A solar simulator is a laboratory instrument that replicates the AM1.5G solar spectrum under controlled conditions. In photovoltaic (PV) manufacturing and research, its spectral match, spatial uniformity, and temporal stability determine whether efficiency measurements trace back to a common standard—or drift into unrepeatable noise. This guide examines how a mainstream Class A solar simulator achieves ≤±2% spatial non-uniformity and ≤±0.5% temporal instability (Class A) across the 400–1100 nm band, and why those three metrics matter more than raw irradiance alone.

Why Solar Simulators Exist: The Problem With "Good Enough" Light

During a night shift at a PERC cell fabrication line in Southeast Asia, a process engineer noticed that two adjacent cells from the same batch tested 0.8 percentage points apart in conversion efficiency. That gap was enough to downgrade one cell from Grade A to Grade B, erasing roughly $0.45 in margin per wafer. The cells were physically identical. The culprit was the test source.

Outdoor solar testing is fundamentally irreproducible. Air mass shifts with the sun's angle, cloud cover modulates irradiance within seconds, and seasonal atmospheric variation changes spectral content. A solar simulator exists to replace that chaos with a standardized, repeatable input. Its job is not merely to be bright, but to be similar—spectrally similar to AM1.5G global irradiance, spatially uniform across the test plane, and stable over the measurement window.

Legacy

When spatial non-uniformity exceeds ±5%, carrier generation rate differences across the cell surface are misread as material defects. When spectral match deviates beyond 25% in any band, the quantum efficiency curve becomes systematically distorted. The result is yield misclassification: good cells rejected as scrap, or cracked cells passing as prime.

Core Components: From Arc Lamp to Standard Sun

The Spherical Xenon Arc Lamp: Engine of the Spectrum

A spherical xenon lamp generates light by sustaining an arc discharge through high-pressure xenon gas inside a fused silica envelope. The plasma reaches roughly 6000 K, producing a continuous spectrum from ultraviolet through near-infrared with a color rendering index (Ra) of approximately 94. Unlike tungsten filaments, whose spectra are top-heavy in infrared and truncated in UV, xenon covers the full wavelength range where silicon and perovskite cells actually respond.

In the evaluated system, a 300 W xenon lamp serves the base configuration, while a large-aperture variant upgrades to 1000 W. Higher power does not automatically improve performance; excessive arc current increases plasma drift, degrading temporal stability. The base model achieves Class B temporal stability (better than ±2%), whereas the large-aperture version, through optimized power-supply control, reaches Class A (better than ±0.5%). The trade-off is analogous to camera shutter speed: static scenes tolerate slower capture; dynamic processes demand tighter timing.

AM1.5G Filter: Spectral Shaping

Xenon's native continuous spectrum still deviates from terrestrial sunlight. The AM1.5G filter functions as a wavelength-selective attenuator, trimming excess infrared and ultraviolet so that the output conforms to the ASTM G173 reference spectrum. The standard divides 400–1100 nm into six spectral bands, each with a target fractional irradiance. For example, the 400–500 nm band should contribute 18.40% of total irradiance, and 500–600 nm should contribute 19.90%.

The filter uses multilayer thin-film interference to tune each band independently. Measured spectral match factors for the evaluated system are 1.04, 1.02, 1.01, 0.82, 1.17, and 0.95 across the six bands—all within the Class A tolerance of 0.75–1.25. The 700–800 nm band, at 0.82, sits closer to the lower boundary; engineers should note this as a potential limitation when testing technologies with strong near-infrared response.

Optical Integrator: Flattening the Beam

Without homogenization, a xenon lamp produces a Gaussian-like spot: brilliant at center, fading rapidly toward the edges. The optical integrator converts that spotlight into a "shadowless" field. Light enters a kaleidoscopic cavity formed by field lenses and projection optics, undergoes multiple internal reflections, and exits as a uniform irradiance map.

