Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
IEC 61758: Fibre Optic Reference Source — Standardised Reference Light Sources for Fibre Optic Testing
✅ Standard at a Glance
IEC 61758 specifies the requirements for fibre optic reference sources used as standards for optical power and loss measurements in fibre optic systems. Developed by IEC Technical Committee 86 (Fibre optics), this standard defines the spectral characteristics, output power stability, connector interface requirements, and calibration traceability for reference sources used in laboratory calibration, field testing, and manufacturing quality control. While the average field technician may take the light source in their test kit for granted, the accuracy of every optical loss measurement ultimately depends on the quality and calibration of the reference source. IEC 61758 provides the standardisation framework that ensures measurement results are consistent, repeatable, and traceable to national metrology standards.
IEC 61758 specifies the requirements for fibre optic reference sources used as standards for optical power and loss measurements in fibre optic systems. Developed by IEC Technical Committee 86 (Fibre optics), this standard defines the spectral characteristics, output power stability, connector interface requirements, and calibration traceability for reference sources used in laboratory calibration, field testing, and manufacturing quality control. While the average field technician may take the light source in their test kit for granted, the accuracy of every optical loss measurement ultimately depends on the quality and calibration of the reference source. IEC 61758 provides the standardisation framework that ensures measurement results are consistent, repeatable, and traceable to national metrology standards.
🔌 1. Types of Reference Sources and Their Characteristics
1.1 Source Classification by Emitter Type
IEC 61758 classifies fibre optic reference sources according to the type of optical emitter used. Each emitter type has distinct spectral characteristics that determine its suitability for different measurement applications:
Stabilised Laser Diode Sources (Class L): These sources use a Fabry-Perot or distributed feedback (DFB) laser diode with integrated temperature stabilisation and power feedback control. Laser sources provide high output power (typically +3 dBm to +10 dBm at 1310 nm and 1550 nm), narrow spectral width (typically 0.1-5 nm for FP lasers, <0.1 nm for DFB lasers), and excellent short-term stability (±0.02 dB over 15 minutes after warm-up). Laser sources are essential for measurements that require high dynamic range, such as long-haul fibre attenuation characterisation and OTDR dead zone minimisation.
Stabilised LED Sources (Class E): These sources use high-power light-emitting diodes with thermoelectric cooling. LED sources provide lower output power (typically -20 dBm to -10 dBm) but much wider spectral width (50-150 nm). The wide spectrum reduces the impact of coherent effects such as speckle noise and connector interference, making LED sources ideal for multimode fibre loss measurements and passive component insertion loss testing where mode-filling conditions must be controlled.
Broadband Sources (Class B): These sources use amplified spontaneous emission (ASE) from an erbium-doped fibre amplifier or a superluminescent diode (SLD). They provide a smooth, broad spectrum (typically 40-80 nm for erbium-based sources, 20-60 nm for SLDs) with moderate output power (-10 dBm to +3 dBm). Broadband sources are used for wavelength-dependent loss (WDL) measurements of passive components, FBG sensor interrogation, and optical spectrum analyser calibration.
White Light Sources (Class W): These sources use a halogen lamp or xenon arc lamp coupled into the fibre. They provide the widest spectral coverage (400-1700 nm or more) but have low output power (typically -30 dBm to -20 dBm) and poor stability. White light sources are used primarily in laboratory equipment for spectral attenuation measurements and component characterisation over broad wavelength ranges.
💡 Engineering Insight
The choice between a laser source and an LED source for loss testing has a significant impact on measurement results, particularly for multimode fibre. IEC 61758 highlights a critical issue known as source-dependent loss: a multimode link may measure 0.5 dB loss with an LED source but 1.5 dB loss with a laser source, because the laser excites fewer modes, creating a different power distribution in the fibre. This discrepancy is normal and expected. The standard requires that the reference source type be matched to the deployed system source type for meaningful loss measurements. For single-mode systems, laser sources are the appropriate choice. For multimode systems with LED-based transmitters, LED reference sources must be used.
