Error minimization in high-accuracy scanning deflectometry

A number of deflectometric devices have been developed and built for measuring optical surfaces for industrial and research applications. These include the scanning facility for the highly accurate and traceable measurement of large near-flat and slightly curved optical surfaces at PTB which is based on the Extended Shear Angle Difference (ESAD) deflectometry technique. ESAD combines deflectometric and shearing techniques in a unique way to enable error minimization and traceability. With this facility, sub-nm repeatability, reproducibility, and uncertainty of topography measurement have been achieved. A central optical element of the ESAD device and of several other deflectometric scanning systems is a pentagon prism which is used to deflect the light beam of an angle measuring device by 90 degrees. With ESAD, the pentagon prism is used to provide the lateral displacement of the measuring beam on the surface of the specimen to perform the shearing (angle difference) measurements, while other deflectometric systems make use of the pentagon prisms to scan the surface under test directly. It will be reported on the ESAD device using optimized opto-mechanical components, and on measurement results. Information on the adjustment of the pentagon prism, the specimen under test, and the axes of the autocollimator which has been achieved with the ESAD device will be presented. It will be reported on the reduction of the systematic errors in the angle difference measurement down to the milliarcsec level. Measurement comparisons involving ESAD deflectometry, an independent deflectometric scanning system, Fizeau interferometry using a mercury mirror, and absolute interferometry applying the skip flat method will be presented. Information on the autocollimator used as an angle measuring device and its calibration will be provided.