Choosing a Light Source for Fiber Bragg Grating Testing and Production

Use a broadband ASE or SLD source to characterize an existing fiber Bragg grating (FBG), a stable fixed or multi-channel source to screen known wavelengths in production, and a dedicated UV or ultrafast writing system to write the grating. An ASE source can illuminate and measure an FBG, but it is not automatically an FBG inscription laser.

Start by separating three FBG workflows

“FBG production” can mean either manufacturing the grating itself or testing finished gratings on a production line. The source choice changes completely between those workflows.

FBG task	Typical source architecture	Parameters that decide fit	Main mistake to avoid
Characterize an existing FBG	Broadband ASE, SLD, or another low-coherence source	Usable spectrum, bandwidth definition, power spectral density, flatness, polarization, and stability	Comparing total power without checking power per nanometer
Interrogate or screen finished gratings	Fixed CW, tunable, or multi-channel source plus a detector or interrogator	Bragg wavelength, channel spacing, wavelength accuracy, repeatability, power at the device, and test throughput	Assuming a narrow fixed source can reveal the full reflection spectrum
Inscribe or write an FBG	UV or ultrafast writing laser, phase mask or scanning optics, and alignment hardware	Writing wavelength, pulse or exposure conditions, beam profile, fluence, fiber photosensitivity, and process timing	Treating a milliwatt-class test source as a writing tool

An FBG reflects a narrow wavelength band determined by its grating period and effective index. A characterization setup therefore needs a source that covers the expected Bragg wavelength, an optical path into the fiber, and an optical spectrum analyzer or interrogator that resolves the response.

What an FBG test source must do

For a finished grating, the first question is not “How many milliwatts do I need?” It is “What spectrum must reach the grating, and how will I measure the response?” A broadband source may illuminate a family of gratings at once. A fixed source may be adequate for a pass/fail check at one known wavelength. A production line with many known grating centers may benefit from multiple individually controlled channels.

Four specifications usually separate a useful FBG source from a merely compatible one.

1. Match the Bragg wavelength and the measurement band

Choose the source from the actual grating center wavelength, not from the fiber name alone. A 1550 nm telecom FBG, a 1064 nm specialty-fiber grating, and a 980 nm grating need different source bands and different detector assumptions.

Then define the usable band. For a single grating, the band may only need to cover the expected shift and reflection width. For a production batch, it may need to cover multiple nominal centers, process drift, and out-of-spec parts. Write the center wavelength, acceptable range, and temperature or strain shift into the request for quotation.

2. State how the bandwidth is measured

“Broadband” is not a complete specification. A range measured at 2 dB, 10 dB, or 20 dB describes a different part of the spectrum. The nanometer span cannot be compared fairly unless the definition is the same.

For example, the current Omni Wavelength catalog lists the C-band ASE source at 1528–1569 nm using a 2 dB definition. Its 1030 nm source lists 1018–1044 nm at 10 dB, while the 980 nm source lists 973–982 nm at 20 dB. Those figures are useful for checking coverage, but they are not a ranking of which source is “broader.”

For an FBG measurement, ask for the spectrum at the requested output power, the level used to define the range, the spectral resolution of the measurement, the flatness or ripple, and the tolerance across temperature.

3. Compare power spectral density, not only total power

An ASE source may have a large total output while delivering insufficient power in the narrow spectral slice that reaches the detector. Power spectral density, usually expressed in dBm/nm, is often the more useful number for broadband FBG work.

The current C-band ASE page lists power spectral density from −6 to +11 dBm/nm, while its output options range from 10 to 500 mW for SM configurations and 10 to 200 mW for PM configurations. The 980 nm page lists at least 10 mW total output and at least −20 dBm/nm power spectral density. These figures show why total power and per-nanometer power should be requested separately.

The required source power still depends on the optical path. Include connector, circulator, coupler, and fiber loss, plus grating reflection, detector sensitivity, and detector saturation. More source power is not automatically better if it creates measurement nonlinearity.

4. Define polarization and fiber delivery

Many FBG measurements are intended to reduce polarization sensitivity, while others deliberately use polarization-maintaining components. Do not assume that PM is always better or that standard SM output is always neutral.

The current C-band ASE page describes SM output as completely unpolarized with PER ≤0.2 dB and PM output as linear with PER ≥23 dB. Its public configurations use G652D/SMF-28 for SM output and PM1550 fiber for PM output, with FC/APC connectors. A polarization-sensitive grating or interrogator may need PM fiber, axis orientation, and a defined PER. A general spectral screening bench may instead prefer low-polarization output.

