A whole bunch of reasons. For audio, one big one is that ceramic capacitors are microphonic - tap on them and they make a boom in the signal. If you've heard of piezoelectric or ceramic microphone elements, they're largely made of the same stuff. The capacitance of X7R and Y7V capacitors (the commonest ceramic types) is both temperature and voltage dependent, which will mess up tuning. A Y7V capacitor might show an effective capacitance of only 50% the nominal value at 0 volts of signal, increase up to +20% with 1 volt and stabilize close to rated value at 2 volts. Unless you're talking C0G ceramics, all the ceramics behave horribly with DC voltage applied. Some of them have only 10% of the rated value when full rated DC voltage is applied. We get away with this for bypass capacitors by using ludicrously big ones - those 0.1 µF X7R capacitors that you see scattered all over our circuits may actually be providing only 0.01 µF of filtering - but that 0.01 µF is enough. That's also why you'll see us using 50 volt rated ceramic capacitors on 12 volt supplies. But the X7R, Y7V, dielectrics excel for that application because their equivalent series resistance is tint. C0G ceramics are fine for use in the signal path, but they only come in values up to a few nF. For larger values (1 nF-10 µf or so), film is really the best option. I mentioned polyester because it's cheap. Polypropylene is seven better. For surface mount, PEN and PPS are the best options - most of the other plastics won't take the heat of reflow. Mica and glass are also great for capacitors that are in the signal path, if you're dealing with frequencies high enough that the capacitors come in big enough values.
@SkippiiKai tolerance, voltage dependence, temperature dependence are all reasons to keep ceramics out of the signal path. Cheap ceramics also tend to be microphonic... just like a ceramic microphone element. The piezoelectric effect couples ambient noise and vibration into your signal. NP0 and C0G ceramics are a possible exception, but those max out at pretty small capacitance values.
Hi Kevin , why did you say it has to be a polyester capacitor, not a ceramic one?
A whole bunch of reasons.
For audio, one big one is that ceramic capacitors are microphonic - tap on them and they make a boom in the signal. If you've heard of piezoelectric or ceramic microphone elements, they're largely made of the same stuff.
The capacitance of X7R and Y7V capacitors (the commonest ceramic types) is both temperature and voltage dependent, which will mess up tuning. A Y7V capacitor might show an effective capacitance of only 50% the nominal value at 0 volts of signal, increase up to +20% with 1 volt and stabilize close to rated value at 2 volts.
Unless you're talking C0G ceramics, all the ceramics behave horribly with DC voltage applied. Some of them have only 10% of the rated value when full rated DC voltage is applied. We get away with this for bypass capacitors by using ludicrously big ones - those 0.1 µF X7R capacitors that you see scattered all over our circuits may actually be providing only 0.01 µF of filtering - but that 0.01 µF is enough. That's also why you'll see us using 50 volt rated ceramic capacitors on 12 volt supplies. But the X7R, Y7V, dielectrics excel for that application because their equivalent series resistance is tint.
C0G ceramics are fine for use in the signal path, but they only come in values up to a few nF. For larger values (1 nF-10 µf or so), film is really the best option. I mentioned polyester because it's cheap. Polypropylene is seven better. For surface mount, PEN and PPS are the best options - most of the other plastics won't take the heat of reflow. Mica and glass are also great for capacitors that are in the signal path, if you're dealing with frequencies high enough that the capacitors come in big enough values.
@@KludgesFromKevinsCave A very comprehensive answer, thank you!
Why can't you use ceramics in a schmitt trigger? You say to only use polys. Is it just the tolerance needed for music, or are there other reasons?
@SkippiiKai tolerance, voltage dependence, temperature dependence are all reasons to keep ceramics out of the signal path. Cheap ceramics also tend to be microphonic... just like a ceramic microphone element. The piezoelectric effect couples ambient noise and vibration into your signal.
NP0 and C0G ceramics are a possible exception, but those max out at pretty small capacitance values.
Ode to Joy of Electronics.
@@justovision Beethoven must be spinning in his grave.