Astron. Astrophys. 363, 1013-1018 (2000)

1. Introduction

The shortest timescales in the variability of the compact objects always was a very interesting topic of X-ray astronomy. The efforts to measure the highest frequency in the variability of Cyg X-1 can be traced back to rocket experiments in 70-ties (e.g. Rothschild et al. 1974, Giles 1981). It was found that Cyg X-1 demonstrates statistically significant variability of the X-ray flux at least up to frequencies of 10 Hz. Since then Cyg X-1 was extensively observed by various satellite missions searching for millisecond and submillisecond time scales variability (HEAO-1, see e.g. Meekins et al. 1984, Wen et al. 1996; EXOSAT, e.g. Belloni & Hasinger 1990; GINGA, e.g. Miyamoto & Kitamoto 1989). At the present time the Rossi X-ray Timing Explorer observatory (RXTE, Bradt et al. 1993) is the most powerful experiment for the investigation of the short time scale X-ray flux variations. It combines high telemetry rate (up to 512 kbps), large effective area of the detectors (6400 cm² in the soft X-ray band), high time resolution (down to 1 µsec) and comparatively low deadtime distortions. The comparison of the characteristics of several X-ray experiments can be found in e.g. Giles et al. 1998.

The shape of the power density spectrum (PDS) of Cyg X-1 in its low/hard spectral state is well established (in the range from to 50-100 Hz) since EXOSAT observations - it can be roughly represented by a constant from Hz up to some break frequency (), then it steepens to the power law with index 1 () and then it breaks again to the power law with the index 1.6-1.8, (see e.g. RXTE results in Nowak et al. 1999). Short observations (10 ksec) of Cyg X-1 by RXTE/PCA with high time resolution (s) showed the absence of significant variability of Cyg X-1 at time scales smaller than 3 msec (Giles et al. 1998).

In this paper we report the systematic study of very high frequency variability of Cyg X-1 in X-rays using large number of RXTE observations (total exposure time close to 190 ksec for the hard state and 10 ksec for the soft state) and carefully treating the deadtime effects.

© European Southern Observatory (ESO) 2000

Online publication: December 5, 2000
helpdesk@link.springer.de