[hpsdr] Mercury: LTC2209

[Eric Blossom]

On Sat, Apr 21, 2007 at 11:07:37AM -0700, Chris Albertson wrote:
> ***** High Performance Software Defined Radio Discussion List *****
> 
> Purely theoretical questions...
> 
> Compare the Mercury with one channel to the USRP with two 65Msps
> channels.  It seems to me that you get the same result in the end if
> you use two 65Msps channels in quadrature or if you use one "real"
> channel that runs a bit more than twice as fast.  After all you can
> divide the fast channel into even and odd samples can get out of phase
> two channels.
> 
> I know it's not the same, just wondering how the two approaches compare

(1) The LTC2208 has better spurious free dyanamic range than the AD9862 on
the USRP.

(2) To be able to take advantage of quadrature sampling, you need to
(somehow) get a quadrature signal to the two A/D's.  The Mercury
design (as I understand it), tries to minimize the amount of
non-linear analog h/w between the antenna and the converter.  This
argues against any kind of mixer etc that could be used to generate
quadrature input for two A/D's.

Eric, K7GNU

 1177183107.0