Ask for the output fiber, connector type, pigtail length, connector key orientation, polarization definition, and the reference plane for all power and stability figures.

Choose the source architecture by workflow

Broadband characterization: ASE or SLD

Use a broadband ASE or SLD source when the goal is to see the FBG reflection spectrum, measure center-wavelength shift, compare reflection strength, or evaluate several gratings across one band. The source should cover the complete wavelength window of interest with enough power spectral density for the analyzer or interrogator.

ASE is often the more flexible choice when the bench needs higher output options, adjustable power, SM and PM variants, or a band-specific spectrum. SLD can be a practical choice when the system is already built around a common center wavelength and needs a simpler, lower-power broadband source. The existing ASE vs SLD broadband-source comparison provides the general source-architecture background; an FBG RFQ still needs the grating band and detector requirements.

For a 1550 nm FBG, the C-band ASE broadband source is a logical catalog starting point because its public spectrum is 1528–1569 nm at 2 dB. The trade-off is configuration-dependent flatness: the page lists ≤2 dB for 10–200 mW, ≤3 dB for 300–500 mW, and a ≤1 dB F1 option. If the measurement compares spectral amplitude across the band, the 100 mW flatness option may be more useful than simply choosing the largest output.

Production screening: fixed or multi-channel output

Use a fixed CW source when every part is expected to have one known Bragg wavelength and the production test only needs a repeatable signal at that point. This can reduce system complexity, but it cannot replace a spectral sweep when the acceptance criteria include center-wavelength shift, reflection width, side lobes, or unexpected secondary peaks.

Use a multi-channel source when the production line tests several known wavelengths or needs parallel outputs. Omni Wavelength’s 24-channel multi-channel fiber light source lists 24 laser channels and 24 output fiber channels. Its public wavelength coverage includes 405–850 nm, 974/976/1480 nm, and C+L bands; the page lists spectral width ≤0.1 nm, output options of ≥10 or ≥20 mW, 10–100% power adjustment, and separate 15-minute and 8-hour stability figures.

That architecture is useful when the wavelength set is known and throughput matters. It is not a substitute for a broadband source or analyzer when the test must reveal the complete FBG reflection spectrum. Confirm the populated channels, center wavelengths, fiber, and connector before ordering.

For visible and near-infrared component work, the 405–940 nm single-mode test light source lists 14 selectable wavelengths, CW operation, 10–100% power adjustment, and SM, MM, or PM options. Its public power table varies by wavelength, from 5 to 90 mW. It is a useful example of a wavelength-selectable test platform, but its listed range does not cover a 1550 nm FBG; do not treat “test light source” as a universal FBG solution.

FBG inscription: a dedicated writing system

Writing an FBG is a material-processing step. The usual system includes a UV or ultrafast writing laser, beam delivery and shaping, a phase mask or another interference method, precision fiber positioning, and process monitoring. The writing source is selected by the fiber’s photosensitivity and the inscription method, not by the output power required for a later spectral measurement.

Published FBG fabrication setups use process-specific UV or ultrafast sources and phase-mask arrangements. The wavelength, pulse energy, beam size, exposure time, and scan method change with the fiber, grating design, and production target.

The public Omni Wavelength pages reviewed for this article describe ASE, CW test, multi-channel, and other fiber-coupled platforms, but do not publish a complete FBG inscription system. The 405–940 nm test source is explicitly described as a CW testing platform. It should not be presented as an FBG writing laser unless engineering has confirmed a separate process configuration.

If the requirement is inscription, send a separate RFQ containing:

• writing wavelength and whether the fiber is photosensitive or hydrogen-loaded;
• pulse width or CW exposure mode, pulse energy or fluence, repetition rate, and average power;
• beam diameter, spatial profile, polarization, and stability at the fiber;
• phase-mask period, scan length, alignment tolerance, and exposure method;
• target Bragg wavelength, grating length, reflectivity, bandwidth, apodization, and chirp;
• fiber coating, stripping, handling, annealing, and post-process inspection requirements.

This prevents a measurement-source quotation from being mistaken for a complete grating-manufacturing proposal.

Current catalog examples and their practical fit

The following values are public configuration examples. They are not universal guarantees for every option in a product family, and the quoted configuration should be rechecked before publication or purchase.

