Why high voltage supply rail for class D amplifier?

[grizedale]

Hi,

Why do Class D amplifier designers prefer high voltage supply rails?


Please find attached here two Full-Bridge , Class D amplifier simulations which run in the free simulator called LTspice by linear.com.

Each simulation is of a 4KHz sine wave which gives 100W of power in the speaker.

One of the Class D amplifiers has an 80V rail and the other has a 40V rail.

Each simulation gives the same output signal (i.e. the same current in the speaker)

……so why do engineers prefer higher voltage rails for Class D amplifiers?……….
……because these simulations show that you can get just the same speaker signal with a lower rail voltage…… and simply using a lower value of filter inductor.



(To run these simulations , you will need to make sure that the text files called “sin” and “sin1” are situated in the same folder as the LTspice simulation files.)

Also, do you know if my representation of the speaker as an 8 Ohm load is correct?…..Does it have an inductance associated with it?…….is this inductance likely to be poorly toleranced.?


The simulation files are in .txt format..........(i had to do this else they wouldnt upload.....anyway, all you need to do to run them in LTspice is save them as .asc files.)


View attachment sin.txt
View attachment sin1.txt
View attachment CLASS D _40V 100W.txt
View attachment CLASS D _80V 100W.txt

 

[FvM]

40 V would be the minimum supply voltage for an ideal class-D amplifier with 100 W power. There are various reasons to provide a margin, e.g. transistor and inductor losses, supply ripple and dynamical variations, keeping a minimal PWM duty cycle to avoid additional distortions, ...

So 50 V would be reasonable minimum supply voltage. Higher values are possibly choosen to for a higher peak output power.

Speaker impedance will be resistive in a first order, some series inductance of voice coil and cable, and complex impedance characteristic at low frequencies representing speaker resonance. Multichannel speakers with passive crossover can have nasty impedance characteristics around the crossover frequencies.

 

[grizedale]

...but its the current in the speaker that matters...who cares what voltage was employed to get that current?

my simulations show that you can get the same current in the speaker with 40v or 80v rail.

So why use 80V?

I dont understand about the margin issue.......there is no more margin issue with 40v than 80v......if you run the simulations you can see this

 

[FvM]

I dont understand about the margin issue.......there is no more margin issue with 40v than 80v......if you run the simulations you can see this

It's a trivial thing. 100 W sine @ 8 ohm = 40 V peak. A real bridge amplifier won't achieve it with 40 V supply. Any supply voltage drop, e.g. controller transient response to load change will cause clipping of output voltage. So at least 50 V supply seems necessary. In case of an unregulated supply, a higher margin is required.