What is the sig noise from a combination of DAC ( e.g AD9142) and IQ MOD (eg ADL5375)? Is it dominated by the DAC noise or the IQ MOD? if it is the latter, then the IQ Mod data sheet gives me my answer.

Assume DAC =AD9142 with 50 Ohm to ground on IP, IN and 100 Ohms load across the IQ MOD input - same set up on QP/QN. Result for ADL5375-05 @ LO =1.9G is RF output = 0.7dBm with baseband IQ input =1Vpk-pk or 0.354Vrms . Noise floor from IQ MOD data

sheet = -160dBm/Hz @ 1.9G, so the Sig/Noise is 160.7 dBc.

What is the noise floor at the RF output due to the DAC noise? For Fout =10MHz, the AD9142 data sheet gives -162dBm/Hz for the Noise Spectral Density.; it does not describe the termination resistor.

Assume 50 Ohm resistor; then -162dBm/Hz = 1.8nV/root(Hz). Compare 1.8nV with 0.354Vrms, and the noise from the DAC at the1.9G RF output is -165 dBm/Hz. This means the noise of the IQ MOD is dominating.

I admit being fast & loose assuming 50 Ohm term for the DAC noise giving 1.8nV, and comparing this directly with 0.354 Vrms. However, if the net result does not degrade Sig/Noise from 160.7dB by more than 2 dB, I am satisfied. Is this a reasonable calculation?

IQ Mods are typically a little noiser than DACS. You can use ADIsimRF to get a better sense for this. It has device models for both DACs and IQ Modulators (www.analog.com/adisimrf)