J
Member
This gets asked often enough that I wrote down my notes.
1. The Y-factor method with a noise source is the standard approach; the uncertainty is dominated by the ENR uncertainty.
2. Cable and adapter losses before the DUT raise the measured NF; you must de-embed them, not ignore them.
3. The second-stage contribution matters: a lossy or noisy analyser front end will limit your floor.
4. Always measure with the DUT powered and settled — thermal drift during the measurement is a silent error.
Happy to hear a better approach.
1. The Y-factor method with a noise source is the standard approach; the uncertainty is dominated by the ENR uncertainty.
2. Cable and adapter losses before the DUT raise the measured NF; you must de-embed them, not ignore them.
3. The second-stage contribution matters: a lossy or noisy analyser front end will limit your floor.
4. Always measure with the DUT powered and settled — thermal drift during the measurement is a silent error.
Happy to hear a better approach.