"john jardine" writes:


"OH1GTF" wrote in message
roups.com...
Hi fellows!

I'm building a software defined radio that is mostly based on the
"Software Defined Radios For The Masses" -articles and are at the point
of designing the DDS board. I have two units of AD9954 DDS chips and
my idea was to use them parallel, other one with 90 degree phase shift
to generate the quadrature signal. This is my first time with DDS, so
understanding the control seem to be a little bit difficult for a first
timer.

Two questions to start with:

1. I suppose for proper control I need four pins connected from the DDS
to the computer (through buffers) I/O update, SDIO, SCLK and Reset, am
I correct?

2. Do I have to drive two SDIO serial signals separately for each DDS
chip from computer so, that I can define the phase shift for the other
DDS (while other one has no phase shift, obivously) ? This came to my
mind since the chips don't have any addresses to indentify them.

Thanks a lot!!!

73 de Janne, OH1GTF

Don't know that particular DDS but did it with two AD9832 used in a vector
impedance arrangement. The master clock and serial clock/data lines were
wired to both chips. Common serial data was sent to both chips but made
'valid' only for the wanted DDS by use of individual "Fsync"(?) lines.
Worked no problem. Chips never got out of step. Phase could be varied at
will.
john

Yep, that will always work. If you aren't tied to the AD9954, I would suggest
the AD9959 or AD9958. You get 4 or 2 channels in one chip with lots of
control. Cost more, but pretty nice part.

Steve.
--
Steven D. Swift, , http://www.novatech-instr.com
NOVATECH INSTRUMENTS, INC. P.O. Box 55997
206.301.8986, fax 206.363.4367 Seattle, Washington 98155 USA