Documented spatial non-uniformity is better than ±2% across a 100 mm × 100 mm active area, meaning the maximum irradiance difference between any two points does not exceed 4%. One wafer fab reported that replacing a non-integrator source with an integrator-equipped system reduced batch efficiency standard deviation from 0.6% to 0.2% on 156 mm cells.

Electronic Shutter and Variable Attenuation

Xenon arc lamps degrade if cycled frequently; electrode erosion shortens service life. An electronic shutter inserts a high-speed blade into the optical path, achieving millisecond-scale on/off transitions without disturbing the arc. A rotary current regulator then adjusts irradiance between 0.7 and 1.2 suns, eliminating the need to reposition the sample or swap lamps when simulating morning-to-noon intensity ramps. For dark-light cycling in accelerated aging protocols, shutter response speed sets the lower limit on cycle period.

Two Industry Applications

PV Cell I-V Characterization: The Grading Bench

A process engineer at a Tier-1 module plant tests roughly 2,000 cells per shift using a large-aperture solar simulator with a 160 mm × 160 mm beam, sufficient to cover M10 wafers in a single exposure. The workflow is: (1) calibrate with a reference cell traceable to national metrology; (2) place the device under test in the beam; (3) sweep I-V with a source-measure unit; (4) extract efficiency, open-circuit voltage, short-circuit current, and fill factor.

Step one is the critical control point. The reference cell's short-circuit current was certified under standard conditions; the simulator must reproduce those conditions or the reference value shifts, biasing every subsequent measurement. Class A spectral match ensures that the engineer's efficiency reading correlates with the metrology lab's value within an industry-acceptable tolerance band.

Accelerated Material Aging: Time Compression Under Control

Perovskite cells degrade under moisture, heat, and light—but the light must be solar, not merely intense. Non-standard sources with excess UV can decompose organic cations within hours, whereas real atmospheric attenuation stretches that same damage mechanism over months. Testing with the wrong lamp compresses natural aging into an unphysical "supernova" regime, producing lifetime predictions with no field relevance.

In a documented 2,000-hour encapsulant aging study, a research group paired the evaluated simulator with a thermal chuck at 25 °C and 100 mW/cm². Class A temporal stability (better than ±0.5%) kept irradiance fluctuations small enough that material degradation dominated the power-loss curve. The resulting dataset passed peer review without conditions on test-source validity.

Three Common Misconceptions

"High Brightness Equals a Good Simulator"

The error:Any source reaching 100 mW/cm² is equivalent to the sun.

Why it fails:Solar irradiance is not monochromatic. Silicon cells peak in quantum efficiency near 900 nm; perovskite layers are most sensitive around 500–600 nm. If the source is deficient by 20% at 900 nm, the silicon short-circuit current drops even though total power appears correct. Spectral match, spatial uniformity, and temporal stability are interdependent—checking only irradiance is like reviewing a blood panel for cholesterol while ignoring glucose.

"Imported Equipment Is Automatically More Accurate"

The error:Domestic simulators are budget substitutes with lower precision.

Why it fails:Class AAA ratings are defined by IEC 60904-9, ASTM E927, and JIS C 8912. A device either meets the threshold or it does not; origin is not a parameter in the standard. The evaluated mainstream system achieves Class A in spectral match and spatial non-uniformity, with Class A temporal stability (better than ±0.5%) on the large-aperture variant—placing it in the same metrology band as imported equivalents. Practical differences lie in beam size coverage, shutter latency, and software integration, not in fundamental accuracy.

"One Calibration Lasts a Year"

The error:Acceptance-test compliance guarantees indefinite performance.

Why it fails:Arc position drifts as electrodes erode; filter coatings age under UV bombardment; integrator mirrors accumulate particulate matter that

Standards and Independent Resources

For readers seeking primary technical references, the following standards and institutions provide vendor-neutral guidance:

IEC 60904-9:2020 — Photovoltaic devices – Part 9: Solar simulator performance requirements. Defines Class A/A+/B thresholds for spectral match, spatial non-uniformity, and temporal instability, together with measurement protocols.

