RFI/EMI Filter Caps - Film or C0G?

I am wondering if there is any audible benefit to use polypropylene caps in input/output RFI filtering or ceramic C0G types would be suitable without sacrificing audio band performance. Here is a sample schematic of the application:

Input:


Output:


PS: For the L1-L4 ferrites I've been using Panasonic EXC-ELSA39.

Someone can probably chime in with better explanation, but I usually always see those as c0g.

Ceramic is very good at conducting at high frequencies, plus usually a little cheaper to boot!

some people swear against ceramics, but I use em for filtering all the time, no problems that I can notice. I just try to keep em out of the signal path, as they can be a bit brittle sounding to me (say, in a tone stack for a guitar amp).

...personal opinion.

for a guitar thing, try to use high-voltage ceramics. COG if possible.

for some "hairy mojo" I had some luck with x7r multilayers as well.

Surely these won't fit into all circuits out there..


To OP: Imo film caps will "sound" better, and ceramics will better filter out the RFI (and add some mojo of their own, whether you like it or not).

In that case, you may want film if your making only one, and cermic cog's if your planning on getting the units ce/ul certified?

There is no audible benefit to using film over cog for a RF filter  (AFAIK).

WRT RF filtering the ceramic may be lower impedance so work better.

At best, film might work "as good" for that task, not likely to be better.

There are good arguments that cog/npo dielectric isn't even needed there over cheaper ceramic dielectrics, but they are not expensive so why worry?

JR

Thank you all for replies! It's just that I've seen people on this forum use film caps for RFI (Igor's PGA2500 for example) and was wondering if it's REALLY worth it.

And what about silver-mica caps, that are available in 220-470pF values?

I recall mica from the old days... IIRC they were decent for low Z at HF but i doubt they are superior linearity to cog/npo.

JR

"Film cap" is a very unprecise terminology; the (for audio applications relevant) electrical differences (mostly voltage coefficient) between a polyester film and a polypropylene film are larger than between a ceramic C0G/NP0 and polypropylene film. In any case ceramic C0G/NP0 is an excellent dielectric for any audio cap, and very appropriate for RF filtering. Note that to fully use their low HF impedance quite some attention to layout might be necessary.

Samuel

X2Y capacitors?  They have two capacitors in one package designed for RF filtering able to replace several components and save board space... see page 6 of the PDF for PCB layout considerations... page 4 for audio RFI filtering in GSM phones...

May have some caveats for SMT hand soldering (like don't directly touch the capacitor's terminals with the iron - cracking of ceramic)...

http://johansondielectrics.com/images/stories/surface-mount/x2y/JDI_X2Y_2010-06.pdf

http://johansondielectrics.com/x2y-products.html

Your link does not link to a specific filter and your question does not make sense.

JR

NPO/CGO ceramics are probably the best capacitors you can buy.  See Cyril Bateman for details.  If the value/voltage/size is right, use them.

Resistors need to obey Ohms Law under all operating conditions and capacitors must meet

I = C dV/dt

Apart from problems with soldering, polystyrenes are microphonic and the way they're made means even 1% bits have poor distribution.  At Calrec, we once stripped every single small polystyrene from a 72 channel Broadcast Desk and replaced them with NPOs.

Silver Mica, the Golden Pinnae darling, has "high" loss.  Replacing the input ceramic capacitor in a condensor mike may increase noise.  Scott Wurcer will show measurements in the next issue of Linear Audio.

For EMI protection, size matters.  Big "film" caps are good aerials to pick up crap.  A common mike "improvement" is to replace electrolytics or ceramics with huge films.  This often results in RFI.

Capacitors at a mike XLR need to be 10n+.  Unfortunately, you can't get NPO/CGO ceramics that size.  But the advantages of small size and good RF outweigh the poor dielectric of non-NPO/CGO ceramics compared to films.

However, if you can hear the slight midrange roughness with these, Cooktown Recording & Ambisonic Productions has Golden Pinnae capacitors hand carved by Virgins from Solid Unobtainium.  Send $100 in used bank notes for a sample.

