[time-nuts] Simple open source microcontroller solution to tune DDS needed

Alberto di Bene dibene at usa.net
Thu Dec 14 11:23:35 EST 2017


The problem with the 9854 DDS is that the analog reconstruction filters that you have to place after its outputs, are 
not guaranteed to have
*exactly* the same phase behavior  between the two sections (I and Q). This may or may not be important, depending on 
the application.

For sure, for an SDR, even small discrepancies between the two phases, let's say 0.1 degree, can lead to a non complete 
cancellation of
the unwanted sideband, when you implement a zero IF SDR radio.

73  Alberto  I2PHD
==============================================================================

On 12/14/2017 5:15 PM, Jerry Hancock wrote:
> I have those 9954/9854 boards and the problem is the drivers for newer operating systems past XP aren’t available.  The AD9854/9954 boards were clones of the AD eval kits and you downloaded the eval kit software from AD.  I got the board(s) working with Windows 7 but it took a lot of screwing around.   I never got it working with Windows 10.  The other issue with those boards are the filters were set really high.  All DDS units need a filter on the output though I’m sure the term “all” can be debated.
>
> If you have a need to generate a quadrature signal and have access to Windows 7, or better yet, XPSP3, then yes, it is a decent option and seemed to work well.  The filters can be handled off-board, especially if you just need 10Mhz as BG7TBL has decent 10Mhz filters for $10 per.  I bought a couple and hung them on my RFTG-U REF0 and REF1 outputs and they make a really nice 10Mhz sine wave.



More information about the time-nuts mailing list