AC coupling questions
don at till.com
Fri May 8 18:47:35 CEST 1998
From: media at mail1.nai.net
Date: Fri, 8 May 1998 02:22:55 -0400 (EDT)
On this subject the FAQ says, "Use 0.47uf metallized polyester or
polyproplyene. Values less than this will attenuate bass frequencies."
Mind you it makes no mention of impedances. Regardless, a .47uF
polypropylene cap is the size of a bumblebee.
I've looked at a number of audio circuits and have found a wide variety of
capacitor values used for AC coupling.
Yeah, that passage above is pretty clearly bogus.
While I'm fairly sure that a
capacitor of any value will block DC, we don't want to create a high pass
filter in the audio range.
Yup. The simple answer (and a really proper answer requires going
into the details of the whole circuit, but for now this is plenty
close...) is that the coupling cap creates a 6dB per octave high pass
filter with the input impedance of the circuit tuned with a -3dB point
at frequency f = 1 / (2 * pi * R * C); F is in Hz, R is the input
impedance in ohms, C is the coupling capacitance in Farads.
So if you have 10k ohm input impedance, a 0.47uF cap will give you a
-3dB point at 34Hz. And if you have a 100k ohm input impedance a 0.1uF
cap will give you a -3dB point at 16Hz.
Is there a capacitor value that is typically used for this in
First you'd need a standard input impedance. Which is (surprise!)
often choosen with reasonable cap values in mind.
Further, is blocking DC always necessary?? I'm guessing typical
coltrol voltages wouldn't hurt modules with op-amp imputs.
No it's not necessary. The ARP 2600 is kind'a famous for not blocking
DC in any of it's modules, and since it's VCO square waves go from zero
to positive there's a serious thumping effect on VCA'd square waves.
On the other hand, inputs that handle DC make a module all that more
useful. You can use your VCF for overshoot portamento, f'rinstance.
So the best thing is to have both DC and AC inputs.
More information about the Synth-diy