I'm trying to understand the theory behind coupling capacitor value selection, specifically the desired -3 dB rolloff frequency. I understand that too low of a value can lead to negative effects, and too high of a value obviously will hamper bass response. Within that range, is there any particular logic for aiming higher or lower? Is the salient factor if and where you want to insert a rumble filter?
Here’s an online calculator, simply enter the connected amplifier’s load resistance. The calculated capacitor value is only a rule of thumb. Whether you can use a smaller than calculated value, only you can judge by listening.
https://v-cap.com/coupling-capacitor-calculator.php
https://v-cap.com/coupling-capacitor-calculator.php
For electrolytic coupling capacitors, use a 2Hz corner frequency to minimize the distortion they can generate.
For film types, decide at what frequency a five degree phase shift is acceptable, and then set the pole
at ten times lower in frequency. Bear in mind that there is a 45 degree phase shift at the -3dB corner
frequency of the dominant pole.
Some circuits can become unstable at low frequencies if the poles are set too low, or too close together.
Both the source impedance and the load impedance enter into the frequency calculation.
A rumble filter will normally need to have a sharper cutoff than first order, and so will cause more phase shift.
For film types, decide at what frequency a five degree phase shift is acceptable, and then set the pole
at ten times lower in frequency. Bear in mind that there is a 45 degree phase shift at the -3dB corner
frequency of the dominant pole.
Some circuits can become unstable at low frequencies if the poles are set too low, or too close together.
Both the source impedance and the load impedance enter into the frequency calculation.
A rumble filter will normally need to have a sharper cutoff than first order, and so will cause more phase shift.
Last edited:
As you are asking this in the tube/valve section: it is usually low frequency stability (preventing motorboating) that determines how low you can go. You have to analyse any feedback loop that has more than one first-order high-pass section (that is, AC coupling or transformer) in it and check that you don't get too much phase shift at the point where the loop gain drops to unity. That usually means that one of these section has to cut off at a substantially higher frequency than the rest. Besides, you have to check unintended feedback via the impedance of the power supply.