2.1

Requirements and characteristics

The main goal of SUSI (Figure 2) is to provide the same features as the Sun for the calibration subsystem of the telescope, ensuring a full illumination, with similar spectral response, spatial uniformity, divergence and irradiance levels to the Sun. The main characteristics and requirements of the instrument are related with the large aperture (>130 mm), divergence level of the beam (0.27° ±0.05°), spatial non-uniformity and irradiance level for a large spectral range (270 to 2385nm, covering the UV1, UV2, NIR, SWIR1 and SWIR3 ranges). The goal in creating a good solar simulator is to have these parameters match the sun’s parameters as close as possible. For Sentinel-5, the requirements are especially stringent, as it is challenging to achieve high irradiance levels, while maintaining a divergence matching the Sun’s, and very good spatial non-uniformity.

Figure 2.

SUSI

Table 1 summarizes the main characteristics of SUSI, based on experimental results from the test campaign.

Table 1.

Main characteristics of the Sun simulator (SUSI)

Light source

The optical design of the sun simulator was driven by the light source. For a continuous light flux, a short arc xenon lamp was selected, due to the good match with the spectral irradiance requirements. Although a Xenon arc lamp is widely used in sun simulators, for the stringent requirements of this instrument, it poses several challenges:

• 1) High thermal power - High optical power and low efficiency of the lamp leads to excessive heating. Sufficient clearance needs to be allowed between the optical components and the source, and the thermal stability needs to be accounted and analyzed in the design.

• 2) Time-dependent angular and spectral distribution – Xenon arc lamps produce their spectral emission creating a plasma between the cathode and anode. The light distribution is asymmetrical: more light emitted around the cathode than the anode. The spectral emissions have a spatial dependence – different wavelengths are emitted in different regions within the plasma. Moreover, the electrodes degrade during the lamp lifetime, which changes the angular and spectral distribution. These effects make the non-uniformity requirement even more challenging.

  To quantify this effect, the spectral angular distribution of a xenon lamp was measured, using a spectrometer in the far field, in the 270-900nm spectral range. The lamp was measured in two different lifetimes, 75h and 500h (Figure 3). Results show both the asymmetry, the spectral dependence and time effect on the angular distribution.

• 3) High frequency intensity fluctuations – Xenon arc lamps have high frequency intensity fluctuations. Considering the purpose of the Sun simulator for the on-ground calibrations, active spectral monitoring of the light source is fundamental to account for these fluctuations.

Figure 3.

The far field angular distribution of a xenon arc lamp was measured in Lusospace’s laboratory, in the spectral range 270-900nm, with the lamp in two different lifetimes, 75h and 500h. The angular distribution is wavelength dependent, and is asymmetric in y axis for 75h, and it spreads for 500h.

2.2

Optical design

SUSI’s optical design is composed of three parts: light source, homogenizer and telescope (Figure 4). A parabolic mirror collects and transmits the light from the light source into the homogenizer. The homogenizer consists on a custom-designed and made aluminum-mirrored kaleidoscope configuration, designed to achieve the intended spatial and spectral uniformity. The stringent spatial uniformity requirement demands special care in the design and optimization process. The measured spectral and time-dependent angular distribution of the Xenon lamp (Figure 3) was modeled in Zemax OpticStudio, and used in the overall design and optimization process. This modeling and respective optimization are crucial to ensure the performance throughout the lamp’s lifetime, especially for the spectral irradiance and spatial uniformity.

Figure 4.

Optical layout of the sun simulator

The telescope is a mirror-based Offner telescope, with an incorporated field mask to ensure the divergence. All mirrors are coated with protected aluminum, and the design ensures the required collimation and magnification to achieve a large exit pupil (>150mm). The telescope was designed to avoid large incidence angles, which would affect both spatial uniformity and polarization. The design was optimized such that the exit pupil is at 3178mm from the simulator’s envelope, although the distance can be easily changed and adapted.

