Typical capacitor values for audio

[JohnRoberts]

Dielectric absorption is very favoable in Styrene caps. Have not used them audio, nor any ceramics, some SMT caps are microphonic, pressure sensitive. Leaded ceramic may be OK.

I don't see a distinction between leaded TH caps and SMD ceramics. I recall back in the 80s visiting a cap manufacturer (in Mexico) who was attaching leads and potting(?) small capacitors. This was before the popularity of SMD and the capacitor dice being leaded looked almost like modern SMD.

The only microphonic caps I have ever encountered were TH and faulty.

I have an old stash of 7uF 300V (mylar?) with very low ESR, like 0.008 Ohm, I use for decoupling.

Large?

JR

 

[Mr. Moscode]

I don't see a distinction between leaded TH caps and SMD ceramics. I recall back in the 80s visiting a cap manufacturer (in Mexico) who was attaching leads and potting(?) small capacitors. This was before the popularity of SMD and the capacitor dice being leaded looked almost like modern SMD.

The only microphonic caps I have ever encountered were TH and faulty.

Large?

JR

Every cap is microphonic. HV caps more so.

To see it connect your cap to a scope. connect that to a power supply through a 1 meg R to energize the cap, like a condensor mic. If it is a HV cap use a high polarizing voltage, if low volts use highest you can safely use. Now tap on the cap and observe the wiggle on the scope. You may need to use a high gain setting on the scope but it's there. This test is most pertinent to caps that are large enough to use as coupling caps.

Electrostatic forces are at work inside the wrap of the cap making it a transducer.

Small value caps may resonate but it would be a high frequencies.

Depending where the cap is in the circuit will determine how much effect that would have in the output.

Likewise circuit boards are also microphonic but that's most noticeable in hi Z, hi voltage, high gain circuits like a tube phono/tapehead/mic preamp. Anywhere a trace on once side crosses a trace on the other you have capacitor.

Another test: Connect the cap to audio power amp through a 1 ohm resistor. Feed a 1khz square wave to the amp. Listen to the cap radiate 1khz noise. Alternately sweep a sine wave through the audio range and listen. If it's resonant in the audio band you'll hear it if your ears are young enough. For us older folks you could mic it.

Anytime I test a power amp at high power under load it sings. Sometimes without a load, too.

 

[JohnRoberts]

Every cap is microphonic. HV caps more so.

To see it connect your cap to a scope. connect that to a power supply through a 1 meg R to energize the cap, like a condensor mic. If it is a HV cap use a high polarizing voltage, if low volts use highest you can safely use. Now tap on the cap and observe the wiggle on the scope. You may need to use a high gain setting on the scope but it's there. This test is most pertinent to caps that are large enough to use as coupling caps.

Electrostatic forces are at work inside the wrap of the cap making it a transducer.

Small value caps may resonate but it would be a high frequencies.

Depending where the cap is in the circuit will determine how much effect that would have in the output.

Likewise circuit boards are also microphonic but that's most noticeable in hi Z, hi voltage, high gain circuits like a tube phono/tapehead/mic preamp. Anywhere a trace on once side crosses a trace on the other you have capacitor.

Another test: Connect the cap to audio power amp through a 1 ohm resistor. Feed a 1khz square wave to the amp. Listen to the cap radiate 1khz noise. Alternately sweep a sine wave through the audio range and listen. If it's resonant in the audio band you'll hear it if your ears are young enough. For us older folks you could mic it.

Anytime I test a power amp at high power under load it sings. Sometimes without a load, too.

the load resistors make even more noise at high power....

Allow me to restate, "the only significantly microphonic caps I've encountered were faulty".

JR

 

[Mr. Moscode]

JR,
You are always welcome to your experience and opinion. This is my experience from my 40 years of manufacturing and experiments. How did I find out? I had attached large coupling caps to a circuit board with 2 sided tape. Pressing or tapping the board created significant wiggle in the output. Remounting the caps by leads only greatly reduced the problem. In a phono preamp for a moving coil cartridge gain at 20hz is 100db for a 1Khz gain of 80db.

However, don't take my word for it, run this experiment and post your result. I'd love to see what you get.

Significance is subjective and in the ear of the beholder.

 

[Cqwet Dbdfte]

I don't see a distinction between leaded TH caps and SMD ceramics. I recall back in the 80s visiting a cap manufacturer (in Mexico) who was attaching leads and potting(?) small capacitors. This was before the popularity of SMD and the capacitor dice being leaded looked almost like modern SMD.

The only microphonic caps I have ever encountered were TH and faulty.

Large?

JR

Cap size or size of stash?

 

[JohnRoberts]

Cap size or size of stash?

A 7uF 300V mylar cap will be physically large.

The long leads may compromise their low ESR in practice.

