[sdiy] Simple clickless FET mute switch?

[Mattias Rickardsson]

2008/9/17 John Mahoney <jmahoney at gate.net>:
> At 08:38 AM 9/17/2008, Mattias Rickardsson wrote:
>> [snip]
>> > To avoid the click it needs to be either zero-crossing.
>> > Or be a gradual fade over at least 10ms.
>>
>> Well, switching a signal instantaneously at a zero crossing also gives
>> a click, albeit a mufflier one. Otherwise, triangle waves would be
>> silent, you know! :-)
>
> Um, no, I don't know! You've completely lost me, there.

When a signal is turned off to zero at a non-zero timing, you will
hear a strong click at the "discontinuity". Now, if you instead turn
it off where the signal would cross zero, the result will not have any
sharp "discontinuity" anymore, but a sharp "corner" - which is also
heard, but it doesn't contain such a lot of high frequencies... so it
might be OK in some applications. But probably not with a bass tone.

Take a square wave as example. Integrate it, and you get a triangle
wave. The "discontinuities", which are just like the switched non-zero
signal above, are removed and you have "corners" instead, like the
zero-aligned switched signal above.

The integrator is in fact a one-pole lowpass filter. -6dB/oct.
So "corners" are like "discontinuities" run through a one-pole lowpas
filter. They are still heard.

How did that sound? :-)

/mr