Boost converter- dynamic output

[PurvR]

Hey everyone I am building a DIY speaker powered by batteries. I'm using lithium Ion battery and boosting the voltage to 12 volt using xl6009 boost converter. The amplifier is 10 watt class d amp and works form voltage range of 7 volts to 15 volts. I was wondering if the boost converter can be modified to change the output voltage according to increase in load as the volume increases, like as I increase the volume the voltage of the boost converter also increases. I think that it'll be more efficient use of the battery, as there is no need for 12 volts for lower volume and as I increase the volume the voltage also increases accordingly.
I can't seem to figure it out how to design such circuit, so I need help.

[David Hess]

It is definitely possible to do and similar to how class-H amplifiers work.

For simplicity I would peak detect the input signal and use that as the reference for the regulator which multiplies it to the final supply voltage, with a fast attack and slow decay time.

[NiHaoMike]

For simplicity I would peak detect the input signal and use that as the reference for the regulator which multiplies it to the final supply voltage, with a fast attack and slow decay time.

You'll need a pretty significant delay in the main signal path for envelope tracking to work well and the savings for such a small amplifier would be largely cancelled out by the power needed to run the logic. Adjusting the supply voltage based on volume setting works nicely. That can be done with another potentiometer on the volume control or program the microcontroller handling the volume control to also set the voltage.

[trobbins]

I would think you need to additively combine the XL6009 reference level with an increment that is derived from the signal amplitude (as David Hess indicates).  That way if there is no signal then there is no change to the base level regulated output voltage (regulated for both input and output level changes).

However the XL6009 has an internal reference that you can't access, which makes your request a bit trickier to implement.  One option is to use a signal based voltage or current supply to pull down the feedback level to the XL6009, and fool it to raise the regulated DC output beyond what it does for line and load regulation control.

[David Hess]

For simplicity I would peak detect the input signal and use that as the reference for the regulator which multiplies it to the final supply voltage, with a fast attack and slow decay time.

You'll need a pretty significant delay in the main signal path for envelope tracking to work well and the savings for such a small amplifier would be largely cancelled out by the power needed to run the logic.

I would just accept clipping on the first peak, with a relatively long decay time constant.

[Zero999]

Hey everyone I am building a DIY speaker powered by batteries. I'm using lithium Ion battery and boosting the voltage to 12 volt using xl6009 boost converter. The amplifier is 10 watt class d amp and works form voltage range of 7 volts to 15 volts. I was wondering if the boost converter can be modified to change the output voltage according to increase in load as the volume increases, like as I increase the volume the voltage of the boost converter also increases. I think that it'll be more efficient use of the battery, as there is no need for 12 volts for lower volume and as I increase the volume the voltage also increases accordingly.
I can't seem to figure it out how to design such circuit, so I need help.

Just one question: why?

You're using a class D amplifier, which is pretty efficient at lower output volumes, even with a relatively high power supply voltage. The additional complexity of altering the supply voltage will gain you very little effiency-wise and modern class D amplifier ICs often need a fairly stable supply voltage, so it could cause instability and more distortyion.

This sort of technique greatly boosts the efficiency of a class-AB amplifier, when if forms a class H amplifier, but is makes no sense for a class D amplifier.

[NiHaoMike]

You're using a class D amplifier, which is pretty efficient at lower output volumes, even with a relatively high power supply voltage. The additional complexity of altering the supply voltage will gain you very little effiency-wise and modern class D amplifier ICs often need a fairly stable supply voltage, so it could cause instability and more distortyion.

It does make quite a difference in quiescent power. I would say to implement it based on volume setting, with 3D printed gears to connect the extra potentiometer. That will make it easy to select a different value for the voltage control.

EDIT: You can also use noncircular gears to change the voltage vs. volume curve, but that starts getting into mathematics and mechanical engineering.

[Someone]

You're using a class D amplifier, which is pretty efficient at lower output volumes, even with a relatively high power supply voltage. The additional complexity of altering the supply voltage will gain you very little effiency-wise and modern class D amplifier ICs often need a fairly stable supply voltage, so it could cause instability and more distortyion.

It does make quite a difference in quiescent power. I would say to implement it based on volume setting, with 3D printed gears to connect the extra potentiometer. That will make it easy to select a different value for the voltage control.

Except we don't know what the amplifier part is or its operating parameters are. H-bridges (class D amps) can be more efficient at higher supply voltages, it all depends on the specifics.

[Zero999]

There's also the quiescent current required by the boost converter. It might make more sense to just disable the boost converter and run directly from the batteries, at low volume settings. It could be implemented with cam and suitable switch on the volume potentiometer. The main downside is there will be a click, both from the speakers, as the supply voltage changes abruptly and obviously the switch.