Sorry about that one.
I'm kind of saying it is feasible with the right approach as long as you don't mind a few smoking amplifiers and popped driver cones on the journey.
I'm kind of saying it is feasible with the right approach as long as you don't mind a few smoking amplifiers and popped driver cones on the journey.
Note how image b from post #2 puts the speaker directly in the feedback loop, so it's an uncontrolled and unpredictable situation. The amplifier is trying to keep itself linear, probably up to a few MHz, and replacing a few mm of PCB traces with metres of cables could make things interesting. An oscilloscope is mandatory equipment for these kinds of mods.
+1 for a Pass F1 style amplifier. An output stage in "open drain" so the decices naturally convert a voltage to current, even without feedback.
It doesn't though. The feedback point is moved from the output at the top of the speaker to the top of the current sense resistor. If you compensate it correctly the amplifier will try to drive a linear, proportional, current through the speaker in the same way it would have tried to place a linear, proportional, voltage across the speaker. I don't disagree that layout will be an issue and also you might get popped cones at speaker resonances but, again assuming you get the compensation right, it will do what you want it to do.
@abstract
The simple resistor method already has some big impact on the quality of sound.
Did not know that the loss calculation looks to favorable.
I thought that if you have a loudspeaker with 4 ohms impedance and you add a four ohm resistor in line you loose half of the amplifiers wattage.
However impedance is not linear so effectively it's less loss.
The simple resistor method already has some big impact on the quality of sound.
Did not know that the loss calculation looks to favorable.
I thought that if you have a loudspeaker with 4 ohms impedance and you add a four ohm resistor in line you loose half of the amplifiers wattage.
However impedance is not linear so effectively it's less loss.
The trick to find out how to modify a mass produced cost effective amplifier is a great thing.
The Behringer in its dsp version has an easy programmable dsp and is good to handle due to its light weight and providing plenty of power.
@MorbidFractal
I am so curious and grateful to hear a modification is not out of reach.
The rest would be a try to implement such a modification. Besides the important theory still a (long) way to go.
The Behringer in its dsp version has an easy programmable dsp and is good to handle due to its light weight and providing plenty of power.
@MorbidFractal
I am so curious and grateful to hear a modification is not out of reach.
The rest would be a try to implement such a modification. Besides the important theory still a (long) way to go.
Last edited:
How about sampling the current from the inductor L.. Wait are we entering a current mode control modulator with intentional self-oscillation i.e omitting the TYPE-II/III pole/zero "prevention of system oscillation".
Lateral thinking: you can always fake the constant current behavior by looking at the impedance curve and applying via DSP (or parametric equalizer) the exact same EQ, meaning resonance boost and steadily rising above 1 kHz
Your ears won't be able to tell the difference.
I have successfully done that in bridged output Guitar amps, where constant current NFB is difficult to impossible to get.
Also applies to Class D amps, where constant voltage is the norm.
Your ears won't be able to tell the difference.
I have successfully done that in bridged output Guitar amps, where constant current NFB is difficult to impossible to get.
Also applies to Class D amps, where constant voltage is the norm.
@JMFahey
Right - dsp is a mighty tool.
But besides frequency response, distortion reduction due to adding a copper ring to the pole piece can be heard if you DIY, here the 50 cent trick
https://www.diyaudio.com/community/threads/visaton-bg20-impressive-mod.408081/post-7576162
Right - dsp is a mighty tool.
But besides frequency response, distortion reduction due to adding a copper ring to the pole piece can be heard if you DIY, here the 50 cent trick
https://www.diyaudio.com/community/threads/visaton-bg20-impressive-mod.408081/post-7576162
Only pure analog is required for low THD outcome, all these DSP grand high cycle count solutions don't exist and isn't capable of outperforming current solutions contextually applied to high speed control loop response making response corrections. So no.