New Class D chip fro Analog Devices

[AAK]

Here's the pdf.

http://www.analog.com/UploadedFiles/..._2_4_6_pra.pdf

[jan.didden]

Yeah, saw it at EDN:

"Jeff Berman, News Editor -- EDN, 1/26/2005


[snip]The new amplifiers, AD1990, AD1992, AD1994, and AD1996, deliver audiophile sound quality with less than 0.0005% total harmonic distortion plus noise, as much as 40W of output power per channel with at least a 30% heat reduction, and a dynamic range greater than 101 dB. The amps comprise power-stage-design capability, advanced sigma-delta modulator, closed-loop-feedback topology, and advanced mixed-signal and power and digital CMOS processes.

"Using sigma in a closed loop gives the best performance, and closed loop is important because it gives nonlinearity in output," Slattery says. "In the past, many Class Ds have used an open loop, which is self-limiting after a certain point." These amplifiers are instrumental in fixing design issues in HDTV and home-theater systems with lower EMI, resulting in lowered interference with other electronic equipment. The amplifiers are currently available for sampling, and full production quantities will be available in April. Sample prices range from $2.90 to $6.72 (1000)."

Jan Didden

[tiki]

Closed loop... before output filter!
The datasheet does contain only one-frequency plots or sweeps, I could not find IMD or similar plots.
The other data look impressive indeed, except the efficiency.
Timo

[phase_accurate]

The other data look impressive indeed, except the efficiency.

You put together the the advantages and disadvantages of delta-sigma modulation class-d amps in one short sentence.

BTW: Post-filter NFB is also possible with SD architectures. Most IC manufacturers' target market is multimedia anyway, where no one cares about such details like post-filter NFB.

Regards

Charles

[classd4sure]

Hello

BTW: Post-filter NFB is also possible with SD architectures. Most IC manufacturers' target market is multimedia anyway, where no one cares about such details like post-filter NFB.

Which brings to mind Sanyo and their IcePower IC... due for release in 2003? Doesn't exist yet does it? I wonder if they're running into the same problems Mueta are having, but I would think Sanyo would be most willing to flood the market with lower power versions.

Regards
Chris

[phase_accurate]

Attached the THD+N of a hypothetical SD amplifier with a 4th order loop, 8 Ms/s sampling rate and post filter NFB takeoff.

Regards

Charles

[phase_accurate]

And here is the IMD for 19 + 20 kHz, each 20 dB below clipping. As one can see expected performance could be excellent, but the high switching frequencies involved would be detrimental to efficiency.

Regards

Charles

[dmfraser]

I have had an amp board with the Sanyo/ICE Power chip in my hands, at the Las Vegas Hilton during CES, in the ICE Power suite and again in New York City during AES. But the delivery date seems to keep slipping. I'm promised samples now some time in 2Q 2005. Meanwhile we keep on using the 100A and 500A.

[phase_accurate]

Hi Dan

Do you know if they will be available to ordinary people (unlike B&O's modules) ?

Regards

Charles

[classd4sure]

Hi Dan,

Thanks for that info, what kind of power levels were these amps achieving?

I have a feeling the average Joe won't be able to get them, but eventually we'll be seeing another Sonic Impact like thread.

Regards
Chris