Power supply for balanced input stereo amplifier

[SMourareau]

I am currently working on a stereo balanced input amplifier project and wanted to share with you an idea I had about power supply.
It is based on following assumption:
- most of the power if used to produce low frequencies
- low frequency signals are mainly the same on two channels
So the idea is to cross the input and output of one of the amplifiers and use the same power supply for both, so that when an amplifier drags current on one supply rail the other drags current on the other rail.
The result is:
- smaller caps
- smaller tranformer
- smaller ripple
- smaller radiation

Regards

 

[AndrewT]

A properly designed balanced amplifier will have a constant current draw on the power supply.
Look up EUVL's F5x for an example.
This constant current draw is very rare among all the various topologies used for amplification.
This constant current draw does give significant performance improvements.

A balanced input for an amplifier does not approach the specification requirements for a constant current draw amplifier.

 

[SMourareau]

Constant current ? That could be possible, even for balanced input, only in case of class A amplifier. My purpose was more generaly about class AB/B amplifiers.

 

[AndrewT]

.............This constant current draw does give significant performance improvements.

A balanced input for an amplifier does not approach the specification requirements for a constant current draw amplifier.

... My purpose was more generally about class AB/B amplifiers.

then you are chasing ghosts as another Member puts it.

 

[CBS240]

A balanced output class AB amplifier has a constant current draw...like this one.🙂

 

[SMourareau]

You can take it all ways you want. balancing input or output or both you will never have a constant draw current with something else than a pure class A amplifier.

Regards

 

[godfrey]

...So the idea is to cross the input and output of one of the amplifiers and use the same power supply for both, so that when an amplifier drags current on one supply rail the other drags current on the other rail.
The result is:
- smaller caps
- smaller tranformer
- smaller ripple
- smaller radiation

Yes, that should give less supply ripple especially with low bass notes. I seem to remember this being done in some commercial amp, might have been NAD?

Starting with balanced inputs certainly makes it easier to reverse one channel.
🙂

 

[AndrewT]

this one.🙂

"this" took me to post66 & the zip in the next.
Where is the schematic?

 

[CBS240]

I misspoke about constant current draw, I meant equal current draw resulting in a zero difference in current flowing from zero volt node (gnd) to each PS cap bank. this is the VAS section, each +/-10V output is for two seperate output followers for BAL output.

 

[ontoaba]

So the idea is to cross the input and output of one of the amplifiers and use the same power supply for both, "so that when an amplifier drags current on one supply rail the other drags current on the other rail".
The result is:
- smaller caps
- smaller tranformer
- smaller ripple
- smaller radiation

Regards

That was bridged amplifier.

-smaller transformer = not really. The power drawn is just same, even bigger because more drop voltages on amplifier (less efficiency). You may get fewer copper needed (smaller secondary cables for winding, because for same power, Vout is doubled, current is half, but power is same.
-smaller caps = yes, for half current like above.
-smaller ripple = no, because the capacitor is also smaller.
-smaller radiation = ??? what radiation.😕

Remember, with same current, same voltages, same power, the load resistance is 4:1 for series and paralelled.

 

[SMourareau]

That is not at all a bridged amplifier

Quietly read again my first post it is not at all a bridged amplifier it is just a way to optimize the use of a single power supply in a stereo amplifier with 2 separate independent outputs.

Regards

 

[magnoman]

Seems to make low frequency sense to me.
Would agree with Godfrey that only ripple would be reduced.

Thanks for the idea,
-Antonio

 

[SMourareau]

Ripple is lower because you discharge both caps from positive and negative rail at the same time.
And for the same reason you also reload these capacitors using both (transformer) secondaries at the same time.
Wich generates lower copper power dissipation
And you may also have a more homogenous current repatition around the transformer.
So lower EMI.

Regards

 

[magnoman]

Yep, I was wrong.
Seems like you would get some reduced resistive losses on the transformer secondaries.

Thanks again
-Antonio

 

[ontoaba]

Ups, my mis understood, sorry.

Still, transformer core should not be reduced, and secondary copper could smaller. The ripple is remain higher if you reduce the caps. And if it is stereo there must be a moment that they draw current not like you wishes.

 

[SMourareau]

You may reduce or not the idea here is to get the best from your tranformer and caps.
Yes I agree it is stereo and the benefit will be significant only if you have low frequencies.
Just because low frequencies are rather mono.
Regards