Split rail supplys for Class D amplifiers?

[eem2am]

Why do Class D amplifiers always run off spilt rail supplies which are centred on 0V (eg -24V, 0V, +24V)?

Is this simply because 50Hz mains transformers can give a greater output voltage if they are set up with split secondarys?

[wwenze]

Not a class D problem; audio amplifiers usually run off split supply.

A simplified explanation is, because the waveform of music goes both positive and negative, to amplify the negative parts, you need a negative power supply.

[Andy5112405]

Any amplifier Class A or B or C or D that runs off a single supply will have a DC potential at its output neccessitating a large capacitor to isolate the speaker from the DC.

If a split rail is used then the speaker output is at 0V so no capacitor is needed.

Most amplifier designs are currently Push-Pull using +/- supplies.

[eem2am]

-but if you don't use a series capacitor (with split rail) then you could get flux walking,..followed by a big bang

[Andy815408]

Please have a read of Rod Elliots' site httP:\\Elliott Sound Products - The Audio Pages (Main Index)

He has written many articles in starightforward English to explain how various amplifiers work and what their power requirements are.

I've never heard of "Flux Walking"

A +/- supply has a centre voltage of 0V. A single + supply has a centre voltage of Vcc/2 (Half the supply voltage).

You are trying to achieve 0V across the speaker when there is no signal applied.

In a +/- supply it is already 0V so no special care is needed. In a single + supply the Vcc/2 must be blocked with a LARGE capacitor.

[eem2am]

even though a split supply, in the ideal world, would give zero DC current in the speaker.........

in the actual world, this does not happen........pulses going positive outweigh those going negative and you end up with a DC component in that speaker coil.

And we all know that class d's always run with one half of the split rail using more power than the other half....

....so flux walking here we come.

-it can also be avoided /protected against if you have speaker coil current sensing.

[Andy815408]

OK. But you still cannot use a single + supply in place of a split +/- supply.

The analogue parts of the circuit will still need +/- if that was the intention of the designer.

I don't doubt that there may be Class D amps out there that use only one supply but stick with the spec. sheets.

Personally I don't like Class D so I'm not an expert, my answers were in relation to split power rails in general.

[!]

Whether you need the output coupling capacitor depends on the specifics of a particular design regardless of whether it is class D or an analog amp

Class D amps don't always run off split rail supplies. To clarify, some amp chips create their own virtual ground between positive and negative rails from the PSU then the speaker return is not at electrical ground potential. You can see an example on page 5 of this datasheet: http://www.audiolabga.com/pdf/TA2024.pdf

50Hz transformers output no difference in voltage because of a split in the secondaries versus one continuous secondary (if all else is equal and you sum the two secondaries' voltages to compare).

It is not true that there is a significant DC current and a "big bang" without a series capacitor, assuming the input to the amp is AC w/o any DC component to it, unless the amp has some other design problem. However, it is always prudent to measure the amp output for DC voltage before hooking a load (speaker) up to it.

Even if a chip spec sheet shows a circuit with PSU having positive and ground input to the amp, you could still use a split supply with positive and negative to the amp and ground (0V) return to the PSU being common ground to the speakers, that could eliminate the need for an output coupling capacitor in "most" designs.

In some class D amps, the output is monitored for DC offset and a compensation circuit feeds an equal but opposite polarity DC offset to center the output at 0V again.

eem2am, you seem to be stating a lot of misinformation as if it were factual which tends to derail a topic.

[Frank Berry]

eem2am,

Flux Walking?
Loudspeakers have been handling asymmetrical waveforms for decades.
That doesn't mean that there is DC across the loudspeaker.
Just because the waveform has a higher amplitude in one direction than the other (on some, but not all recordings) doesn't mean that there is some "residual" DC on the loudspeaker.