Modern metrology labs and large service providers require periodic verification of their fiber-optic test and measurement equipment. The accuracy of this equipment depends largely on the calibration quality of the power meters.
Many factors must be considered when performing absolute power calibration and linearity tests of power meters: the uncertainty calculations, the test methods, the necessary equipment, and the industry standards. Also important are the tests performed during optical light source and variable attenuator verification.
Many companies find it advantageous to have an in-house calibration verification system for fiber-optic power meters, light sources, and variable attenuators.
Verifying Power-Meter Calibration
Power meters must be verified at regular intervals to ensure that the optical calibration constants—characterized by detector responsivity in amperes per watt of light received—are stable over time (Figure 1). The verification process is performed during power-meter calibration using a working reference standard. The quality and reliability of the response linearity and absolute power measurements are very important, especially when seeking a high degree of absolute accuracy (<±1%).
The two key issues characterizing power-meter calibration are absolute power calibration at one or more wavelengths and optical linearity with varying input power levels. Two methods are used to verify power-meter calibration:
Sequential—determines the calibration factor between the reference measurements and the power meter under test as well as total uncertainty.
Superposition—determines optical power linearity.
Sequential Method
The sequential method compares the reading of the power meter under test to that of a reference power meter. The reference power meter requires better accuracy and is traceable to the primary reference of a national standards laboratory, such as the National Institute of Standards and Technology (NIST), the U.K. National Physical Laboratory (NPL), the German Bureau of National Standards (PTB), or the National Research Council of Canada (NRC). To determine the calibration factor between the reference measurements and the power meter under test, the reference power meter must provide high precision with a maximum ±2% uncertainty.
A calibration verification system, including a light source and a reference power meter, ensures greater control over the entire calibration verification process. Calibration reference parameters include source spectral bandwidth, fiber type, connectors, environmental operating conditions, and the recommended power level.
Power stability must be considered when choosing the appropriate source for calibration verification. The spectral bandwidth of the selected source should not exceed 10 to 12 nm. A broadened linewidth distributed feedback laser with optimized coherence and ±0.003-dB power stability is a reliable source for calibration. Fabry-Perot lasers, with effective spectral widths of a few nanometers, are inexpensive sources for moderate precision calibration systems.
Fiber size should not vary. Differing fiber types will direct light to different portions of the central detector area. The connector and connector adapter also influence measurements. To minimize uncertainties, perform calibration with the connector and adapter to be used for measurements.
Ideal environmental operating conditions typically are 23° C with 50% relative humidity. As for the choice of a standard power level, NIST recommends -10 dBm. In addition, efficient absolute power calibration requires that the time between readings be as short as possible.
These parameters should be taken into account when evaluating power measurement uncertainty, which is divided into systematic and random. The calibration factor is calculated using
CF = PRPM(watt)/PDUT(watt)
where: PRPM = measured reference power
PDUT = measurement from the power meter under test
To perform precise calibration verification measurements, connectors, fiber ends, and detectors should be clean at all times. Test jumpers, connectors, and adapters must be identical to those used during calibration. A null measurement, with a protective cap on the detector, should be performed on the power meter under test prior to each test session. This compensates for the effects of dark-current detection and internal electronic offsets.
The source and power meter must stabilize after power up, according to specifications. The optical light source should be stable to minimize fluctuations in output power, better than ±0.003 dB over 15 min and even ±0.001 dB over a few minutes. After stabilization, perform an offset nulling on both the reference power meter and the power meter under test.
Take measurements on the reference power meter and the power meter under test. Make sure not to displace the patchcord when disconnecting the fiber from the reference power meter and connecting it to the power meter under test.
To have adequate statistics for the calibration factor, several sets of power readings are recommended. Calculation of the calibration factor and total uncertainty often is performed automatically by the software.
Superposition Method
Optical power linearity is important for optical component loss and other relative measurements, such as attenuator calibration and absolute power measurements at power levels other than the reference power.
A high-accuracy method, known as the superposition method, consists of splitting the power in two paths and recombining them on the power meter under test (Figure 2). The power from one path is first measured, then that of the other path. Finally, total power is measured from both paths simultaneously. If the sum of the individual powers differs from the total power, there is nonlinearity.
Global linearity is the sum of all local linearities, starting from the reference power level where nonlinearity equals zero. For higher power measurements, an optional optical amplifier can be used. To perform optical linearity measurement on fiber-optic power meters, connect the light source to the variable attenuator with an appropriate patchcord. Make sure all connectors and fiber-optic adapters are clean.
Connect each branch of the coupler to the two variable attenuators, each variable attenuator to a 1 × 2 coupler, and the coupler common port to the reference power meter. Perform power balancing. Disconnect the reference power meter, connect the power meter under test, and enter the power reading.
