Here are the basics. Light is sent into a fiber and reflects back from the FBG. The reflected light travels back to the instrument’s photo detectors and is compared to wavelength reference artifacts so that the instrument can determine the position of the center wavelength of the FBG. Wavelength information is converted to engineering units, e.g., 1.2 picometers of wavelength shift could correspond to 1 microstrain. The actual translation is given by the gage factor supplied with the FBG sensor.

When more than one FBG is present on a fiber (this is often the case), the instrument will use one of two schemes to discriminate between one FBG and the next. Time division multiplexing (TDM) systems use the known speed of light in the fiber to discern which signal is reflected from which FBG along the fiber path. Theoretically, 100 or more FBGs can be on the same fiber at the same nominal center wavelength.

The most utilized scheme is wavelength division multiplexing, or WDM. WDM FBGs are at distinctly different nominal center wavelengths from their neighbors, and the interrogator uses these unique FBG wavelengths to keep track of which sensor is which. Sensor capacity on each fiber is determined by the range that each sensor will measure and the total spectral range of the instrument. WDM ranges are now very large and can also accommodate more than 100 sensors per fiber.

We (Micron Optics) have chosen the WDM approach for a few reasons.

1) Scanning speed is not a function of the number of sensors, and the scan speed can be up to 2kHz.

2) Swept laser sources are powerful enough to split into four fibers for simultaneous measurement of >300 sensors.

3) FBGs can be highly reflective. This coupled with the high dynamic range of the instrument make the system much more flexible for measurements over tens of kilometers of fiber.

4) WDM is compatible with Micron Optics fast, narrow line, wide range lasers that are stable over time and temperature, and are mechanically robust in environments where they will encounter shock and vibration.

There are other approaches, but they have significant drawbacks:

a) Broadband source, Dispersive element, Diode Array
Limitations: This method cannot achieve the required wavelength measurement repeatability and resolution with commercially available diode arrays. Low broadband source power limits the ultimate needed combination of channel count/sensor capacity and dynamic range/distance to sensors.

b) Broadband source, Optical Spectrum Analyzer/Multi-line wavelength meter
Limitations: Laboratory OSAs are large, slow, expensive, and do not have a wide operating temperature range. Multi-line wavelength meters acquire data at slow speeds only, and are not mechanically robust. Low broadband source power limits the ultimate needed combination of channel count/sensor capacity and dynamic range/distance to sensors.

c. OTDR/TDM systems
Limitations: Low loss budget precludes a solution with the required number of sensors and/or channels, and data acquisition rates scale down with increasing sensor counts. Minimum physical grating spacing limits some applications.

d. External Cavity Tunable Laser, Power Meter, Wavelength Meter
Limitations: External cavity tunable lasers are slow, expensive, and do not have a wide operating temperature range or the required mechanical robustness. The addition of power meters and wavelength meters add to the bulk, complexity, and cost, as well as reduce reliability and speed. Polarization properties of the narrow line lasers may not be an ideal match for all sensing applications.

Tom Graver
Director, Optical Sensing
twgraver@micronoptics.com