Class -D audio amplifier

patent 6,084,450

I need to design a class d audio amplifier, but I only konw Class AB, and know little about of class D, anyone have any good suggestion or kindly give me some relative Class D material.
thank you very much.

class-d_audio_ii_evaluation_board

I'm desing class d, but this desing is made of a Integrated circuit

Give details. How is ur experience after assemble.

It seems Class D power amplifier is hot now.
However in high frequency domain Class D PA is seldom used. Why?

a demo board:

Hi,
Class D PA is seldom used in high frequency domain because of the THD is bigger than it is in low frequency domain.

ynhe

alphi said:

a demo board:

Click to expand...

Direct link:
/http://wwwlea.upb.de/lehre/Projektseminar/Class-D_Audio_II_Evaluation_Board.pdf

Are you sure about that ynhe? I thought it was because the losses went way up (efficiency goes way down) which removes the advantage of going with class d.

currently several integrated chips with switching frequency of 500KHz to 1 MHz are available like those from IRF or PHILIPS, the provide over 80% efficiency and very low THD (better than 0.1 %), of course thy are more noisy and have lower performance and worse THD at higher frequency but they have not cross over non-linearity like class AB amplifers and at lower frequency (lower than ~1KHz) there is no comptitors at all thus the are used in sub woofer drivers with very high power and very high fidelity sound.
in portable applications like headphones, mobile handsets or laptops they are prefered to class AB apmlifiers due to more battery life.

BEST!

Go to china www.21ic.com forum,you can find several class-D amplifier schematics.and much other articles.

I recalled M. Smith and K. Smedley at UC-Irvine had some low THD class D audio amplifier. The control scheme is called 'one cycle control'. Paper is published in IEEE tran Power Electronics.
Check it out. It's so easy to be implemented in IC.

xshou said:

I recalled M. Smith and K. Smedley at UC-Irvine had some low THD class D audio amplifier. The control scheme is called 'one cycle control'. Paper is published in IEEE tran Power Electronics.
Check it out. It's so easy to be implemented in IC.

Click to expand...

Hi, could you upload the file or tell me the paper name.

woops, how can I delete this repost

here is the link to power electronics lab at UCI
http://www.eng.uci.edu/pel/pel.html

their publications:

Z. Lai and K. Smedley "A New Extension of One-Cycle Control and its Application to dc-ac converters," IEEE Transactions on Power Electronics, Jan. 1996, Vol. 11, No.1, P99-104.

M. Smith and K. Smedley, "A New PWM Controller with One-Cycle Response", IEEE Transactions on Power Electronics. Jan. 1999.

M. Smith and K. Smedley, ìPWM Controller with One-Cycle Response,î US Patent 6,084,450, July 4, 2000.

Mark Smith founded a company turning it into product:
http://www.powerphysics.com/
they have pretty good white paper:
**broken link removed**

Recent effort to implement one-cycle-control in IC level, see:
An integrated one-cycle control buck converter with adaptive output and dual loops for output error correction
Dongsheng Ma; Wing-Hung Ki; Chi-Ying Tsui;
Solid-State Circuits, IEEE Journal of , Volume: 39 , Issue: 1 , Jan. 2004
Pages:140 - 149

Class D audio amp needs several possible improvements:
1. distortion: by higher modulation freq(stupid), by pre-disortion in PWM (I think about this idea for a while but just don't have time to rigorously pursue it myself; given the possibility of either in digital or analog domain, you can write quite a few papers), by white noise injection (my other interesting idea, try to lower harmonic), or by multiple level modulation(unipolar, bipolar, tri-state, quad, ....), etc.
2. response: fast, it's more about control scheme. You can design numerous nonlinear controllers, how about fuzzy controller? or sliding mode controller?
3. size: reduce bulky external conponents, check Virginia Polytech Power Electronics Center recent work on modulization and high integration of power parts (active and passive). If possible on chip level.
4. efficiency: not so important in class D, but just as "usually" mentioned by professors. You can try some new softswitch topology.

I think if you can combine all these, pretty much a Ph.D. dissertation is ready. Good luck.

It's so hot here.Certainly class-d is becoming a hot topic now.
I have designed a simple open-loop class-d amplifier in CMOS process,the function is ok.
I plan to design a close-loop one and hope to get help from all of you.
I am also glad to provide my hands.

Are there any EAD tools dedicated to power elctronics IC design? We usually use Cadence as our analog IC design tool. But, what about power elctronics IC design. Thanks.

ynhe said:

Hi,
Class D PA is seldom used in high frequency domain because of the THD is bigger than it is in low frequency domain.

ynhe

Click to expand...

The reason we can get high efficiency of class D power amplifier is, the power transistors operate in switching mode(similar with switched mode power supply).

But in high frequency domain with high power, this condition is difficult to realize.
On the other hand,if the high frequency transistors operate in switching mode, their nonlinear performance will be very bad. Much more intermodulation products will be produced. That's unacceptable.

Class D amplifier generates a series of square wave with varying duty ration depending of the value of the input signal. If your desired signal has the frequency which is slightly lower than the switching frequency, let's say your signal frequency is 20kHz and the switching frequency is 40kHz, the performance of the amplifier will be bad and large attenuation is needed to suppress the switching signal. That's why the performance is bad in high frequency.

go to this link:
http://www.geocities.com/CapeCanaveral/9096/index.htm

carrier said:

go to this link:
h**p://www.geocities.com/CapeCanaveral/9096/index.htm

Click to expand...

Thanks, but I can't open this link, could you check it?