In general, a detector fulfils the condition of linearity (see § III.2.) only for a limited range of the input signal level. There are two effects which define the boundaries of this dynamic range:
• At very low levels of the input signal, the detector’s output is largely dominated by noise. Noise is a random temporal fluctuation of the output signal which occurs even when the input signal is constant. The absolute level and the frequency distribution of these variations depends on the physical properties of the detector and the subsequent electronics. For many detectors, noise is largely independent from the absolute level of the input signal and can be neglected for input signals above a certain minimum level. However, for very low input signals, the output signal is dominated by noise and does no longer quantify the physical quantity which should be determined. The lower limit of the measurement range, which is posed by noise, is quantified by the noise equivalent input. The CIE defines the noise equivalent input as the value of the respective physical quantity (radiant power or luminous flux, irradiance or illuminance, ...) that produces an output signal equal to the root mean square noise output. As the shape of the noise signal depends on the temporal resolution that can be achieved of the recording electronics (often characterised by the electronics’ time constant), the noise equivalent input is defined for a stated frequency and bandwidth. Unless otherwise stated, a 1 Hz bandwidth is usually considered. Depending on the detector’s characteristics, its noise level can be reduced by longer detector integration times or by averaging subsequent measurements of the same input signal.
• At high levels of the input signal, the detector’s output signal no longer increases proportional to its input signal, and thus the detector no longer fulfils the condition of linearity (see § III.2.). Instead, physical limits of the light sensitive element and / or the electronics cause saturation of the output signal, which increases less than proportional to the input signal and finally reaches a constant level. To a certain extent, subsequent correction of the detector’s output signal can account for the effects of saturation and thus extend the detector’s dynamic range. This correction has to be based on a thorough laboratory investigation of the detector’s dynamic behaviour and still leads to higher measurement uncertainties at high levels of the input signal.
The detector dynamic range depends on the photodiode type The overall measurement system dynamic range will depend on both the detector and electronic meter’s range capabilities. For example, a typical silicon photodiode is capable of measuring more than 2 mA of current before saturating, however the upper current measurement range of the meter may be limited to 200 µA.
This range covers extremely low intensity levels, for instance the quantification of erythemally active UV radiation, or very high intensity levels, which are used for industrial UV curing processes.