I had a demo of an MXG with integrated arbs with "16 bit" DACs. When I turned off waveform scaling, ALC, etc, so that my amplitude is only controlled by the I and Q values, the unit behaved as if the DACs are 14 bits. Discussion with Agilent via phone and email did not result in any satisfactory explanation.

Are the DACs 14 bits? i.e. if I were to see a schematic of the DAC circuit, would it be 14 bits? Not, "oh, after interpolation it performs like a 16 bit DAC" What if the waveforms are DC? Would I see 2^16 different outputs if I kept Q=0 and changed I from -32767 to +32767? Because that's what I did, and it looks like the DACs are 14 bits. If I made I=sin(x) and Q=cos(x) so that my amplitude is constant, can I vary x and get 16 bits of phase resolution?

Are the DACs 14 bits? i.e. if I were to see a schematic of the DAC circuit, would it be 14 bits? Not, "oh, after interpolation it performs like a 16 bit DAC" What if the waveforms are DC? Would I see 2^16 different outputs if I kept Q=0 and changed I from -32767 to +32767? Because that's what I did, and it looks like the DACs are 14 bits. If I made I=sin(x) and Q=cos(x) so that my amplitude is constant, can I vary x and get 16 bits of phase resolution?

The DAC’s in the MXG are 16 bit DAC’s but you only get 16 bits range/resolution if your MXG has Option UNV; Enhanced dynamic range.

If you do not have Option UNV the higher bits are loaded with noise to limit the dynamic range.

Regards -