Public source example	Current listed parameters	Where it can fit	Important trade-off or question
980 nm ASE broadband source	973–982 nm at 20 dB; ≥10 mW; power spectral density ≥−20 dBm/nm; DOP ≤2.5%; Hi-1060 fiber and FC/APC	980 nm FBG characterization or low-polarization fiber testing	The band is relatively narrow; confirm the grating center and whether total power or PSD is the limiting requirement
1030 nm ASE broadband source	1018–1044 nm at 10 dB; 10–1000 mW; fixed or 10–100% tunable power; SM Hi-1060 or PM980; PM options listed from 10–100 mW	1 µm-class FBG characterization and sensing-related test benches	The 10 dB definition and the PM power boundary must be carried into the RFQ; do not infer PM availability at 500 or 1000 mW
1064 nm ASE broadband source	1040–1080 nm or 1030–1080 nm at 10 dB, depending on configuration; 10–1000 mW; SM or PM fiber; fixed or 10–100% tunable versions	1064 nm FBG and component characterization when the 1 µm optical path is already defined	Ask which spectrum range applies to the exact model and whether the detector and grating use that full range
C-band ASE broadband source	1528–1569 nm at 2 dB; SM 10–500 mW; PM 10–200 mW; PSD −6 to +11 dBm/nm; flatness ≤2 dB at 10–200 mW and ≤3 dB at 300–500 mW	1550 nm FBG reflection and spectral-response measurement	The bandwidth definition and flatness change the comparison; choose the exact power and F1 option from the acceptance requirement
24-channel multi-channel fiber light source	24 laser channels and 24 output fiber channels; 405–850 nm, 974/976/1480 nm, and C+L coverage; spectral width ≤0.1 nm; ≥10/≥20 mW options	Parallel production screening at known wavelength points	Confirm channel population, wavelength tolerance, fiber type, and whether a fixed channel can expose the measurement feature you need

For a broader explanation of how to read wavelength, bandwidth, power, fiber, and stability values together, see Laser Specification Sheet Explained: 15 Parameters Buyers Should Understand. The existing testing-light-source application note is also relevant when the FBG bench is part of a larger component-testing workflow.

What to include in an FBG light-source RFQ

Send the supplier one complete requirements block instead of asking for “a light source for FBG.” Include:

• FBG center wavelength or wavelength list;
• reflection bandwidth and allowable wavelength shift;
• required source bandwidth definition: 2 dB, 10 dB, 20 dB, FWHM, or another method;
• minimum power spectral density and total power at the grating or detector;
• detector sensitivity and saturation limit;
• SM, MM, low-DOP, or PM output, including minimum PER if needed;
• fiber type, connector, pigtail length, key orientation, and return-loss limit;
• short-term and long-term stability with warm-up time and temperature range;
• benchtop, module, or multi-channel package and required control interface;
• whether the source is for characterization, production screening, or FBG inscription;
• for inscription only: pulse, beam, phase-mask, alignment, and process-monitoring requirements.

If the specification is incomplete, the supplier may return a source that reaches the nominal wavelength but cannot produce a reliable FBG measurement or a repeatable production result.

Conclusion

Choose the source architecture from the FBG process. Use ASE or SLD for broadband characterization, fixed or multi-channel output for known-wavelength production screening, and a dedicated UV or ultrafast process system for inscription. Then compare the complete configuration: wavelength definition, PSD, flatness, power at the device, polarization, fiber, connector, stability, package, and acceptance data.

If you need a matched configuration, contact Omni Wavelength engineering sales with the FBG wavelength, measurement method, power budget, fiber and connector requirements, package preference, and whether the request is for testing or inscription.

Frequently asked questions

Can I use an ASE source to write an FBG?

Usually no. ASE is suited to broadband illumination and characterization. FBG inscription normally requires a dedicated UV or ultrafast writing system with beam delivery, a phase mask or another interference method, fiber positioning, and process control.

Should I choose ASE or SLD for FBG testing?

Choose between them after defining the grating band, required power spectral density, and package. SLD can be practical for a known band and moderate power. ASE is often more flexible when you need broader band coverage, adjustable power, low-polarization behavior, or SM/PM and module/benchtop choices.

Is a 1550 nm source always the right choice for an FBG?

No. 1550 nm is common in telecom and sensing, but the correct wavelength is set by the grating, fiber, detector, and optical fixture. 980 nm, 1030 nm, 1064 nm, and other bands can be appropriate when the FBG and measurement system are designed for them.

Do I need PM fiber for FBG measurements?

Only when the grating or the downstream measurement path is polarization-sensitive or built around PM components. A general screening bench may prefer low-polarization output. Confirm DOP, PER, fiber type, axis orientation, and connector details together.

How much power does an FBG test source need?

There is no universal number. Start with detector sensitivity and saturation, then work backward through grating reflection and optical losses to the source connector. For broadband sources, specify power spectral density as well as total output power.