Still, I don't like a standalone DDS for the spurs. If you look at the data sheet for the DDS chip N3ZI uses, certain divide ratios provide fairly horrible spurs. Better would be to use it to drive a bank of lo-noise PLLs so the spurs are stripped off.
Wonder how a typical standalone DDS unit like we have been discussing here would compare to a current production ricebox used as a vfo to drive a plate modulated class-C amplifier?
Something to be concerned about is that a nearby spur may be way down, say 50 dB at the output of the DDS VFO, but due to the limiting action of the class C amplifier that follows, the signal-to-spur ratio at the output will be much lower, maybe only 20-30 dB down. The DDS VFO's in riceboxes are followed by
linear amplifiers, which theoretically should preserve the original signal-to-spur ratio at the ricebox output.