An IEC 60904-9 class is a useful summary, but the document that actually defends a measurement is the uncertainty budget: a line-by-line accounting, following the GUM (Guide to the expression of uncertainty in measurement), of every effect that moves the result and how much each contributes. If you are specifying a solar simulator, ask for the calibrated budget, not only the class.
The lines that usually dominate
- Reference-cell calibration: the traceable anchor for irradiance. A WPVS reference cell read through a precision transimpedance amplifier removes shunt-resistance drift, a hidden source of error in older designs.
- Spectral mismatch: the correction between the simulator spectrum, the reference cell and the device under test. A tighter spectral match (a higher IEC grade) makes this term smaller and better known.
- Spatial non-uniformity: how much the result depends on where the module sits in the test plane.
- Temporal instability: drift of the light during the I-V sweep, short-term and long-term.
- Temperature: the device and reference must be at, or corrected to, 25 C for an STC rating.
- Electrical measurement and contacting: voltage, current and series-resistance effects from how the device is wired.
How the terms combine
Independent contributions are combined in quadrature (root-sum-of-squares) into a combined standard uncertainty, then multiplied by a coverage factor, usually k = 2 for about 95% confidence, to give the expanded uncertainty quoted on a certificate. Because the terms add in quadrature, the largest one or two lines dominate the total, which is why reducing the biggest contributor matters far more than shaving the smallest.
Why simulator class shows up everywhere
The simulator's three IEC criteria are not one line in the budget, they influence several. Spectral match drives the spectral-mismatch term, non-uniformity is its own line, and temporal stability affects the electrical measurement. This is why a reference-class instrument lowers the whole budget at once. With a low-uncertainty reference and the DragonBack method, Pasan systems reach repeatability on the order of 0.2% on demanding modern modules.
You can build and explore a full GUM-compliant STC uncertainty budget for a solar simulator, including the spectral-mismatch calculation, with the Pasan uncertainties tool.
Frequently asked questions
- What goes into a solar simulator uncertainty budget?
- Reference-cell calibration, spectral mismatch, spatial non-uniformity, temporal instability, temperature, and the electrical measurement and contacting. Each is a line, combined per the GUM into a total.
- How are uncertainty contributions combined?
- Independent terms are combined in quadrature (root-sum-of-squares) into a combined standard uncertainty, then multiplied by a coverage factor, usually k = 2 for about 95% confidence, to give the expanded uncertainty on a certificate.
- Why does simulator class affect so much of the budget?
- The three IEC 60904-9 criteria feed several lines at once: spectral match drives the spectral-mismatch term, non-uniformity is its own line, and stability affects the electrical measurement. A higher class lowers the whole budget.
