Polarimetric studies of pulsar radio emission have contributed significantly to our understanding of both the geometry involved, and the emission properties of the radiation generated in the magnetospheres of these stars. A promising technique involving high-quality polarimetric measurements has been put forward by Blaskiewicz et al. 1991 (1991, hereafter BCW), and followed up by (von Hoensbroech & Xilouris 1997). The classical rotating vector model as developed by Radhakrishnan & Cooke (1969) was modified by BCW to include first-order relativistic corrections. They predict a time-lag between the centroid of the pulsar-profile and the steepest point of its polarization position angle (PPA) curve. This time lag is proportional to the emission altitude. While the BCW method has been applied successfully to decimeter wavelengths, presumably associated with high-altitude emission regions in the context of the radius-to-frequency mapping model (Cordes 1978), very little is known about emission heights at cm-wavelengths, which correspond to regions much closer to the pulsar surface. Such an investigation is initiated by our observations.
In the present paper, a procedure is presented that dynamically accounts for the differential gain variations introduced by the observing system. In addition, corrections are applied for spurious polarization introduced to the measurements by the system characteristics. The calibrated data are presented here while the procedure to derive emission altitudes from this data set as well as the significance of a stratification of the lower magnetospheric regions in pulsar magnetospheres is discussed in an accompanying paper (von Hoensbroech & Xilouris 1997).