Coupling cap physical size matters ?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I have gotten a few different 2.2uf coupling capacitors to play around with yesterday and I feel that the physical size of capacitors does matter even with same or comparable values. I did search the forum but most discussions are on PS caps. Here's what I have :-

JLH 1996 dual rails class A amp
Tannoy MKII speakers
ipod touch line out as source

Cap 1) West-Cap 0.68uF 50V
Cap 2) Siemens 2.2uF 100V MKL (CELLULOSE ACETATE)
Cap 3) ERO 2.2uF 100V MKT1813

* I never have any WIMA on hands to compare.

At first, the input cap was two West-Caps in parallel. Then I changed them to Siemens. Finally tried the ERO. Ended up back to West-Cap three in parallel (to make up to about 2uF).

The West-Cap is the largest in terms of physical size with the lowest voltage. Siemens and ERO are about same size but different dielectric materials. The Siemens are actually very good but gave me a somewhat different sound that I'm not used to. I can't say its good or bad but just different and I don't know how to discribe it. The ERO gave me the familiar sound of the MKT type caps but the bass seemed a little weak. The tripled West-Cap has the best warm sound, liquid feel and the details I like.

Overall, the physical size of the "West-Cap array" is much larger compare to a single Siemens or the ERO. I also noticed (from pictures) the highly regards WIMA are also physically large as well. Therefore, I want to conclude physical size does matter.

What do you guys think ?
 
If you're going to use a capacitor in the signal path, go with a film type (usually starts with 'poly', as in polyester, polystyrene or polypropylene). Personally, I prefer direct-coupled but that is not always feasible (especially if you are using a single-ended supply).

For capacitors in the DC power supply, lean towards units with low ESR and ESL ratings (ESR rating is dominant for most amplifiers, but also consider ESL if you are using Class D) for the 100uF and higher valued caps, plus some tantalum in the 50uF range if the large capacitors are 1000-2000uF or more, and be sure to bypass them with ceramic caps in the 0.1 - 2.2uF range.

Generally speaking, for a capacitor of a given type, capacitance and voltage rating, the larger cases provide better service then the smaller cases. Conversely, for a given capacitor case volume, a unit with a lower voltage rating will perform better (just so long as it's not too low).

For 100uF and larger values, electrolytic capacitors are usually selected. They tend to wear (or dry) out after several years (typically 10, but could be less depending on the manufacturer), so they have to be replaced. As an alternative, you can get film capacitors in the 1000-3000uF range with adequate voltage ratings for pretty much any audio power supply application, but they are expensive (well over $100 each). I haven't read anything by anyone who's used them for high-end audio power supply applications. Personally, I doubt they would be useful.


Jeff
 
Physical size can matter for coupling caps as larger caps will have more stray capacitance to nearby conductors, so may pick up more hum and interference. In bad cases they may even lead to instability. This may be audible, but might not always be recognised for what it is.

The larger form factors will generally provide better protection against dielectric failure. In practice, the larger size (especially for surface mount components) is usually about a millimeter. I very much doubt this could increase noise sensitivity by any measurable amount.
 
Jeffin90620 said:
The larger form factors will generally provide better protection against dielectric failure. In practice, the larger size (especially for surface mount components) is usually about a millimeter.
I didn't realise we were talking about surface mount components. For conventional components some expensive caps can be 5 times the size of an ordinary cap, and then the problems I described can occur.
 
Size is everything - well maybe

As stated before, size is typically due to dielectric thickness increase which related to voltage withstand. but.....not always.

Construction and electrode type/size has a big impact on size. Take film capacitors for example...metalized film has an electrode thickness around 600 to 1000 angstroms, whereas discrete foil is around 6 um. Makes a big difference in size. Also makes a big difference in ESR and ESL. The diameter to length ratio as has a large impact on ESR and ESL.

The larger the diameter, and the shorted the length, the lower the ESR & ESL

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