A new highly precise angle measuring table has been developed for the Physikalisch-Technische Bundesanstalt, which is to be used as a comparator to calibrate angle standards. First results of the determination of the comparator's errors of measurement obtained by two calibration methods are reported. Considering the results obtained, it is to be expected that the desired uncertainty of measurement of can be achieved.

At the Physikalisch-Technische Bundesanstalt, autocollimators are calibrated with the aid of the WMT 220 angle comparator to ensure their direct traceability to the SI unit of plane angle, the radian (rad). It is shown that high-resolution electronic autocollimators can thus be calibrated with an uncertainty of 0.007 arcsec (k = 2) in reproducible measurement steps down to 0.005 arcsec. Calibrations in very small measurement steps close to the autocollimators' resolution can inform about possible short-period deviations and resulting aliasing effects in measurements with autocollimators.

No abstract

No abstract

A new method for the self-calibration of divided circles is presented which is based on a known prime factor algorithm for the discrete Fourier transform (DFT). The method, called prime factor division (PFD) calibration, is of interest in angle metrology specially for self-calibrating angle encoders, and generally for a significant shortening of the cross-calibration between two divided circles. It requires that the circular division number N can be expressed as a product N = R × S, whereby the factors R and S are relatively prime integer numbers. For the self-calibration of a divided circle, N difference measurements between R angle positions in a regular distribution and one reference angle position determined by S are evaluated by a two-dimensional DFT, yielding the N absolute division errors. The factor R is preferably chosen small, down to a minimum of R = 2, whereas the factor S may be as large as appropriate for the division number N of interest. In the case of a cross-calibration between two divided circles, the PFD method reduces the number of measurements necessary from N2 to (R + 1) × N. Experimental results are demonstrated for the calibrations of an optical polygon with 24 faces (prime factor product 3 × 8) and a gearwheel with 44 teeth (prime factor product 4 × 11).