--- In [email protected], "D. J. Wang" <[EMAIL PROTECTED]> wrote: > > Hi Janne, > Building a DDS with AD9951/4 is easier than you think. I had copied >the design by DL5MGD and it was pretty easy and it works fine even >though I used only a 400 MHz clock (80 x5). His web page is: >http://www.dl5mgd.de/dds/AD9951.htm , since it was so easy I built two >of them, one for use as a standalone/LPT control SG, the other is part >of a Vector Network Analyzer designed by IW3HEV, I simply replaced the >AD9851 with an 9951. > I would give the I0CG 500 MHz clock a try when I have more free time. >I used the DDS SG currently for my SDR setup, it works pretty good >although I did notice some birdies. > > 73 de D.J. Wang / NM3R
Thanks for the reply Wang! Yeah, it shouldn't be too hard! I've actually visited dl5mgd's sites before, but now I understood it much better :) I just started to draw circuit board with Eagle. I have the old version 4.01 and just noticed that it doesn't have TQFP 48 on the library and couldn't find it from the internet, so I had to make it manually, pheeewww. Just a question about your reply. Did you mean with "Building a DDS with AD9951/4 is easier..." same as "Building a DDS with AD9951 / AD9954 is easier..." or "Building a DDS with AD9951 frq/4 is easier...". Is it acceptable to use the multiplication of the clock frequency? I've heard that it may result some phase noise problems? It would be quite easy to use just the same multiply -method like you, instead of building direct 400..500MHz oscillator. Reading the AD9954 datasheet about 20 times gives somekind of idea how to access the AD9954 tuning words & stuff, but I think it might be much easier to study someone's earlier work. Maybe I should go with the dl5mgd's design? 73 de Janne, OH1GTF
