# Capacitors and filtering

I noticed that in many audio op-amps, RC constant is designed so that the cutoff frequency is about 16 Hz. I think that's because the human ear responses to frequencies higher than 22 Hz. But I am a bit confused. If the RC component decides the cutoff frequency, why do they always use values of about 100kOhm and 0.1uF? Can't it also be a different combination, such as 1kOhm and 10 uF. Is that because the current issue?

• Cost. A 100nF capacitor will cost less than a 10$\mu$F one. Apr 4 '12 at 13:08
• There is no absolute reason, but for example capacitors scale worse than resistors, so bigger capacitors are harder to handle than bigger resistors. But also, both are reasonable values, instead of using GigaOhms and picoFarads Apr 4 '12 at 13:09
• Also power: Depending where it's used a larger resistor is likely to result in a lower-power circuit. But 100k is more-or-less the maximum value for a tight-tolerance resistor. Higher values are likely to be skewed by parasitic leakages. So they're using basically the maximum reasonable resistor value, then choosing the capacitor to get the RC they want. Apr 4 '12 at 15:22
• And between the three of us, we may have an answer.... Apr 4 '12 at 15:22
• Give us an example schematic and we may be able to offer some better insight.
– user3624
Apr 4 '12 at 16:09

1) the reason there is a 16KHz cut off: While our ears do have a hearing range of 20-22KHz the actual range that we use in our every day life for hearing is about 6-10KHz (at most)

If you are listening to music, one of the things to realize is that pitches, musical notes use a logarithmic scale. so A4 is at 440Hz while A5 880 Hz A6 is at 1760 Hz what this means that there are only a few notes are left to be able to play above 16kHz. Just for reference, the highest note played on a piano would be 7.9kHz which is a B

http://peabody.sapp.org/class/st2/lab/notehz/

2) 100K ohm resistors are cheap and common, same-thing with the capacitor. its the most cost effective choice.

• IMO, your statement about the logarithmic scale of musical notes is missleading. Sure it is true for pure tones, but music of real instruments generate several other frequencies when you play just one note. Apr 16 '12 at 19:12
• this is true, but there are varying debates about the importance of frequencies above 10kHz. In addition, you also have to think about sub-audible and super-audible frequencies, people also think that these hold significant importance or play a noticeable roll in our hearing. Apr 16 '12 at 19:15

Life is all about tradeoffs. 0.1uF is much cheaper than 10 uF 0.01uF is not practical. Because the load R becomes 10MOhm.

Some Op Amps have input offset currents which translates into an offset DC voltage at high gain or with high resistance bias or filter resistance... which is the whole point .. no dc...

Thats why some PC microphones which have DC on them are then AC coupled but still when boosted to max 20dB or whatever in the audio channel and people hear a DC pop when you key up your microphone in a chat room. due to the high gain offset.

;) ;)