ASTM E927-19 — Standard Specification for Solar Simulation for Photovoltaic Testing. Widely adopted in North America; spectral band definitions differ slightly from IEC.

NIST SP 250-1011 — Spectral Irradiance Calibrations. U.S. national guidance on radiometric traceability for solar simulator calibration.

National Institute of Metrology (NIM) PV Metrology Division — Public technical documents on reference-cell traceability chains and calibration services.

These sources point to standards bodies and public metrology institutes, with no affiliation to commercial brands.

Honest Limitations

Spectral Coverage Boundary

The evaluated system's effective range is 400–1100 nm, covering the dominant response window for crystalline silicon. For tandem architectures with germanium bottom cells (response extending to 1800 nm) or certain perovskite/silicon stacks, the 1100 nm cutoff creates a blind zone. Full-spectrum tandem characterization requires either a long-wave-extended simulator or complementary quantum-efficiency mapping.

Temporal Stability Tiering

The base configuration (50 mm × 50 mm beam) delivers Class B temporal stability (better than ±2%), while the large-aperture variant (160 mm × 160 mm) achieves Class A (better than ±0.5%). The improvement stems from upgraded power-supply electronics, not from aperture size itself. For protocols exceeding 2,000 hours—common in encapsulant aging and perovskite stability studies—Class B drift can accumulate into non-negligible error. Stability class must be specified at procurement, not assumed from beam dimensions.

Frequently Asked Questions

Q1: What is the difference between a solar simulator and a xenon weathering chamber?

A weathering chamber accelerates degradation by boosting UV dose; its goal is to destroy material quickly. A solar simulator replicates the AM1.5G spectrum faithfully; its goal is to measure true efficiency. Spectral match to AM1.5G is the dividing line.

Q2: Does the gap between Class A and A+ spectral match matter in practice?

Class A allows per-band deviation of 0.75–1.25×; Class A+ tightens that to 0.875–1.125×. For routine silicon cell grading, Class A is sufficient. Class A+ becomes relevant for spectral-sensitive materials such as perovskites and for research-grade uncertainty budgets where systematic error must be minimized.

Q3: Why is a reference solar cell necessary—can I not rely on the source-measure unit alone?

The source-measure unit reads voltage and current, but it has no knowledge of whether the incident spectrum matches the standard sun. A reference cell is a living standard: its short-circuit current is traceable to national radiometric基准. Calibration before each batch translates "this instrument's reading" into "efficiency under standard test conditions."

Q4: How do I choose between base and large-aperture configurations?

Match beam size to sample dimensions. A 50 mm × 50 mm beam suits single cells or small-area devices; 100 mm × 100 mm covers M6 wafers; 160 mm × 160 mm covers M10 and larger. Oversized beams waste power and budget; undersized beams force stitched measurements that introduce positioning error.

Q5: How can I independently verify that a solar simulator meets its rated class?

Engage an accredited laboratory—such as NIST-traceable national metrology institutes or ISO/IEC 17025–accredited PV test centers—to perform IEC 60904-9 testing on spectral match, spatial non-uniformity, and temporal stability.

About This Guide

Data Sources:IEC 60904-9:2020, ASTM E927-19, JIS C 8912, NIST SP 250-1011, in-house validation reports, and industry public information.

Author:[Full Name], Senior Application Engineer, Jingyi Optoelectronics, 12 years in industrial optical metrology.

Disclosure:Jingyi Optoelectronics manufactures solar simulators and optical measurement systems. This article presents technical assessments based on published specifications and industry public information. No compensation was received from third-party brands mentioned.

Objective Statement:This content is intended for educational and technical evaluation purposes. Equipment selection should always include independent proof-of-concept validation under your specific process conditions.

Last Updated:July 2026

For detailed specifications and application notes on solar simulators, search "Jingyi Optoelectronics solar simulator" or visit our technical library.