SMPS must be rated twice the class D power

[eem2am]

Hello,

If you are building an SMPS to power a class D amplifier, then you need the SMPS to be rated to twice the maximum average power of the D amplifier......do you agree?

sorry I posted this in Class D forum but realise it would have been better here....sorry....


http://www.diyaudio.com/forums/class-d/218662-class-d-amp-needs-big-power-supply.html#post3142616

 

[dfuhr7gm]

as i know, Class D amps have more than 95% efficiency ... so for a 100W Amp you'll need a 105W SMPS ... take a 120W smps and you're on the safe side

 

[Ouroboros]

Most people get away with an SMPSU that only provides about 50% of the class-D output nominal power, provided that it can provide the full short-term peak power, or has a suitably-sized reservoir cap bank. It is very rare to drive any amplifier to have an average power output of more than 33% of the nominal power rating with any normal programme material.
The class-D amps I design for voice-alarm systems have to meet EN54 part 16, where the amplifier has to be able to drive all channels simultaneously to the full rated load for at least 1 minute with a 1kHz sine-wave at +45 degrees C ambient, so that is a different matter!

 

[marce]

You are better running an SMPS near its capacity, or at least so it does not go into burst mode, otherwise you get allsorts of noise, often into the audio frequecny range.

 

[Ouroboros]

Any properly designed SMPSU will not go into burst mode at any mormal load. It might if there is no load at all. It will drop from continuous mode into discontinuous mode of course but the PWM should work down to quite small duty cycles.

The XP-Power HCP1000 (1kW) SMPSUs I am using are quite happy down to a load of less than 50W for example.

 

[marce]

Sorry ment discontinous mode, just recently seen design where when the SMPS went into discontinous mode there was low frequecny noise generated, that was audible. A biger cap on the output did solve the problem but it is somthing o be aware of.