2.3

Structure and subsystems

Figure 5 shows the CAD design of SUSI. The opto-mechanical support of the mirrors and field mask were designed with sufficient degrees of freedom for alignment and beam pointing. The baffle structure was selected to minimize straylight in the exit pupil, and highly anti-reflecting black coating is applied to straylight critical areas. This includes SUSI’s front surface: SUSI’s beam will be used through a window in the thermal vacuum chamber. The coating in the front surface prevents back reflections in the mirror to be reflected back to the vacuum chamber.

Figure 5.

CAD design of SUSI

Thermal and structural analysis were made to ensure the performance of the system in ISO 8 cleanroom conditions (22 ± 0.5°C), including a pointing stability of 36 arcseconds.

SUSI’s design includes a monitoring spectrometer, to perform active monitoring of the high frequency fluctuations of the light source, and a high speed shutter for the on ground calibration test campaign. It also includes an alignment cube and laser tracker balls for alignment purposes.

3.1

Spectral irradiance

The spectral irradiance was measured using a calibrated spectrometer for UV1, UV2 and NIR, and with IR photodiode and filters for SWIR1 (1620 and 1653nm) and SWIR3 (2318 and 2367nm). The results showed full compliance with the requirements. Table 2 shows the irradiance full each spectral band, and Figure 7 shows the spectrum for the range 270 to 773 nm.

Table 2.

Irradiance measurement results (photons.cm-2 nm-1 s-1)

Figure 7.

Spectral response from 270 to 773nm

3.2

Beam divergence

The beam divergence was measured considering the optical design of the system. Considering that the divergence is defined by a field mask, a theodolite, focused into infinity, was placed in the exit pupil position, and pointed in the opposite direction of the beam. With the focused image of the field mask in the theodolite, its angular aperture was measured. The divergence half-angle was measured to be ±0.318°, precisely the design value (which was maximized to improve the spectral irradiance).

3.3

Spatial non-uniformity and beam diameter

The beam diameter was measured to be 151mm. For such a large beam diameter, and for the intended measurement accuracy (0.05%), a dedicated test setup was developed. A fiber tip spectrometer was setup on top of a 2D stage, to perform a 2D scan of the entire exit pupil, with 3x3mm pixels (Figure 6). During scanning, SUSI’s shutter and monitoring spectrometer were used to normalize the data with dark current and source intensity fluctuations measurements, respectively.

The reported spatial non-uniformity values are peak-to-peak, according to IEC 60904-9 definition. Figure 8 presents the measured irradiance patterns for representative wavelengths of each spectral band. The pp spatial non-uniformity values are presented below each plot: Central are the values for a 60mm diameter region of the exit pupil, while Wide are the values for a 151mm diameter exit pupil. The results show that the irradiance patterns are extremely homogeneous. Across all spectral regions of interest, the spatial non-uniformity is below 1% for the central exit pupil; for the wide exit pupil, it is below 5%. The reported values are, to the best of the authors’ knowledge, the highest value of spatial homogeneity for large aperture sun simulators.

Figure 8.

Measured irradiance patterns for representative wavelengths in each spectral band, in percentage. The values below each plot are the pp spatial non-uniformity for the central inner exit pupil (60 mm diameter) and wide exit pupil (151mm)

3.4

Temporal stability of spatial homogeneity

The fact that the angular distribution of the Xenon lamp changes over the lamp’s lifetime might have an impact on the spatial homogeneity. As such, the spatial non-uniformity was measured 7 times, spread across 240 hours, with the lamp always turned on (approximately the estimated time required for Sentinel-5 test campaign), in UV1, UV2 and NIR spectral ranges (Figure 9). The results have confirmed the stability of the spatial homogeneity, maximum variation occurring in NIR, of 0.15%.

Figure 9.

Homogeneity stability. With the Xenon lamp always turned on, the spatial non-uniformity was measured 7 times spread across 240 hours.