looks like you need to plan for PIO's.
 
Junction said:
So invest in good caps for EQ, blocking applications not so much..  
But this is the problem, one wishes to invest in decent quality caps, but how does one know which ones are decent quality ...
have been looking at Wima caps and a few types I have found are MKS2, MKS4, MKP2, MKP4, MKP10 and then there are the yellow MKT type. So which ones would be most desirable??

OK, blocking applications where the pole frequency is set well below the audio passband, there should not be significant terminal voltage to excite mechanisms like dielectric absorption or voltage coefficient, so the focus needs to be on non-ideal (not perfect capacitance) terms like ESL and ESR.

I like to think of this in terms of how much work the cap is doing (not the proper physics term), by thinking about how much current the cap is passing.  From simple inspection a capacitor in a passive loudspeaker crossover is doing a lot of work, a 10 uF cap loaded by a couple hundred K, not so much.

The bottom line is this depends somewhat on the application. I am not a huge fan of simple listening tests, but prefer to answer questions like this with, how much does the circuit deviate from ideal? And how nasty is that deviation? Null tests say compared to a buffered (closer to perfect) film cap, DC path, or some such strategy can help quantify and quality that non-ideal error term. (note the DC comparison null will include some ideal C terms). 

I am inclined to ignore claims of night and day differences for simple (light lifting) blocking applications. OTOH there are a few circuit nodes where electrolytic caps may be found working relatively hard.  Probably the most common these days is in series with mic preamp gain pots. At high gain that cap could be driving a few ohms. While not as bad as a speaker crossover due to low voltages involved, the current is enough to drive the connected feedback network resistors to full scale, so not zero current.

It has been decades since I did any bench testing in this regard, but I would warn that null testing will also reveal phase shift caused by ESL, so analysis of the null residual needs to be thorough. 

have fun...

JR
   
 
Insomniaclown said:
How about tantalums? I have seen several pieces of neve inspired gear that have a Tantalum mod offered. What is the advantage with tantalums over lytics?
Tantalums are smaller and don't dry out like aluminium electrolytics. They also have zero tolerance for reverse voltages, so you can't use them for ac signals unless they have a large enough dc bias to prevent that. They are unreliable in low impedance (high current) applications because they tend to go short circuit. They fail regularly when used as decoupling caps on voltage regulator outputs, for example. They are very reliable in high impedance circuits with proper dc bias however. Neve stuff from the '70s is full of them, and most of them still work.
 
While tantalums are rather unloved* for audio path use, in my one bench test for this a few decades ago (in the gain leg of a phono preamp operating into 360 ohms), I observed significantly less phase shift at 20kHz with same value tantalum, no doubt due to lower ESL compared to aluminum electrolytics back then. 

JR

**Tantalum gets a bad rap for audio use due to terminal voltage effects like dielectric absorption. While I am not convinced that DA is a big deal in DC blocking application, for that and several other reasons I see little compelling reason to use them. The only tantalum cap I ever specifically designed into a circuit, was in an attack/release circuit in the decoder for an old, and quickly obsolete, consumer record NR system (CX) from CBS. I learned that the encoder used for mastering CX records used a tantalum cap in it's time constant circuit, and IMO the DA that tantalum was disliked for, could actually matter in an attack/release circuit because of the asymmetrical charge/discharge impedance. Why "they" used tantalum in such an application I will never know, but I tried to match the response in all it's ugliness for accurate playback (ignoring the fact that CX was level threshold dependant, so odds of accurate playback were about zip. 

*** Tantalum for non audio use gets a bad rap for catching on fire... modern low impedance aluminum, designed for use in switchers, will cover most such applications with less risk of releasing smoke.
 

Latest posts

Back
Top