Question:
Is there is any current mode (high output impedance) class d IC at all by a reputable manifacturer or are they all voltage sources with low output impedance?
Background:
I want a current mode / current source / transconductance amplifier: I.e. high output impedance of 200 ohms or more.
Class D efficiency would be nice and I don't need a lot of power, 10W @ 8 ohms is plenty so ICs like the TDA8932 would be perfect if it wasn't a voltage source.
I have tried to wrap a current feedback loop around the TDA8932 but it doesn't work. I can push the output impedance up to ~ 10-20 ohms but it becomes unstable. It rests on the knife edge where if I change the output resistance or use a bit more feedback it starts to oscillate at full power.
Is there is any current mode (high output impedance) class d IC at all by a reputable manifacturer or are they all voltage sources with low output impedance?
Background:
I want a current mode / current source / transconductance amplifier: I.e. high output impedance of 200 ohms or more.
Class D efficiency would be nice and I don't need a lot of power, 10W @ 8 ohms is plenty so ICs like the TDA8932 would be perfect if it wasn't a voltage source.
I have tried to wrap a current feedback loop around the TDA8932 but it doesn't work. I can push the output impedance up to ~ 10-20 ohms but it becomes unstable. It rests on the knife edge where if I change the output resistance or use a bit more feedback it starts to oscillate at full power.
Why such an impedance ..... It is to sound a store?
It might be necessary to go through a transformer at the output of the amplifier.
But not knowing the impedance of your speaker it will be difficult.
If you connect them in parallel ... Impedance decrurat ....
2 hp of 200 ohms in parallel gives 100 ohms, 4 hp 50 ohms and so on.
Up to you .
It might be necessary to go through a transformer at the output of the amplifier.
But not knowing the impedance of your speaker it will be difficult.
If you connect them in parallel ... Impedance decrurat ....
2 hp of 200 ohms in parallel gives 100 ohms, 4 hp 50 ohms and so on.
Up to you .
I want a high output impedance because it reduces the distortion produced by my speakers. Here are some measurements with just a series resistor so limited output impedance of 32 ohms. Note specifically the orange third harmonic which falls significantly from 2 khz and upwards. If I use a large series resistor it defeats the purpose of using an efficient class d amp though.
There are working circuits with open loop class AB chipamps like the LM1875, but then I don't get the nice class D efficiency.
Attachments
In my case a CBT shaded array of 24x SB65WBAC25-4. Ideally I'd want cheap enough amps I can build 8 amps for each side with 3 speakers in series each. The effective impedance per bank would be 16 ohm but 2 watts of power for each speaker is plenty hence 6W @ 12 ohm ~ 10W @ 8 ohm with respect to max voltage.
I would assume for such a high output impedance.
The output inductors of the amplifier would need to be changed.
For such low wattage a inductor less design could work
switching frequency is very high. likely using GanFets
Definitely no expert, far as I know whatever opamp is used
typically needs to be very stable at unity gain.
And the bandwidth well above 400 kHz
when current sensing a high frequency amplifier.
just a guess the instability limits your hitting lies within
the opamp capabilities
The output inductors of the amplifier would need to be changed.
For such low wattage a inductor less design could work
switching frequency is very high. likely using GanFets
Definitely no expert, far as I know whatever opamp is used
typically needs to be very stable at unity gain.
And the bandwidth well above 400 kHz
when current sensing a high frequency amplifier.
just a guess the instability limits your hitting lies within
the opamp capabilities
Hi,
I recall you had a nice insurance based amp class A design, what was wrong with that?
If purpose of doing this project is about going green, I don't think there is that much mileage considering your amplifier power consumption is only a few 10s of watt, and you are not listening to music 24 hours a day.
Back to your topic, afraid I don't know. But class AB can be made current source by a modification to its feedback loop via a current sense resistor. It becomes sensitive to connection issues and it could send tehe amplifier into oscillation if there is poor contact.
I am also interested in this topic...
Oon
I recall you had a nice insurance based amp class A design, what was wrong with that?
If purpose of doing this project is about going green, I don't think there is that much mileage considering your amplifier power consumption is only a few 10s of watt, and you are not listening to music 24 hours a day.
Back to your topic, afraid I don't know. But class AB can be made current source by a modification to its feedback loop via a current sense resistor. It becomes sensitive to connection issues and it could send tehe amplifier into oscillation if there is poor contact.
I am also interested in this topic...
Oon
Current feedback around standard class-D chips can be done using a magnetic current transformer that monitors speaker current. The output on the secondary is injected into the feedback loop. Did this some 10 yrs ago. If memory serves me the max stable impedance was in the ballpark of 100 Ohms or lower.
Yup, I do have a few working class A amps which I can use as fallback if nothing else. It's fun to explore different options though. And I want to compare and see if I liked the sound of my class As more than my class Ds beacuse they were current sources or if I just like class A more.
I have tried the following circuits, I might have messed up the polarity on the feedback in the pictures but ignore that. Both transformer and the opamp version work but when I push the output impedance higher and higher it becomes crazy unstable. I also tested if the lack of gain is the problem by adding another OPA1632 in front of the TDA8932 but that wasn't the problem.
The problem are the high frequencies, if I limit the feedback to lower frequencies by bypassing the current RSense with a capacitor then I can use more feedback and get output impedance to 40-50 ohm @ 1 khz before it becomes unstable again.
But yeah, if it is this unstable then the inefficient class A amps are probably a better option, they aren't as likely to kill my speakers if as much as breathe on the amps =)
I have tried the following circuits, I might have messed up the polarity on the feedback in the pictures but ignore that. Both transformer and the opamp version work but when I push the output impedance higher and higher it becomes crazy unstable. I also tested if the lack of gain is the problem by adding another OPA1632 in front of the TDA8932 but that wasn't the problem.
The problem are the high frequencies, if I limit the feedback to lower frequencies by bypassing the current RSense with a capacitor then I can use more feedback and get output impedance to 40-50 ohm @ 1 khz before it becomes unstable again.
But yeah, if it is this unstable then the inefficient class A amps are probably a better option, they aren't as likely to kill my speakers if as much as breathe on the amps =)