The choice between a laser source and an LED source for loss testing has a significant impact on measurement results, particularly for multimode fibre. IEC 61758 highlights a critical issue known as source-dependent loss: a multimode link may measure 0.5 dB loss with an LED source but 1.5 dB loss with a laser source, because the laser excites fewer modes, creating a different power distribution in the fibre. This discrepancy is normal and expected. The standard requires that the reference source type be matched to the deployed system source type for meaningful loss measurements. For single-mode systems, laser sources are the appropriate choice. For multimode systems with LED-based transmitters, LED reference sources must be used.
1.2 Spectral Characteristics and Power Stability
IEC 61758 defines the key spectral parameters that must be specified for every reference source and provides the measurement methods for verifying compliance:
Centre wavelength (λc): The nominal operating wavelength, typically 850 nm, 1300 nm (multimode), 1310 nm, 1490 nm, 1550 nm, or 1625 nm (single-mode). The standard specifies the tolerance on the centre wavelength: ±10 nm for LED sources, ±5 nm for FP laser sources, and ±1 nm for DFB laser sources. A wavelength error of 10 nm at 1550 nm causes approximately 0.003 dB/km of attenuation measurement error for standard single-mode fibre due to the wavelength dependence of Rayleigh scattering.
Spectral width (FWHM): The full-width at half-maximum of the source spectrum. IEC 61758 specifies the minimum spectral width for LED sources (typically 50 nm for multimode sources at 850 nm and 1300 nm) and the maximum spectral width for laser sources (typically 5 nm for FP lasers, 0.2 nm for DFB lasers). The spectral width affects the measurement’s sensitivity to wavelength-dependent loss features.
Output power stability: The maximum variation in output power over a specified time interval. IEC 61758 classifies stability into three grades: Grade 1 (±0.01 dB over 15 minutes, ±0.05 dB over 8 hours) for laboratory reference standards, Grade 2 (±0.03 dB over 15 minutes, ±0.15 dB over 8 hours) for field test equipment, and Grade 3 (±0.1 dB over 15 minutes, ±0.5 dB over 8 hours) for general-purpose sources.
| Source Class | Output Power | Spectral Width (FWHM) | Wavelength Tolerance | Stability Grade (Typical) | Primary Applications |
|---|---|---|---|---|---|
| L (Laser, FP) | +3 to +10 dBm | 1-5 nm | ±5 nm | Grade 2 | Single-mode loss testing, OTDR |
| L (Laser, DFB) | +3 to +10 dBm | <0.1 nm | ±0.5 nm | Grade 1 | WDM channel testing, high-precision |
| E (LED) | -20 to -10 dBm | 50-150 nm | ±10 nm | Grade 2-3 | Multimode loss testing, component IL |
| B (Broadband) | -10 to +3 dBm | 20-80 nm | ±2 nm | Grade 1-2 | WDL, FBG interrogation, OSA calibration |
| W (White light) | -30 to -20 dBm | >300 nm | N/A (continuous) | Grade 3 | Spectral attenuation, component characterisation |
🔬 2. Calibration and Traceability Requirements
2.1 Power Calibration Traceability
IEC 61758 requires that the optical output power of reference sources be traceable to national metrology standards through an unbroken chain of calibrations. The standard specifies the calibration hierarchy: Primary standard (at a national metrology institute, such as NIST in the US, PTB in Germany, or NIM in China) uses a cryogenic radiometer or electrically calibrated pyroelectric detector to realise the optical watt with uncertainty <0.1%. Transfer standard (a calibrated reference detector) transfers the calibration from the primary standard to calibration laboratories. Working standard (the reference source itself) is calibrated against the transfer standard and used for day-to-day measurements.
The calibration uncertainty budget must include contributions from the primary standard uncertainty, the transfer standard uncertainty, the source stability, the connector interface repeatability, the polarisation dependence of the measurement system, and the temperature coefficient of the source output. IEC 61758 specifies that the expanded uncertainty (k=2) of a working standard reference source must not exceed ±0.2 dB for Grade 1 sources, ±0.5 dB for Grade 2 sources, and ±1.0 dB for Grade 3 sources.
2.2 Spectral Calibration
The centre wavelength of the reference source must also be calibrated, as the wavelength directly affects the fibre attenuation measurement through the wavelength dependence of Rayleigh scattering and the water peak absorption at 1383 nm. IEC 61758 specifies that wavelength calibration be performed using an optical spectrum analyser (OSA) that is itself calibrated against known gas absorption lines (typically acetylene or hydrogen cyanide reference cells providing wavelength references accurate to ±0.02 pm).
