If I put my notes here, I might be able to find them again later!
Five-times-bandwidth
Posted 3rd January 2012 at 05:44 PM by rjm
If we accept that conventional wisdom that the audio bandwidth extends from 20 Hz to 20 kHz, a good rule of thumb for the f(-3 dB) high and low cutoff points of the frequency response of each audio circuit element is 5x the bandwidth, or 4 Hz to 100 kHz. In practice most designs tend to shift that range a little to the higher frequencies, so perhaps 5 Hz - 200 kHz, or 4 - 250 kHz.
Personally I "tune" my circuits to 4 Hz. That is, the time constants are adjusted to about 4 ms. Capacitance is usually cheap enough to go even longer, but the influence on sonics is typically net negative.
The high frequency side is more interesting, since many circuit elements naturally run into the megahertz range, the the question is do you actively try to prevent that, and if so, where and how?
The biggest issue is bypassing: a small value electrolytic (100 uF) is probably fine up to 100 kHz or so, but quite useless at 2 Mhz. The textbook solution is normally to connect a small capacitor in parallel, ceramic caps are recommended by electronic engineers, and frowned upon by audiophiles.
The related issue is stability, not just by insufficient power supply bypassing, but induced by poor layout, or high capacitance loads, or what have you.
In short I am unconvinced about the benefit of "high speed" circuits for audio, but remain open-minded about them, when due and proper care is taken.
Personally I "tune" my circuits to 4 Hz. That is, the time constants are adjusted to about 4 ms. Capacitance is usually cheap enough to go even longer, but the influence on sonics is typically net negative.
The high frequency side is more interesting, since many circuit elements naturally run into the megahertz range, the the question is do you actively try to prevent that, and if so, where and how?
The biggest issue is bypassing: a small value electrolytic (100 uF) is probably fine up to 100 kHz or so, but quite useless at 2 Mhz. The textbook solution is normally to connect a small capacitor in parallel, ceramic caps are recommended by electronic engineers, and frowned upon by audiophiles.
The related issue is stability, not just by insufficient power supply bypassing, but induced by poor layout, or high capacitance loads, or what have you.
In short I am unconvinced about the benefit of "high speed" circuits for audio, but remain open-minded about them, when due and proper care is taken.
Total Comments 0