Verifying Light Sources and Attenuators
In addition to providing accurate power-meter calibration verification, some testing applications offer the possibility to verify light sources, attenuators, and other optical components for compliance to stated specifications. All optical light sources should be verified regularly for output power stability, sensitivity to reflections, and center wavelength. It is equally important to test variable attenuators for optical return loss, insertion loss, linearity, and repeatability.
Light-source output power fluctuates as a function of time. Source-power stability must be stated both on a short-term (for example, 15 minutes) and a long-term (>1 hour, 8 hours, or even 24 hours) basis, depending on the operating conditions and applications. Typical acceptable values are ±0.003 dB to ±0.005 dB for short-term and ±0.03 dB to ±0.05 dB for long-term.
To record the source-power level and stability, the optical light source is connected to a power meter, and its output is monitored as a function of time. The power meter wavelength must match that of the source, its dynamic range must be appropriate, and its resolution must be better than source stability specifications.
Source-power stability is calculated from the Bellcore Standard TR-TSY-000887, Generic Criteria for Fiber-Optic Stabilized Light Source, Section 5.1.2.
Source-Power Stability (dB) = Pmax-Pmin
where:
Pmax = maximum detected power
Pmin = minimum detected power
The resulting value is the total power variation over the selected time span.
Some light sources are sensitive to reflections, mainly because of their coherence length. Back reflections may cause output power to fluctuate dramatically. As a result, it is necessary to test the source sensitivity to reflections by monitoring the source output power against back reflections. The back reflection generator (VBR) must provide sufficient reflection range, its wavelength must match that of the source (angle physical contact connectors are mandatory), and all connectors and fiber-optic adapters must be absolutely clean.
For better accuracy, use two power meters. The first power meter is connected to the output port, nulled, and automatically set to the source wavelength by the software. This power meter will measure source-power stability.
The second power meter is connected to the monitor port of the VBR to measure reflection, taking into account the reflection from the output port in addition to that generated by the VBR. Calibration is performed automatically before measurements are taken.
Some applications require minimizing the uncertainty of the light-source center wavelength. The uncertainty is about ±10 nm but can reach ±30 nm in some cases. If a light-source central wavelength cannot be precisely located or if its spectral distribution cannot be defined precisely, this uncertainty can have dramatic effects on the power- meter calibration. Typically, a wavelength meter can measure the source center or peak wavelength with ±0.02 nm accuracy throughout the 400-nm to 1,600-nm range.
Attenuators, like light sources, can only be verified for their stated specifications. Insertion loss limits the minimum attenuation obtainable. Return loss can be critical when the attenuator is used with a laser.
Linearity is the capability of the attenuator to provide the requested attenuation values, and it needs to be tested over the available attenuation range. The attenuator also must be tested for repeatability since it is a critical parameter when attenuation scans are performed back and forth on each side of preset attenuation values.
Investing in an In-House Calibration System
Companies that require equipment to be calibrated on a regular basis often consider sending instruments to an external calibration service. Although the process guarantees proper calibration, the time delays involved can slow down testing schedules, ultimately reducing productivity.
Investing in an in-house verification system can be profitable. On one hand, it spot-checks whether your instruments are still within specifications. On the other, it circumvents unnecessary manufacturer calibration, saving time and calibration costs. Only out-of-specification instruments must be returned to the manufacturer for recalibration.
An efficient calibration and verification system can perform accurate, repeatable measurements through integrated software with step-by-step instructions allowing simple absolute power and nonlinearity verification. The calibration factor and total uncertainty automatically are calculated during power-meter calibration.
PC-based integrated test systems are flexible and easy to use for instrument calibration or verification. Thanks to the integrated application software, all calibration criteria are easily selected from a single setup window, test parameters can be configured, general information about the manufacturer and device under test is readily accessible, and time-saving features like pass/fail testing may be available (Figure 3).
With a PC-based system, test results are available in both graphical and table formats that can be analyzed in commercial data-base and spreadsheet programs. For instance, once data has been collected, it can be integrated into a spreadsheet program running on the same system.
About the Author
Farida Souiki is the product manager for EXFO Calibration Systems. Previously, she held various engineering positions for the Algerian National Oil and Gas and worked in international sales and marketing for different organizations. She also participated in commercial and economic activities for diplomatic missions. Ms. Souiki holds a B.S. in engineering from Sheffield University, U.K. EXFO Electro-Optical Engineering, 465 Godin Ave., Vanier, QC G2E 3A7, (418) 683-0211, www.exfo.com.
Copyright 1998 Nelson Publishing Inc.
September 1998
|
|