Why use Class AB amplifier modules in Active speakers when Class D already delivers?

[danishchelsea]

I have understood that modern Class D designs already deliver enough power for all requirements, many of the recommended amplifiers are Class D (Eg: Boxem)

But why do the speaker companies who make Active speakers like KEF, Genelec etc go for Class D for Bass drivers and Class AB for Tweeter drivers? Any engineering reason behind this?

 

[-Matt-]

Look what happens to distortion for high frequencies...

(Class AB, Topping B200)

(Class D, Boxem A 4216/E2, Purifi 1ET6525SA)

 

[RPoitou]

Watching this with great interest to see what answers you are given ........

 

[NTK]

KEF's reasons are given in section 3.4.1 of the LS60 W white paper.
https://assets.kef.com/product-support/ls60-wireless/LS60W_Whitepaper.pdf

[Edit] BTW, Genelec uses all class-D amplifiers for their "One" series studio monitors.

 

[tmtomh]

KEF's reasons are given in section 3.4.1 of the LS60 W white paper.
https://assets.kef.com/product-support/ls60-wireless/LS60W_Whitepaper.pdf

[Edit] BTW, Genelec uses all class-D amplifiers for their "One" series studio monitors.
View attachment 483674

Interesting - thanks! KEF's explanation makes sense: they say the woofer and midrange generally require sustained power, as the crest factor at those frequencies in most musical signals is low to moderate. In contrast, they say the crest factor of higher frequencies in most musical signals is high - that is, relatively little power is required at the higher frequencies, but there's more power required for instantaneous peaks. One thing we know is that Class D amps tend to have gobs of thermally efficient power but little dynamic headroom, while Class AB amps tend to be less thermally efficient but have gobs of dynamic headroom. So if the typical continuous power demands at high frequencies are low, the thermal penalty of Class AB is minimized, while if the instantaneous headroom demands at high frequencies are significant, then the headroom of Class AB comes in handy.

I hasten to add that I have no idea if there is actually any benefit in practice to the KEF LS60's 100w AB tweeter amp with its headroom versus the Genelec 8351b's 150w Class D tweeter amp with its higher power capacity but probably no real headroom. I suspect in practice it's six of one, half-dozen of the other.

And it's also interesting to me that, if I read the white paper correctly, the LS60 uses 100w of Class D amplification for the midrange driver - given what they say in the white paper, that seems very strange since by their logic the midrange driver should require more continuous power than the tweeter. But perhaps I've read the white paper incorrectly in that respect?

 

[-Matt-]

And it's also interesting to me that, if I read the white paper correctly, the LS60 uses 100w of Class D amplification for the midrange driver - given what they say in the white paper, that seems very strange since by their logic the midrange driver should require more continuous power than the tweeter. But perhaps I've read the white paper incorrectly in that respect?

My reading of it is that... The MF driver does need more continuous power. They deliberately spec the HF amp with more continuous power than needed, to ensure that it can meet the peak power requirements... In order to keep the high frequency distortion low.

The reason behind the high-power specification for both MF and HF amplifiers, especially the HF one, is due mainly to the crest factor of music at the intended frequency passband. Ensuring the amplifiers have the rail voltage available to supply the power to amplify a sudden sharp peak in the signal is crucial to avoid harmonic distortion.

 

[patoulol]

I imagine that it is inaudible to have a square signal of class D, so, we use class D

 

[-Matt-]

The OP asked about engineering reasons...

Sometimes the engineer is targeting the highest possible technical performance (even if inaudible).

Sometimes the target is sufficient performance to meet the spec at the lowest possible cost.

Different performance targets will result in different hardware selection.

 

[voodooless]

Look what happens to distortion for high frequencies...

View attachment 483672
(Class AB, Topping B200)


View attachment 483673
(Class D, Boxem A 4216/E2, Purifi 1ET6525SA)

I very much doubt that these AB amps used in these active speakers have a better than 85 SINAD at any frequency.

The only reasons to use an AB amp is marketing story and cost…

 

[-Matt-]

But why do the speaker companies who make Active speakers like KEF, Genelec etc go for Class D for Bass drivers and Class AB for Tweeter drivers? Any engineering reason behind this?

This also prompts the question...

For passive speakers with separate HF and LF binding posts is it interesting to explore bi-amping with class D for LF and class AB for HF?

(Or does the crossover circuit somehow negate the potential advantages even when the links are removed)?