The power, peak wavelength, and spectral width of semiconductor lasers are three fundamental parameters characterising their output characteristics, corresponding respectively to the three dimensions of "how much energy", "where the colour is", and "how pure the colour is".
Testing of these three parameters is not an isolated operation but shares the same drive and temperature control platform. The reliability of test results is highly dependent on rigorous control of operating conditions and the traceability of instrument values.
I. Introduction: Why These Three Parameters Constitute a "Minimal Sufficient Set"
The division of labour among the three parameters in device screening and application matching: power determines the available energy budget, peak wavelength determines matching with optical fibres, detectors, or pump absorption bands, and spectral width is directly related to coherence, transmission dispersion, and sensing resolution. Together they form the basic basis for judging "whether a semiconductor laser can be used and where it should be used".
II. Common Foundations of Semiconductor Laser Testing: Condition Control Before Parameter Reading
1. Stability of Drive and Temperature Control
Both the power and wavelength of semiconductor lasers are sensitive to junction temperature. The typical drift of peak wavelength with temperature is on the order of 0.3 nm/K, and the threshold current also varies exponentially with temperature. During testing, the heat sink temperature and operating current must be recorded, and data should be the average of repeated tests (n≥10). If temperature control has not reached steady state, subsequent readings of the three parameters all lose comparative significance.
2. General Principles of Instrument Selection
Instrument accuracy should be at least 4 times better than the error limit of the measured parameter, with digital instruments having at least 3 significant digits and within their metrological validity period. The range and wavelength response of the optical power meter must cover the operating band of the device under test; the resolution of the optical spectrum analyser must be significantly smaller than the expected spectral width, otherwise the measured "spectral width" is actually a convolution with the instrument function.
III. Semiconductor Laser Output Power Testing: Path Selection from Continuous Wave to Pulse
1. Continuous Wave Power Testing
Configure the laser, drive power supply, attenuator (where applicable), and laser power meter as per the diagram. Adjust the beam so that the spot falls within the central area of the power meter‘s receiving surface. Key operations: before measurement, block with a shutter and zero the meter; select the range so that the reading falls within an appropriate part of the full scale. For devices with pigtail or free-space output, consider integrating sphere collection or direct incidence calibration methods respectively.
2. Pulsed Peak Power Testing
Under narrow pulse conditions, the average power meter reading cannot be directly equated to peak power. Peak current must be obtained by dividing the average current by the duty cycle, while the optical pulse waveform is acquired by a photodetector and oscilloscope; peak optical power is then derived from the pulse width and repetition rate. If the pulse width enters the nanosecond range, the bandwidth and rise time of the detection chain become the main sources of uncertainty and must be corrected or stated.
3. Result Presentation
When reporting power values, the test conditions must be attached: continuous wave or pulsed, operating current, heat sink temperature, and number of tests. "Output power" without conditions is not reproducible.
IV. Semiconductor Laser Peak Wavelength Testing: The Role and Limitations of the Optical Spectrum Analyser
1. Test Configuration and Procedure
The laser is coupled into the optical spectrum analyser via collimation or optical fibre. Set the scan range to cover the expected wavelength ±20 nm, and select the scan resolution according to the device spectral width. Apply the specified forward current, record the relative spectral intensity versus wavelength curve (I‑λ curve). The wavelength corresponding to the maximum relative spectral intensity on the curve is the peak wavelength. For multi‑longitudinal‑mode devices, if the power of each mode is close, the peak wavelength may jump between modes; the mode competition state must be noted in the report.
2. Resolution Constraints
The minimum resolution of grating‑type optical spectrum analysers near 1550 nm is approximately 0.01 nm (about 12 GHz). For narrow‑linewidth DFB or external‑cavity lasers with spectral widths far smaller than this, the "peak wavelength" position measured by the spectrum analyser is still credible, but the spectral shape has been broadened by the instrument function and cannot be used for linewidth analysis. In this case, peak wavelength can be obtained with higher precision using a wavemeter, while linewidth must be transferred to the unconventional means in the next section.
V. Semiconductor Laser Spectral Width Testing: From Direct Scanning to Delayed Self‑Heterodyne
Discussion is layered by spectral width magnitude – this is the core logic of this section. Different spectral widths correspond to fundamentally different testing principles, and mixing methods will lead to order‑of‑magnitude errors.
1. Conventional Spectral Width (≥0.1 nm level): Direct Measurement by Optical Spectrum Analyser
From the I‑λ curve, find the shortest wavelength λ₁ and longest wavelength λ₂ at 50% of the maximum relative spectral intensity, and calculate the full width at half maximum as Δλ_FWHM = λ₂ − λ₁. Applicable to FP cavity lasers and most multimode devices. Note: the scan resolution of the optical spectrum analyser should be set far smaller than the expected FWHM; otherwise, instrument convolution correction must be applied to the measured value.
2. Narrow Linewidth (MHz level and below): Delayed Self‑Heterodyne/Self‑Homodyne Method
When the spectral width is narrower than the resolution of the optical spectrum analyser, direct scanning fails. Use a delayed self‑heterodyne setup: the laser is split into two paths, one passing through a frequency shifter (or phase shifter) and a long delay fibre (delay time τ_d > 6τ_c, where τ_c is the coherence time). The two paths are combined and received by a detector, and the beat signal is fed into an RF spectrum analyser to record the power spectral density curve. In data processing, if the laser spectrum is Lorentzian, the −3 dB width of the beat spectrum is 2 times the linewidth under test; if Gaussian, it is √2 times; actual semiconductor lasers are mostly Voigt profiles, with the correction factor between the two. When estimating using the −20 dB width, the FWHM corresponding to a Lorentzian profile must be divided by a factor of √99 ≈ 10 (the conversion relationship between a 20 dB drop in power spectral density and the field spectrum width).
3. Necessary Information for Data Reporting
The spectral width report must state: test method (direct spectrometer scanning/delayed self‑heterodyne), lineshape assumption and correction factor, operating current, and temperature. Values from different methods cannot be directly compared.
VI. Semiconductor Lasers: Correlation of the Three Parameters and Standardisation of Test Reports
Power, peak wavelength, and spectral width are not independent variables: increasing the drive current raises power while also causing junction temperature to rise, which in turn red‑shifts the peak wavelength and may broaden the spectrum (depending on device structure); uncertainty in temperature control simultaneously contaminates all three readings.
Therefore, a reproducible test report should contain at least: device identification, heat sink temperature, operating current (continuous/pulsed and duty cycle), test method and instrument model with calibration status for each parameter, number of tests, and dispersion.
The value of reporting the three parameters together lies not in three isolated values but in their self‑consistency at the same operating point – if power is abnormal while wavelength is normal, or wavelength drifts while power is stable, this in itself is a clue for diagnosing device or test chain problems.
Shenzhen Zhongwei Inspection is a professional laser equipment testing and certification body in South China, with over ten years of deep engagement in laser product testing. It has long provided performance parameter testing and laser safety testing services for lasers, laser modules, and finished laser products to domestic and international customers.
Its laser performance testing includes semiconductor laser testing, solid‑state laser testing, fibre laser testing, and gas laser testing, with high efficiency and professional reliability.
If you have related testing needs, welcome to contact us for consultation and cooperation!