Then what are "film" caps best for?
Thanks in advance.
Best,
Bruno2000

ricardo said:

NPO/CGO ceramics are probably the best capacitors you can buy.  See Cyril Bateman for details.  If the value/voltage/size is right, use them.

Resistors need to obey Ohms Law under all operating conditions and capacitors must meet

I = C dV/dt

Apart from problems with soldering, polystyrenes are microphonic and the way they're made means even 1% bits have poor distribution.  At Calrec, we once stripped every single small polystyrene from a 72 channel Broadcast Desk and replaced them with NPOs.

Silver Mica, the Golden Pinnae darling, has "high" loss.  Replacing the input ceramic capacitor in a condensor mike may increase noise.  Scott Wurcer will show measurements in the next issue of Linear Audio.

For EMI protection, size matters.  Big "film" caps are good aerials to pick up crap.  A common mike "improvement" is to replace electrolytics or ceramics with huge films.  This often results in RFI.

Capacitors at a mike XLR need to be 10n+.  Unfortunately, you can't get NPO/CGO ceramics that size.  But the advantages of small size and good RF outweigh the poor dielectric of non-NPO/CGO ceramics compared to films.

However, if you can hear the slight midrange roughness with these, Cooktown Recording & Ambisonic Productions has Golden Pinnae capacitors hand carved by Virgins from Solid Unobtainium.  Send $100 in used bank notes for a sample.

Click to expand...

Kemet makes COG/NPO SMD up to .47 Uf.. not cheap and not small for SMD (2220 so .22x.20" square), but perhaps not crazy for a tight application (with too much money to spend).

JR

ricardo said:

At Calrec, we once stripped every single small polystyrene from a 72 channel Broadcast Desk and replaced them with NPOs.

Click to expand...

Why would Calrec use polystyrene caps in a high cost console if c0g ceramics were better?  Cost?
Thanks again in advance!
Best,
Bruno2000

bruno2000 said:

Why would Calrec use polystyrene caps in a high cost console if c0g ceramics were better?

Click to expand...

This was a defining moment for us (early 80's).  Before, we assumed the Golden Pinnae were right.  Then we found out.

ricardo said:

bruno2000 said:

Why would Calrec use polystyrene caps in a high cost console if c0g ceramics were better?

Click to expand...

This was a defining moment for us (early 80's).  Before, we assumed the Golden Pinnae were right.  Then we found out.

Click to expand...

This is what I love about this forum.
Thanks!
Best,
Bruno2000

I am pretty far from a golden ear... but I was always a fan of polystyrene because they are an excellent audio cap dielectric and cheap, two things I admire in a cap.  My early recollection of cog/npo was only found in small low value caps used around crystals for frequency stability and the like. Over the decades as I got into larger scale manufacturing production processes the polystyrene caps turned out to not be robust enough. My old business partner was able to kill them by drying off a PCB after a water wash (damn that acid flux) with a high pressure air hose. I never even tried to run them thought a proper wave soldering process line with inline board washers and such.

The first time I searched out a NPO/COG ceramic big enough to use in an audio filter was around 12 years ago for a SMD DSP design after the FAB I was using melted the audio grade film SMD audio caps I specified. Even thought these caps were designed for and supposed to survive the SMD temperature curves (Panasonic IIRC) , the contract manufacturer (in TN) managed to kill them.

Either I was blissfully unaware that npo/cog were a good alternative back a few decades ago (possible I guess), or these only became widely used for audio more recently as SMD and lead free solder processes put the last coffin nail in film caps for SMD production (more likely).

For HF like RF a NPO/COG ceramic is actually better than polystyrene ever could be, because the polystyrene construction of wrapping a long strip of film into a cylinder, has more inductance than typical ceramic cap construction. 

JR 

bruno2000 said:

Then what are "film" caps best for?

Click to expand...

Alas, for response shaping stuff (EQ etc), where you must have significant signal voltage across them, "films" are the most cost/space effective solution.

See Cyril Bateman for details of the choices.  Note the amounts of distortion he measures are quite small for most "films".  Yus pays your money & takes your choice.

For the Golden Pinnae who sneer at NE5534, see what he chooses for ultra low distortion drivers.