SMT caps in signal path

[ChrisA]

Looking for a line of good sounding small value caps for use in the signal path but I want surface mount. Maybe this does not exist.

My preferred building technique is to try and use SMT when I can. Not only is this quicker to build and cost less but the length of the traces can be kept to a minimum.

But I've hear very little about SMT caps used in audio. So far I've only used SMT caps for power supply bypass. And they work great for that because of the (literally) zero length leads have zero inductance and they take so little space.

[thetube0a3]

I used either NP0 or C0G 0805 5% ceramics on my active xovers. Have not had any issues with them.

I don't remember the brand however I can check when I get home if you like, Murata comes to mind.

The opamps I used are SO-8 TL-072B's I believe. I swear by SMT when Possible

they standard 12dB\octave and 24dB\octave linkwitz-riley designs and tho many might not think they are the worlds best crossovers I like them.

[theAnonymous1]

I use ceramics, but I don't suffer from Audiophilia Nervosa so it's not an issue.

For anything close to being "audiophile grade" in SMD format your choices are limited. Look for PEN and PPS film types.

[thetube0a3]

theAnonymous1 : nice Never heard of Audiophilia Nervosa before. I'm glad I don't suffer from that either.

[Nelson Bass]

Originally Posted by thetube0a3

theAnonymous1 : nice Never heard of Audiophilia Nervosa before. I'm glad I don't suffer from that either.

How can you be sure without knowing the disease?!! ;-)
Google is your friend.... ;-)

Madrigal lab use smt/smd components everythere and a few Wima polyester caps so...
Forget the snaike oil and use those nice SMDs... :-)

[twest820]

Oh, that's a nice link. Maxim's data suggests favoring big packages and overspecing on voltage. The smaller value lines of Taiyo Uden's *MKxxx series are billed as film replacements but I haven't found any measurements of how low the distortion actually is. Panasonic's EC*U series includes some fairly large values.

I DC couple and use a digital crossover so my need for caps in the signal path is minimal. But I am planning to use a 1uF ECPU part in the DC servo RC tank of an amp I'm working on. And my experience has been that if the caps are hitting 0.01% distortion some attention is needed to the op amps and power stages to get performance to be limited by the caps at typical signal levels.

[gootee]

What do you mean by "signal path"? In a power amplifier, there is a small-signal path from the source to the input pins of the amplifying component(s). It basically stops there. But the output's signal path goes through the power supply and the high-current portion of the amplifying component(s). So basically everything is "in the signal path".

[twest820]

Depending on the design and what one chooses as the threshold for relevancy it can be useful to make a distinction between stuff included in the control loop---the signal path---and stuff that's not. With good layout, balanced interconnects, and reasonable PSRR the supply/transformer/power cord/yadda yadda is fairly well excluded. With unbalanced interconnects the ground loop formed by the interconnect, amp's power cord, amp's grounding, source's power cord, source's ground, transformer leakage inductances, AC lift capacitors, and whatever wall wiring is relevant are, IMO, best included.

[theAnonymous1]

Originally Posted by twest820

Maxim's data suggests favoring big packages and overspecing on voltage.

Thanks for that link. It explains why the 1uF 100V 1812 X7R caps I use sound perfectly fine to me.

[Neil Davis]

PPS is probably the lowest distortion cap. You can get values up to 1uF in SMD packages.

See the study at this link: Capacitor Sound

Here's another good article:

http://www.kemet.com/kemet/web/homep...Capacitors.pdf

Be careful about which solvent you use to clean the flux--the manufacturer only recommends iso and ethyl alcohols. Also, the recommended soldering temperature for PPS is lower than most caps.

They look cool, too
PPS_caps.jpg