Re: [Discuss-gnuradio] 2.5 GS/sec 14-bit A/D converter

[Larry Doolittle]

Tim -

On Mon, May 02, 2005 at 01:46:48PM +0930, Tim Ansell wrote:
> So where do we request free samples from :P

Good question.  I'd be happy with a data sheet, for the moment.
I'm in the low-latency business, so this chip _may_ be useless to me.

> The question is how would you get this into a computer? If we round it
> up to 16 bits, it makes 4.6 Gigabytes per second. Would a FPGA even be
> able to keep up? Could people a lot smarter then me, discuss how this
> chip could even possibly be used?

Look at the table again (as corrected by Krzysztof?  I can't verify
because lists.gnu.org is currently refusing port 80 connections):

>Parameter                                    GaAs            SiGe
>
>Maximum clock frequency                       > 5 GHz        > 15 GHz
>SINAD (signal to noise+distortion)           > 70 dB         > 110 dB
>Eff. resolution bits at 2.5 GHz carrier:
>10 MHz BW                                      14 bits         18 bits
>100 MHz BW                                     11 bits         14 bits

So what you need coming out is N bits at something over
twice the bandwidth.  For the SiGe chip, that's 14 bits at > 200 MS/s.
Even with generous provision for filter guard bands and compatibility
with maximum accuracy, you have 20 bits at 300 MS/s, well within reach
of a modern FPGA.

The patent looks -- well, too obvious to be patentable [*].  They seem
to be trying to patent a mixer followed by a sigma-delta converter, in
particular the case where the same frequency is used for the mixer LO
and the sigma-delta modulator clock.

       - Larry

[*] but then, what do I know.  I'm an engineer, so I must not count
as someone "skilled in the art".

signature.asc
Description: Digital signature