I have found one of the bad things about these AD9850/51 units is that they are
often shipped with a bad crystal.
Sometimes. also. they have a 3.3v unit installed and the device is running at 5v.
It will run ok for a while, then rather suddenly fail.
Note the crystal. Some, but not all are marked for 5v or 3.3v operation. If you have
a frequency counter, you can connect it temporarily through a 100 pf. capacitor or so
to the output pin of the oscillator, and measure for any output. It should be about 125 MHz
for the AD9850 and about 30 MHz for the AD9851. The exact number does not matter; what you are
trying to do is verify that the oscillator is working. Quite often, they do not.
If it is bad, it is easily replaced by anything from 30 to 34 MHz for the AD9851 and 125 MHz for the
AD9850. Try to determine both the working voltage of the device AND the designed voltage for the
crystal oscillator. Match them. Then calibrate the combined working unit.
I have recovered several bad units in this manner. YRMV.
IF a device is not working, it is likely to be a bad solder joint or a bad crystal oscillator. The
AD98xx devices themselves are rather rugged. Either can take a 5V supply (but NOT a12v one!).
Now, how well they work is another story. I have heard several explanations for this, most commonly a
mismatched filter string.
john
AD5YE