JR

 

[Cqwet Dbdfte]

Body 25mm long 14mm diameter.
These were specially designed for pulse discharge, end foils are spray metallised, then epoxy covered. Every single cap tested at least 80,000 pulses, 300V into 1 Ohm, then checked for ESR at 10KHz.

 

[JohnRoberts]

That value might also be suitable for passive loudspeaker crossovers, and a better dielectric than aluminum electrolytic for audio frequency filter poles.

Roughly 1" long is not typical spacing for PS decoupling capacitors.

JR

 

[Cqwet Dbdfte]

Right, nothing I do is typical, these caps are used in my 2-way second order crossover, using Dynaudio elements, no longer sold separately. Also given some of these caps to friends for speaker use. To a man with a hammer everything looks like it needs a nice film cap.
In power supplies these are a bit low in voltage, where I prefer photo flash caps for their capacity and low ESR, works great with silicon rectifiers with inrush current limiters on transformer primary.
Replacing vacuum rectifiers with silicon has some benefits, including reducing some static loading on power transformer, less heat, bigger caps. Yes, 1500uF instead of 40uF does make a difference. Old school power supplies were "good enough" and vaccum rectifiers does provide the R in an RC network, which can of course be substituted with power resistors. Some silicon rectifiers have noisy characteristics that will be tamed with said resistors. Then there is also SiC diodes that should work great for this, which I have yet to install in my power amps. A classic harmonics generator circuit has an inductor in series with a diode, where the diode capacitance plays into harmonics generation, for a power supply with a big inductance this would not be so bad, except the transformer leakage inductance can be excited, typically this is tamed with R or C in parallel with diodes, which also works. Transformer can be swept for self resonance and filter constructed for this frequency, which I have yet to try.

 

[jBam]

Hi All - been caught up with Family and intense pre-christmas work BS!... Just the other side of survival mode now

So - I'm at a point where I've loaded up a Mouser cart based on advice here and lots of research... I'm not sure if it's... laughably expensive, or if the quality I'm aiming for just costs this much. For now, it's hovering around AU$100 for my Capacitor selection (say USD$70 on a good day for Aussies).

I've based my Cap values on the initial approx. span that CJ gave early on on page 1, supplemented by other comments / advice here and a range of example circuits I've had a look at. This is roughly (for audio signal related, not power stuff) 10pF through to (only?) 4.7uF. Seems to cover the audio spectrum for filtering with reasonable resistor values and also covers various other uses for e.g. OpAmp stability... I'm typically grabbing it in 10, 22, 47, 100... intervals.

The intent is: A selection of capacitors that represent the quality I'm going for in "ultimate" designs I hope to develop up. This means:

• C0G ceramic for the lower end values (10pF to approx. 0.047uF)
• Film Polypropelene for the mid range values (0.047uF to 4.7uF)
• I've added a few electrolytics... not 100% sure in this space yet, and for now will steer clear of power supply design... but yeah - a few 1000uF.

Numbers of units? Typically going first break point of 10 each for caps valued less than AU$1 each (EDIT: actually... it's around 60c each). Larger Film Caps (0.47uF to 4.7uF) start getting expensive, so I'm just grabbing 2 of each.

So...

• Considering I'm looking to get a collection of higher quality components, is AU$100 sounding ok? Not a lot to spend for Christmas I guess ...
• Do those C0G / Film PP ranges of values above sound good?... it's roughly where the $ point crosses over (although I could go lower with film... they just seem to be getting physically quite small below 0.047uF)
• Is there an obvious Electrolytic for audio path type stuff I should add?... I don't want to add too many poor shelf-life items... is my assumption that "4.7uF film caps should get me to 20Hz filter range" kinda safe??

As always - really appreciate any advice

I can share the current cart, but don't expect / want people wasting there time in there.... it's also all out of order from various changes, so it'd be annoying to look through lol...

 

[PermO]

Spending $70- on a wide range of caps does not seem extraordinary to me.
I think your selection is quite sensible and should get you a long way experimenting with circuits.

Maybe add just a few bipolar electrolitics, 10uF - 100uF - 200uF panasonic / nichicon ... these are sometimes used at headphone outputs, THAT balanced line drivers, for anything audio that is not powersupply related.

DIY is far from economic, DIY is expensive, if you try to make it cheap it;s going to end up expensive, your designs will never be more economic than mass produced China stuff.
There's a breaking point where it starts to make sense, that's with the high end stuff and the vintage "can no longer have that" stuff. It's cheaper to build a LA2A than to buy one, goes for a lot of these vintage classics.

I think I spend well over 500,- in parts last week
Maybe 50,- of that was passives, caps, fuses, inductors.
450,- on some hardware, connectors, rotary switches, boxes.... and shipping costs.

 

[jBam]

Great thanks @PermO ! Good to know I'm not far off the mark, and thanks for the lytics tips - I'll add some if them to the list