Hi,
Ceramic caps change with time (aging) and thus are not considered good for audio line level application. After all, we dont want a filter
whose corner shifts with age.
A friend of mine who has good experience in electronics was of opinion that if SMD ceramics are used at a voltage lesser than 1/3 of their
rated voltage then they don't change the value with age and thus can be used for audio crossover.
Whats your opinion? Are new SMD ceramics usable in this application?
Regards,
WA
Ceramic caps change with time (aging) and thus are not considered good for audio line level application. After all, we dont want a filter
whose corner shifts with age.
A friend of mine who has good experience in electronics was of opinion that if SMD ceramics are used at a voltage lesser than 1/3 of their
rated voltage then they don't change the value with age and thus can be used for audio crossover.
Whats your opinion? Are new SMD ceramics usable in this application?
Regards,
WA
The facts are:
C0G (aka NP0) ceramic are absolutely fine, as good as any film cap. However they don't go up to high values of capacitance due to their low capacitance-density. If you can get C0G capacitor with high enough capacitance and voltage for a crossover it will be fine. Note that large value C0G caps and film caps are being faked - only buy from reputable suppliers.
Other ceramic's are heavily non-linear (and unstable with voltage and temperature) and have no place in an audio signal chain, but can be used for supply decoupling or RF suppression.
Whether SMD or through hole makes no difference to this, thats just the packaging (but through-hole have more inductance)
No opinions were used in this posting, see the data: https://www.collinsaudio.com/Prosound_Workshop/Capacitor-Sound.pdf
C0G (aka NP0) ceramic are absolutely fine, as good as any film cap. However they don't go up to high values of capacitance due to their low capacitance-density. If you can get C0G capacitor with high enough capacitance and voltage for a crossover it will be fine. Note that large value C0G caps and film caps are being faked - only buy from reputable suppliers.
Other ceramic's are heavily non-linear (and unstable with voltage and temperature) and have no place in an audio signal chain, but can be used for supply decoupling or RF suppression.
Whether SMD or through hole makes no difference to this, thats just the packaging (but through-hole have more inductance)
No opinions were used in this posting, see the data: https://www.collinsaudio.com/Prosound_Workshop/Capacitor-Sound.pdf
Do they even make those awful MLCC capacitors (that change capacitance so greatly with voltage) as thru-hole parts? They appear to be an 'innovation' of surface mount technology.
I'd hesitate to even use C0G/NP0 in production as a some parts procurement person might find that "these other ceramics of the same value are cheaper" and end up with a production run with the wrong parts.
I'd hesitate to even use C0G/NP0 in production as a some parts procurement person might find that "these other ceramics of the same value are cheaper" and end up with a production run with the wrong parts.
Member
Joined 2009
Paid Member
I also like C0G ceramics.
(when I need higher value capacitors I do use electrolytic and I've had good experience with Nichicon MUSE Bipolar types, large enough values to keep corner frequencies below audio)
(when I need higher value capacitors I do use electrolytic and I've had good experience with Nichicon MUSE Bipolar types, large enough values to keep corner frequencies below audio)
Pretty much everything is multilayer these days - even the thru hole stuff. Finding old school discs is hard these days - except in the kV stuff. You can make npo’s “multilayer” too - all the high value ones are.
If the corner frequency is low enough distortion due to nonlinear dielectric doesn’t even happen. Bypass caps are just as much in the signal path as the blocking cap is. It keeps the power rail at AC ground, so that AC voltage doesn’t develop across it. AC currents still pass right through it, and an assumption is made of negligible AC voltage across it. Blocking cap isn’t suppose to have AC voltage across it either, unless it’s the one that sets the corner frequency. Good design practice would only have that ONE set the corner frequency - all the others would be significantly lower corner frequency.
If the corner frequency is low enough distortion due to nonlinear dielectric doesn’t even happen. Bypass caps are just as much in the signal path as the blocking cap is. It keeps the power rail at AC ground, so that AC voltage doesn’t develop across it. AC currents still pass right through it, and an assumption is made of negligible AC voltage across it. Blocking cap isn’t suppose to have AC voltage across it either, unless it’s the one that sets the corner frequency. Good design practice would only have that ONE set the corner frequency - all the others would be significantly lower corner frequency.
I use COG/NPO ceramic SMD caps in the compensation circuits of my designs. MLCC X7R and other types are reserved only for supply bypass applications where absolute value is not too critical.
Unless high precision and stability are really needed, as in an active crossover or resonant EQ circuits (think Gyrators), I have happily used ceramics (yes, the dreaded disc types too) all over the place since the dinosaurs roamed the Earth,
All Guitar amp and many commercial Audio/home type amp makers agree.
Fully understand that HiFi amp builders diss them and prefer better types... to each his own 🙂
Champagne is good but beer has its place too 😉
All Guitar amp and many commercial Audio/home type amp makers agree.
Fully understand that HiFi amp builders diss them and prefer better types... to each his own 🙂
Champagne is good but beer has its place too 😉
MLCC just means multi-layer construction, which means more capacitance in a smaller volume with very low inductance - that's absolutely a good thing - the particular dielectric used is independent from the construction technique. SMT caps have lower inductance and thus always perform better all else being equal. Yes plenty of MLCC through-hole caps, both C0G and otherwise.
The procurement thing is a definite risk if you don't have control of or good liason with the production line, but it applies to many kinds of parts, not just ceramic caps.
Nope, not the case. Using a X7R or similar cap for decoupling will reduce distortion because the rail voltage will be attenuated, so any contribution from non-linear rail voltages via the PSRR will be reduced. But the PSRR is often enough on its own to handle this - decoupling is primarily about stability and preventing unintended feedback via the supply rails.
Anyway you'd never sensibly use a non-linear ceramic cap inline in the signal path because they are strongly microphonic. And any such voltage is directly added to the signal, not attenuated by the PSRR.
Last edited: