The problem with the ceramic capacitors is that the only dielectric accepted in signal path is NP0(C0G) but the highest capacitance available for such capacitors is around 0.47uF at the moment. If you parallel two of them you should be fine. At 1uF the alternative would be film capacitor rather than electrolytic.
Member
Joined 2009
Paid Member
It is my opinion that you can use ceramic providing it is NP0/C0G type of ceramic. I really like these capacitors and use them in lots of places. But for an input cap I like larger values and I prefer to use bi-polar electrolytics from Nichicon audio range. They can be very small in size and you can use a 10uF value - I encourage you to look at this option.
fine as long as its NP0,
https://en.wikipedia.org/wiki/Ceramic_capacitor#/media/File:MLCC-Kap-versus-Frequenz-engl.svg
see image for explanation.
other cermaics have a freqvency dependent capacity and are not quite suited as input caps.
however i prefer some smaller foil types paralelld.. as long as it gives around 5-10 uF its more than enough. the -3 dBL point will be far out of the recorded material, any further reduction of 3fL is useless.
https://en.wikipedia.org/wiki/Ceramic_capacitor#/media/File:MLCC-Kap-versus-Frequenz-engl.svg
see image for explanation.
other cermaics have a freqvency dependent capacity and are not quite suited as input caps.
however i prefer some smaller foil types paralelld.. as long as it gives around 5-10 uF its more than enough. the -3 dBL point will be far out of the recorded material, any further reduction of 3fL is useless.
I've used 4.7 uf aerovox 50 v ceramic as an input cap. John C would not approve. It sounds waay better than the poppity from the store tantalum caps (both of them) I replaced. Your luck buying tantalum may be better than mine.
I think the 50v rating ceramic on a 1 v signal linearizes the passthrough signal somewhat. My speakers are 20 db down from 1 W level harmonic distortion (actually rated that by manuf.) so if yours are .0003% hd you might hear the difference to electrolytic. Not very likely, IMHO.
I think the 50v rating ceramic on a 1 v signal linearizes the passthrough signal somewhat. My speakers are 20 db down from 1 W level harmonic distortion (actually rated that by manuf.) so if yours are .0003% hd you might hear the difference to electrolytic. Not very likely, IMHO.
Last edited:
Member
Joined 2009
Paid Member
I started my projects years ago with a film cap at the input, usually MKP 0.47uf. One day I tried a bipolar electrolytic and it sounded better - by that I mean subjectively I preferred it.
You can use capacitance values that are large enough to ensure that the signal induced distortion is negligible, that low frequency response is great, and with bipolar you have much lower distortion to start with. My best sounding amplifier uses it on the input (and feedback shunt if I remember).
I would encourage you to hear some different capacitors with your own ears, as there is no substitute for your own experience. You can make a pcb footprint that allows you to install two different options fairly easily.
My favourite is a tiny cap, a 5mm diameter footprint (2mm lead spacing), 10uF 25V that cost less than a dollar from the Nichicon MUSE (audio grade) range.
https://www.digikey.ca/product-detail/en/nichicon/UES1E100MDM1TD/493-10824-1-ND/4317799
You can use capacitance values that are large enough to ensure that the signal induced distortion is negligible, that low frequency response is great, and with bipolar you have much lower distortion to start with. My best sounding amplifier uses it on the input (and feedback shunt if I remember).
I would encourage you to hear some different capacitors with your own ears, as there is no substitute for your own experience. You can make a pcb footprint that allows you to install two different options fairly easily.
My favourite is a tiny cap, a 5mm diameter footprint (2mm lead spacing), 10uF 25V that cost less than a dollar from the Nichicon MUSE (audio grade) range.
https://www.digikey.ca/product-detail/en/nichicon/UES1E100MDM1TD/493-10824-1-ND/4317799
Attachments
Member
Joined 2009
Paid Member
AudioPrecision for some years now mention once in a while, class II smd ceramics have changed for the better. Now film smd are often smaller size than COG (classI), inferior to COG these days, often more expensive and almost as hard to source. ClassII is smaller yet and inferior again. The difference is not as big as it used to be though. And so it happens that better measuring and sounding new OEM ampboards now have classII ceramics on inputs. (Waiting to be upgraded for magazine reviews and marketing by people that will resell them in nice cabinets, but not necessary to better previous generations OEM ampboards that used through hole filmcaps).
Piezo effect
Ceramic capacitors shows Piezo effect and so its not recommended in audio path.
https://product.tdk.com/en/contact/...lectric_effect.pdf?disppage=1&faq_category=02
Power Tip 49: Avoid these common multi-layer ceramic capacitor pitfalls | EE Times
Ceramic capacitors shows Piezo effect and so its not recommended in audio path.
https://product.tdk.com/en/contact/...lectric_effect.pdf?disppage=1&faq_category=02
Power Tip 49: Avoid these common multi-layer ceramic capacitor pitfalls | EE Times
Member
Joined 2009
Paid Member
Indeed, these kinds of ceramics are pretty awful, they are not the recommended type for audio but they can be effective for power rail decoupling - sometimes they are better than the audio-approved Ceramics because they are lossy and less inclined to parasitic ringing.
For audio, use the C0G/NP0, they are excellent, better than run of the mill film types. I use them for feedback loop compensation.
For audio, use the C0G/NP0, they are excellent, better than run of the mill film types. I use them for feedback loop compensation.
Recent quote from Hypex thread, shortened, I removed name poster
Ampboard signal input coupling caps: classII ceramics. After opampstage signal coupling to TI-chip: classII ceramics. Twice through unfit for audio parts, yet very close to best, best Icepower yet. Do we now conclude the Icepower100AS has superior circuitry and is superior to Ncore, but underperforms a little because it uses unfit for audio coupling capacitors ? Twice.
Xxxxxxxx didn't refer to ceramics as secret weapon, but to TI-chip. He might not have noticed signalpath classII ceramics, visually I mean, he did notice listening, we all know, he must have.
Ampboard signal input coupling caps: classII ceramics. After opampstage signal coupling to TI-chip: classII ceramics. Twice through unfit for audio parts, yet very close to best, best Icepower yet. Do we now conclude the Icepower100AS has superior circuitry and is superior to Ncore, but underperforms a little because it uses unfit for audio coupling capacitors ? Twice.
Xxxxxxxx didn't refer to ceramics as secret weapon, but to TI-chip. He might not have noticed signalpath classII ceramics, visually I mean, he did notice listening, we all know, he must have.
for SMD audio try solid tantalum instead.
ceramics NP0 will be huge and costly if you could even locate a source. other crystalline dielectrics are problematic in the signal/gain path.
I typically use X7R for audio coupling - definitely not officially recommended for audio but subjectively not the lowest hanging fruit for SQ in my experience. The TDA2050 will be responsible for considerably more loss of transparency than almost any ceramic cap. I do tend to oversize by at least 100% though, so if your application calls for 1uF use 2.2uF as a minimum. Also avoid the lower voltage ratings unless you're seriously pushed for space.
now that i have the article in front of me LA12 -- C0G have consistent THD% performance, with higher voltage units better.
polypropylene film, as contrast with polypropylene film-foil metalized had better performance.
authors Sam Groner and Scott Wurcer.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.