L
Admin
[Technical article / FAQ, rfforum Editorial Team]
An EDFA that "only" adds a few dB of optical noise figure can erase tens of dB of RF SFDR on an analog RoF hop. Optical ASE folds into the photodetector as broadband RF noise; the same amplifier that rescues optical budget can be the term that kills spur-free dynamic range.
1. What EDFA actually does to RF SFDR
- Optical NF and ASE density raise the RF noise floor after PD square-law conversion; SFDR shrinks from the noise side even when laser RIN and modulator linearity are unchanged
- Gain saturation and pump settings move ASE; a cold desk EDFA setting is not the soaked, traffic-loaded operating point
- Cascaded EDFAs compound: each stage adds ASE that later stages amplify — Friis-like thinking belongs on the optical+RF cascade, not on the RF LNA alone
2. Freeze list for an analog RoF SFDR claim with EDFA in path
- Optical NF / ASE note (or EDFA PN + gain/pump setting) on the same sheet as RF SFDR
- Photocurrent / optical power at the PD under the SFDR test tone plan
- Whether SFDR was two-tone RF into the link with the EDFA at product gain, not bypassed
- Tone spacing and measurement BW — ASE is broadband; RBW choices change the apparent floor
3. Habit
Never publish analog RoF SFDR with an EDFA in the drawing but absent from the noise budget. Archive optical NF / ASE next to every RF SFDR traveler. A bypassed-EDFA desk number is not a product number for an amplified hop.
4. Quick sanity
- Measure RF noise floor with EDFA at product gain vs optical attenuator substituting the same photocurrent without the EDFA ASE — the delta is roughly the ASE tax you must budget
- If SFDR collapses only when EDFA gain is raised to hit optical budget, the link is ASE-limited; buying a quieter RF LNA after the PD will not buy the SFDR back
Discussion: on your last SFDR miss after adding an EDFA, did the fix land in EDFA NF/gain, optical power at the PD, or RF tone plan — and do you require an ASE/NF line on every amplified RoF SFDR sheet?
— rfforum 编辑部 / Editorial Team
An EDFA that "only" adds a few dB of optical noise figure can erase tens of dB of RF SFDR on an analog RoF hop. Optical ASE folds into the photodetector as broadband RF noise; the same amplifier that rescues optical budget can be the term that kills spur-free dynamic range.
1. What EDFA actually does to RF SFDR
- Optical NF and ASE density raise the RF noise floor after PD square-law conversion; SFDR shrinks from the noise side even when laser RIN and modulator linearity are unchanged
- Gain saturation and pump settings move ASE; a cold desk EDFA setting is not the soaked, traffic-loaded operating point
- Cascaded EDFAs compound: each stage adds ASE that later stages amplify — Friis-like thinking belongs on the optical+RF cascade, not on the RF LNA alone
2. Freeze list for an analog RoF SFDR claim with EDFA in path
- Optical NF / ASE note (or EDFA PN + gain/pump setting) on the same sheet as RF SFDR
- Photocurrent / optical power at the PD under the SFDR test tone plan
- Whether SFDR was two-tone RF into the link with the EDFA at product gain, not bypassed
- Tone spacing and measurement BW — ASE is broadband; RBW choices change the apparent floor
3. Habit
Never publish analog RoF SFDR with an EDFA in the drawing but absent from the noise budget. Archive optical NF / ASE next to every RF SFDR traveler. A bypassed-EDFA desk number is not a product number for an amplified hop.
4. Quick sanity
- Measure RF noise floor with EDFA at product gain vs optical attenuator substituting the same photocurrent without the EDFA ASE — the delta is roughly the ASE tax you must budget
- If SFDR collapses only when EDFA gain is raised to hit optical budget, the link is ASE-limited; buying a quieter RF LNA after the PD will not buy the SFDR back
Discussion: on your last SFDR miss after adding an EDFA, did the fix land in EDFA NF/gain, optical power at the PD, or RF tone plan — and do you require an ASE/NF line on every amplified RoF SFDR sheet?
— rfforum 编辑部 / Editorial Team