⚠️ Critical Calibration Note
One of the most common sources of error in field loss measurements is the use of incorrect reference source calibration due to connector interface wear. The calibration of a reference source is typically performed with a reference-quality connector (IEC 61755 Grade A) terminated on a precision launch cable. After hundreds of field connections, the launch cable connector end-face degrades, changing the coupled power by 0.1-0.5 dB. This degradation is not detected by internal monitoring photodiodes, which measure the laser output before the connector interface. IEC 61758 addresses this by requiring that the reference source’s output power be specified at the reference plane (the output connector end-face), not at the internal source monitor. Regular inspection and replacement of the output connector is essential for maintaining calibration integrity.
One of the most common sources of error in field loss measurements is the use of incorrect reference source calibration due to connector interface wear. The calibration of a reference source is typically performed with a reference-quality connector (IEC 61755 Grade A) terminated on a precision launch cable. After hundreds of field connections, the launch cable connector end-face degrades, changing the coupled power by 0.1-0.5 dB. This degradation is not detected by internal monitoring photodiodes, which measure the laser output before the connector interface. IEC 61758 addresses this by requiring that the reference source’s output power be specified at the reference plane (the output connector end-face), not at the internal source monitor. Regular inspection and replacement of the output connector is essential for maintaining calibration integrity.
💡 3. Practical Guidelines for Reference Source Selection and Use
3.1 Selecting the Appropriate Source for the Application
IEC 61758 provides guidance on selecting the appropriate reference source class and grade for different measurement applications. The selection must balance measurement accuracy requirements against cost, portability, and ease of use:
Laboratory reference standard (Grade 1): For calibration laboratories and manufacturing QC facilities that require the highest accuracy. These systems typically use DFB laser sources with full thermoelectric temperature control, precision current sources, and optical feedback stabilisation. They are rack-mounted, require 30-60 minutes warm-up time, and need annual recalibration. Cost: $10,000-$30,000.
Field test equipment (Grade 2): For installation and maintenance technicians performing loss measurements on deployed fibre links. These instruments use FP laser or Fabry-Perot-stabilised LED sources with moderate temperature control and battery operation. They achieve adequate stability after 5-15 minutes warm-up and are calibrated every 12 months. Cost: $1,000-$5,000.
General-purpose source (Grade 3): For basic continuity checking, loss estimation, and non-critical measurements. These low-cost sources use uncooled FP lasers or basic LEDs with minimal stabilisation circuitry. They are suitable for quick checks and troubleshooting but should not be used for acceptance testing or guaranteed loss measurements. Cost: $200-$800.
| Application | Recommended Source Class | Recommended Grade | Wavelengths Required | Special Requirements |
|---|---|---|---|---|
| Long-haul SMF link acceptance | L (DFB or FP) | Grade 1 or 2 | 1310, 1550, 1625 nm | High stability, connector inspection |
| Multimode link acceptance | E (LED) | Grade 2 | 850, 1300 nm | Mode conditioning, mandrel wrap |
| PON/FTTx testing | L (FP) | Grade 2 | 1310, 1490, 1550 nm | Triple-wavelength, through-filter |
| Component insertion loss | E (LED) or B (Broadband) | Grade 1 or 2 | Application-specific | Mode control, polarisation control |
| WDL measurement | B (Broadband) or L (tunable) | Grade 1 | Tunable or broadband | Spectral calibration required |
| OTDR verification | L (FP or DFB) | Grade 2 | Per OTDR wavelengths | Pulse width matching |
3.2 Best Practices for Accurate Measurements
IEC 61758 outlines several best practices that significantly improve measurement accuracy when using reference sources:
Warm-up stabilisation: Allow the source to reach thermal equilibrium before making measurements. The warm-up time depends on the source design: thermoelectrically cooled sources require 5-15 minutes, while uncooled sources may require 30-60 minutes. Use the source’s stabilisation indicator (if available) or monitor the output power until it stabilises to within ±0.02 dB over a 5-minute period.
