L
Admin
[Technical article, rfforum Editorial Team]
Analog RF-over-fiber budgets are where optical dB and RF dB get confused. Photodetection is fundamentally square-law in optical field (linear in optical power for the photocurrent), so a 1 dB drop in optical power is about 2 dB in RF power — until saturation, clipping, or a different detection regime breaks the rule.
1. The basic mapping
For a link that is still linear, RF power out scales with (optical power)² into the detector. Optical attenuators, connector dirt, and splice loss therefore bite twice as hard on the RF side. Teams that pad "1 dB optical margin" and expect 1 dB RF margin under-budget the link.
2. Where the rule bends
- Near PD / MZM saturation or clipping, incremental optical loss may not map cleanly to 2× RF
- AGC or EDFA gain control can mask optical loss until the noise floor jumps
- Dispersion-induced RF fading is frequency-dependent and is not a flat optical-loss term
3. Budget hygiene
Write two columns: optical power at each node, and predicted RF gain / NF. Convert optical penalties with the square-law factor explicitly. Measure RF gain vs a calibrated optical attenuator sweep to confirm you are still in the linear region.
4. Habit
Any change to patch cords or connectors on a live RoF run should be followed by an RF power check, not only an optical power check.
Discussion: do you keep a lab plot of RF dB vs optical attenuator setting as a standard bring-up artefact, and at what optical power do you usually see the square-law slope break on your PD?
— rfforum 编辑部 / Editorial Team
Analog RF-over-fiber budgets are where optical dB and RF dB get confused. Photodetection is fundamentally square-law in optical field (linear in optical power for the photocurrent), so a 1 dB drop in optical power is about 2 dB in RF power — until saturation, clipping, or a different detection regime breaks the rule.
1. The basic mapping
For a link that is still linear, RF power out scales with (optical power)² into the detector. Optical attenuators, connector dirt, and splice loss therefore bite twice as hard on the RF side. Teams that pad "1 dB optical margin" and expect 1 dB RF margin under-budget the link.
2. Where the rule bends
- Near PD / MZM saturation or clipping, incremental optical loss may not map cleanly to 2× RF
- AGC or EDFA gain control can mask optical loss until the noise floor jumps
- Dispersion-induced RF fading is frequency-dependent and is not a flat optical-loss term
3. Budget hygiene
Write two columns: optical power at each node, and predicted RF gain / NF. Convert optical penalties with the square-law factor explicitly. Measure RF gain vs a calibrated optical attenuator sweep to confirm you are still in the linear region.
4. Habit
Any change to patch cords or connectors on a live RoF run should be followed by an RF power check, not only an optical power check.
Discussion: do you keep a lab plot of RF dB vs optical attenuator setting as a standard bring-up artefact, and at what optical power do you usually see the square-law slope break on your PD?
— rfforum 编辑部 / Editorial Team
