L
Admin
[Technical article / FAQ, rfforum Editorial Team]
Mach–Zehnder modulator bias is often set for "max RF" or "quad" by a one-time tweak. On links that run for hours, bias drift moves you off quadrature: second-harmonic rises, even-order terms grow, and the RF gain you froze at turn-on is no longer the operating point.
1. What bias actually owns
- Quadrature (quad) bias nulls even-order distortion and sets a repeatable RF slope; peak-transmission bias maximizes optical power but is a different linearity point
- A second-harmonic null at the detector is a practical lock signal for quad; chasing max fundamental alone can walk you off that null
- Temperature, photorefractive effects, and supply ripple move the null; a bias tee and slow feedback loop are part of the RF-over-fiber product, not optional lab gear
2. Freeze list for an MZM link claim
- Bias mode (quad vs peak), second-harmonic level, and whether a bias-lock loop was closed — on the same sheet as RF gain / SFDR
- Optical power and photocurrent at the detector (ties to RIN / shot terms elsewhere)
- Warm-up time before the number was frozen, and whether drift was re-checked after thermal soak
- Drive level / OMI: compression and harmonic growth depend on how hard the MZM is driven
3. Habit
Never publish MZM RF gain or SFDR without bias mode and a harmonic / lock note. A "max RF" tweak at cold start is not the same operating point as a locked quad after soak.
Discussion: on your last MZM drift hunt, did you lock on second-harmonic null, optical power, or RF fundamental — and do you require a post-soak re-check before freezing the link number?
— rfforum 编辑部 / Editorial Team
Mach–Zehnder modulator bias is often set for "max RF" or "quad" by a one-time tweak. On links that run for hours, bias drift moves you off quadrature: second-harmonic rises, even-order terms grow, and the RF gain you froze at turn-on is no longer the operating point.
1. What bias actually owns
- Quadrature (quad) bias nulls even-order distortion and sets a repeatable RF slope; peak-transmission bias maximizes optical power but is a different linearity point
- A second-harmonic null at the detector is a practical lock signal for quad; chasing max fundamental alone can walk you off that null
- Temperature, photorefractive effects, and supply ripple move the null; a bias tee and slow feedback loop are part of the RF-over-fiber product, not optional lab gear
2. Freeze list for an MZM link claim
- Bias mode (quad vs peak), second-harmonic level, and whether a bias-lock loop was closed — on the same sheet as RF gain / SFDR
- Optical power and photocurrent at the detector (ties to RIN / shot terms elsewhere)
- Warm-up time before the number was frozen, and whether drift was re-checked after thermal soak
- Drive level / OMI: compression and harmonic growth depend on how hard the MZM is driven
3. Habit
Never publish MZM RF gain or SFDR without bias mode and a harmonic / lock note. A "max RF" tweak at cold start is not the same operating point as a locked quad after soak.
Discussion: on your last MZM drift hunt, did you lock on second-harmonic null, optical power, or RF fundamental — and do you require a post-soak re-check before freezing the link number?
— rfforum 编辑部 / Editorial Team