Connector interface maintenance: Inspect and clean the output connector before every use. A contaminated connector can reduce coupled power by 0.5-3 dB and introduce unpredictable measurement errors. Use connector inspection microscopes and appropriate cleaning tools (lint-free wipes with isopropyl alcohol or specialised dry cleaning systems).
Reference condition establishment: For loss measurements, establish the reference condition using a reference patch cord of the same type and length as the launch cable. The reference condition defines the 0 dB baseline. IEC 61758 specifies that the reference condition be re-established at least every 30 minutes during continuous testing or whenever the measurement setup is changed.
💡 Engineering Insight
The most frequently overlooked factor affecting reference source measurement accuracy is temperature-induced power drift. A laser diode’s output power changes by approximately 0.3-0.5 dB/°C without temperature stabilisation, and even stabilised sources have residual temperature coefficients of 0.01-0.03 dB/°C. When a field technician takes a light source from an air-conditioned vehicle (18 °C) to a sunny rooftop (40 °C), the source may take 30-60 minutes to restabilise. During this period, loss measurements will have a systematic error that exactly equals the power drift. The standard recommends that field technicians keep the reference source in its operating environment for at least 30 minutes before performing acceptance-critical measurements, and that the source be placed in a shaded, ventilated location during rooftop testing.
The most frequently overlooked factor affecting reference source measurement accuracy is temperature-induced power drift. A laser diode’s output power changes by approximately 0.3-0.5 dB/°C without temperature stabilisation, and even stabilised sources have residual temperature coefficients of 0.01-0.03 dB/°C. When a field technician takes a light source from an air-conditioned vehicle (18 °C) to a sunny rooftop (40 °C), the source may take 30-60 minutes to restabilise. During this period, loss measurements will have a systematic error that exactly equals the power drift. The standard recommends that field technicians keep the reference source in its operating environment for at least 30 minutes before performing acceptance-critical measurements, and that the source be placed in a shaded, ventilated location during rooftop testing.
❓ Frequently Asked Questions
1. How often should a reference source be recalibrated?
IEC 61758 recommends a calibration interval of 12 months for Grade 1 and Grade 2 reference sources, and 24 months for Grade 3 sources. However, the standard also recommends drift monitoring between calibrations: compare the source output against a stable reference detector at least monthly and plot the deviation on a control chart. If the drift exceeds half the specification tolerance, the source should be recalibrated immediately regardless of the scheduled interval. Sources that are used daily or in harsh environments may require more frequent calibration (every 6 months).
2. Can a reference source be used as a general-purpose light source for other fibre optic tests?
Yes, but with caution. A calibrated reference source is designed for power and loss measurements where output power stability is the primary requirement. Using it as a general-purpose source for communication system testing, as a transmitter replacement, or for applications requiring modulated signals is not recommended, as reference sources typically provide continuous-wave (CW) output only and may not have the modulation bandwidth or extinction ratio required for communication system testing. Additionally, using a reference source as a general-purpose source accelerates connector wear and may degrade its calibration.
3. What is the impact of connector type on reference source calibration?
The connector interface is part of the reference source’s calibration. If a source is calibrated with one connector type (e.g., FC/APC) and used with another (e.g., SC/APC), the calibration is invalidated because the connector adapter introduces a different insertion loss. IEC 61758 requires that the reference source’s calibration be connector-specific. If a source must be used with multiple connector types, it should be calibrated with a universal adapter (typically a 2.5 mm ferrule-based design that accommodates SC, FC, and ST connectors) and the adapter loss should be included in the uncertainty budget.
4. How does the reference source affect the uncertainty of an optical loss measurement?
The reference source contributes several components to the total measurement uncertainty budget: source power stability (typically 0.02-0.1 dB), calibration uncertainty (0.1-0.5 dB depending on grade), connector repeatability (0.05-0.2 dB), wavelength uncertainty (0.001-0.01 dB/km depending on wavelength accuracy and fibre type), and polarisation dependence (0.01-0.05 dB). In a typical field loss measurement setup (Grade 2 source, field-quality connectors, single-mode fibre at 1550 nm), the combined uncertainty is approximately ±0.3-0.5 dB (k=2). This means that a measured loss of 10.0 dB represents a true loss between 9.5 dB and 10.5 dB with 95% confidence. The reference source is typically the largest single contributor to this uncertainty.
