@john k... You've got yourself a live one here. 🙂 He's been on the forum a week and already 160 posts enlightening everybody. 🙂
Dave.
Dave.
That is because distortion is measured and plotted by the nominal loading of the speaker.
In current out amplifiers, like class D, you have a critical impedance loading which is the lowest load it can handle stably. When you rate an amplifier's loading range, they use the 3 to 5 times rule, that is applied to the sizing the coupling used. So, for instance, an amplifier that has a critical impedance of 2 ohms, its nominal impedance is 6 to 10 ohms. Outside of this range, the coupling device plays a role as its capacitive reactance (its dynamic resistance in AC across frequencies) increases due to the function of the RC circuit formed from the coupling capacitor and the speaker load. This of course, changes the bandwidth of power transfer from the amp to the speaker.
Different amplifier topologies were developed to address distortion and power delivery in the desired operational range to provide better handling of the variable load.
This is a headphone amp.
This is comparing line an operational amplifier with different load impedance:
Try reading something authoritative on class D amps where you will see that the loudspeaker damps the output filtering
https://www.analog.com/en/technical-articles/fundamentals-of-class-d-amplifiers.html
Also go see what the type of feedback arrangement that might be used works and does.
Last edited:
I tried the constant LC ratio on the 1mH on the tweeter. Works well. Leaving the R's alone 47u and 47mH improves things further. Thought it might do from RF factors. Higher L values can mean higher Q but the penalty can be it's series resistance. If the 10ohm gets to 1 the tweeter goes below 4ohm. At 100 ohm it just doesn't work well. No signs of ringing anyway.
This behaviour might help if I ever need a notch filter.
This behaviour might help if I ever need a notch filter.