Zetex new Acoustar ZXCW8100 class D

[sonnya]

The Datasheet is online and it looks very good.

Like PWM frequency up to 2MHz!!!!!! its a bit higher than what TI can come up with . hahahahaha


Actually you can only choose between 529KHz , 1.058MHz and 2.116MHz!

It uses 32bit processing and can recieve data at rate of up to 192KHz

I thought all should now about it

[phase_accurate]

Like PWM frequency up to 2MHz!!!!!! its a bit higher than what TI can come up with

I'd just like to know why they don't state the amp's efficiency in their data-sheet

apart from that: 50 Watts seem to be a little whimpy in my opinion for bothering with all the hassle to implement a class-d amp.

IMHO a discrete class AB amplifier costing the same (taken the numbers DIYers usually build into account) would easily outperform this one.
Keep in mind that it doesen't have any NFB and therfore the power-supply has to be of very high quality.

Regards

Charles

[Brian Brown]

This looks like it could be a really decent chip.

The main thing that I think sets it apart from TI's equibit chipsets is that it handles 32bit audio data, which would interface very nicely to an SACD via a DSD-PCM converter such as the NPC SM5816F.

The advantages of the higher switching frequency aren't clear-cut. Higher frequencies can decrease efficiency and increase distortion as the switching time of the output transistors becomes a bigger percentage of the output cycle. The selection of the output transistors and the overall design of the output stage become more critical as frequency goes up. If this can be done successfully, then the higher frequency will be better. A higher frequency certainly simplifies the output filter.

It should also be noted that the Zetex can only run up to 2.116MHz when it's in its 'conventional PWM' HPWM mode. To improve definition, its RPWM mode skews the timing to the opposite H-Bridge transistors (in a way that sounds similar to Equibit to me). In RPWM mode, the max frequency is 1.058MHz. Furthurmore, it appears that for most output stages that they recommend 529KHz (which isn't that far off from TI's 384KHz).

For PCM with a 192KHz sampling rate with 32bit word size, a direct conversion to PWM would require over an 82THz carrier frequency! (192KHz x 4294967296counts)
Even 44.1KHz - 16bit (44.1KHz x 65536counts) would require 2.9GHz, which is near our technology's current processing limits, forget about output switching devices.
This is why it's necessary to have some type of compensation technique to take the dynamic resolution beyond what simple PWM can do.

At this point from the data sheets, I wouldn't even want to begin to speculate whether the Equibit or Acoustar chip sets will sound better. (But my expectations are that both should be very good.)

At this point, Zetex is only offering samples and eval kits of the ZXCW8100S18 to 'qualifying customers'. If they follow the history of their ZXCD1000 analog input Class D chip, perhaps we'll be able to buy it from Digi-Key in a year.

We're just at the beginning of a new era. These are first generation digital amplifier IC devices. I expect Analog Devices and Philips to be in the game in a year. Personally I find it very exciting.

Regards,
Brian.

[Rarkov]

I have a friend who works for Zetex in Manchester, UK...I think I might get a few smaples from him...sounds more promising than their ZXCD1000!

Gaz