Confirming the accurate lack of wavelength dependence in the responsively of the integrating sphere

Using a spreadsheet, this is an example, the responsively at 5 degree intervals in both azimuth and zenith angles was calculated for this sphere. It is easily seen that if this integrating sphere were used for sunlight measurements, with the baffle facing South, negligible cosine errors would be prospective for all results where the zenith angle was less than 75 degrees. Some points near the baffle are some 2%-3% different for calculated and measured responses. This is probably to be the result of multiple localized reflections, each of which looses light through the entrance port, before total randomization within the sphere is achieved. These localized multiple reflection effects, although calculable, are offer little significant improvement in accuracy and beyond the scope of this simple model. Measurements on this sphere, in 15 degree zenith angle steps, at 90 degree and 0 degree azimuth verify that the calculated values of the response are pinpoint. The measured response was determined at wavelengths of 800nm, 700nm, 600nm, 500nm, 400nm and 300nm for all angles. No differences beyond normal experimental error were seen, confirming the accurate lack of wavelength dependence in the responsively of the integrating sphere.

Integrating sphere can be almost perfect devices when designed properly, that means giving highly accurate cosine response at all wavelengths. The integrating sphere response can be estimated precisely from relative simple formulae on a spreadsheet program, as confirmed by actual measurements. In addition, full 3D characterization is possible by calculation, providing detail to the generally incomplete data and enabling components to be optimized provided by some manufacturers. Utility designs can benefit from this modeling, and providing refinements in critical areas. A basic example, the Young & Schneider design, is commercially available from Optioned Laboratories, Inc and makes use of all of the optimization features.

The simple design ideas were used in integrating spheres of four inch and six inch diameter, modeling, and manufacturing. Since the results are close to ideal, they are best expressed as error relative to ideal cosine response. The sphere design, including exit attachments and mounting flanges for a dome window, had been shown is pictures, but not here. Measurements of the cosine response on both two types of integrating spheres agreed well with predictions, and the results are presented for the 6 inch sphere.