why delta sigma class D

[terminator12]

Delta sigma class D amplifiers will have a finite noise floor at audio frequencies due to the quantization process (quantization noise). Whereas a PWM-Class D does not suffer from any quantization noise (because the pulse widths are continuous in an analog modulation scheme). Therefore PWM Class D is more linear than the Delta sigma ones.

Then, (apart from the EMI problem mitigation) why at all was Delta sigma class D modulation invented?

thanks,

 

[Andrew Eckhardt]

I imagine because by the time you make a conventional PWM modulator that beats the sigma-delta converter it's more expensive than a sigma-delta converter.

 

[AP2]

I imagine because by the time you make a conventional PWM modulator that beats the sigma-delta converter it's more expensive than a sigma-delta converter.

Hi,
yes, add that it is very difficult to get very good performance.
need a lot of knowledge and a lot of development to solve some problems of PWM Class D
But result is very very good, intermodulation is not present at the midle high-frequency sound.

 

[terminator12]

Hi,
need a lot of knowledge and a lot of development to solve some problems of PWM Class D

Could you tell me what some of those problems asosciated with PWM-Class D's are which are yet to be understood clearly?

Also I thought we need a much higher switching frequency for a Delta sigma Class D when compared to PWM ones thereby causing an efficiency loss.

thanks,

 

[phase_accurate]

First of all: Noise and distortion is not the same. The thread-starter made some mix-up here.

For any power amp from some dozens of Watts upwards I don't see much advantage in using delta-sigma modulation.

But for small power applications there are advantages you can profit from (well - they actually exist for large power as well but the high idle-switching frequency makes delta-sigma a little impractical for large power applications):

Delta-sigma can be easily integrated (switched-capacitor topologies may be used).

A guaranteed minimum on-time of the output devices eases MOSFET driver design.

Dead-time can be implemented digitally with high series consistency.

Delta-sigma does spread EMI more evenly than PWM.

Regards

Charles

 

[AP2]

Could you tell me what some of those problems asosciated with PWM-Class D's are which are yet to be understood clearly?

Also I thought we need a much higher switching frequency for a Delta sigma Class D when compared to PWM ones thereby causing an efficiency loss.

thanks,

Hi,
First of all I refer to a class D PWM with professional features.
(Or simply is developed)
first problem is the triangular wave. must be geometrically perfect but should not have spurious high frequencies in the corners, this is not easy to obtain. maximum percentage of modulation, harmonics and noise depend on it.
Another problem is the circuit after the comparator and the driver.
I see that many do not call attention in this circuit, relying on chip manufacturers.
so it is more effective if those who design the amp, then the driver accurate designs with the features you need.
Another problem to be solved is the feedback, need 2 fb. then repeat that we need a good knowledge. simulators certainly help but it is likely that takes us off the road, he only mathematically analyzed in the circuits ...as well as manufacturers of chips ...

Regards

 

[AP2]

First of all: Noise and distortion is not the same. The thread-starter made some mix-up here.

For any power amp from some dozens of Watts upwards I don't see much advantage in using delta-sigma modulation.

But for small power applications there are advantages you can profit from (well - they actually exist for large power as well but the high idle-switching frequency makes delta-sigma a little impractical for large power applications):

Delta-sigma can be easily integrated (switched-capacitor topologies may be used).

A guaranteed minimum on-time of the output devices eases MOSFET driver design.

Dead-time can be implemented digitally with high series consistency.

Delta-sigma does spread EMI more evenly than PWM.

Regards

Charles

Hi,
only see an advantage in the delta-sigma.
development costs very low
production cost very low.
disadvantages:
definition sound at high frequency is very low. (After 60% volume)
a lot of IMD especially in the midle range (voice trumpet etc). proportional to power.

I do not understand how can mix the "inventions" that are only for the business market with the audiophile music lover.

regards

 

[phase_accurate]

So we do basically agree, don't we ? The four points that I mentioned are definitley advantages for the manufacturer.

While you may personally strive to get the best sounding class-d amp ever (which is of course a honest and reasonable goal) you don't have to forget that high-quality class-d amps are only a veeeeeeeery small part of the class-d market (in terms of unit count and money). By far the largest part are low power mass-market devices.


Regards

Charles