We studied the properties of the low-frequency quasi-periodic oscillations detected in a sample of six black hole candidates (XTE J1550-564, H 1743-322, XTE J1859+226, 4U 1630-47,GX 339-4, XTE J1650-500) observed by the Rossi XTE satellite. We analyzed the relation between the full width half maximum and the frequency of all the narrow peaks detected in power density spectra where a type-C QPO is observed. Our goal was to understand the nature of the modulation of the signal by comparing the properties of different harmonic peaks in the power density spectrum. We find that for the sources in our sample the width of the fundamental and of the first harmonic are compatible with a frequency modulation, while that of the sub-harmonic is independent of frequency, possibly indicating the presence of an additional modulation in amplitude. We compare our results with those obtained earlier from GRS 1915+105 and XTE J1550-564.

We report the discovery of kHz quasi periodic oscillations (QPOs) in three Rossi X-ray Timing Explorer observations of the low mass X-ray binary (LMXB) XTE J1701-407. In one of the observations we detect a kHz QPO with a characteristic frequency of 1153 +/- 5 Hz, while in the other two observations we detect twin QPOs at characteristic frequencies of 740 +/- 5 Hz, 1112 +/- 17 Hz and 740 +/- 11 Hz, 1098 +/- 5 Hz. All detections happen when XTE J1701-407 was in its high intensity soft state, and their single trial significance are in the 3.1-7.5 sigma range. The frequency difference in the centroid frequencies of the twin kHz QPOs (385 +/- 13 Hz) is one of the largest seen till date. The 3-30 keV fractional rms amplitude of the upper kHz QPO varies between ~18 % and ~30 %. XTE J1701-407, with a persistent luminosity close to 1 % of the Eddington limit, is among the small group of low luminosity kHz QPO sources and has the highest rms for the upper kHz QPO detected in any source. The X-ray spectral and variability characteristics of this source indicate its atoll source nature.

We report variability of the X-ray source, X-7, in NGC 6946, during a 60 ksec Chandra observation when the count rate decreased by a factor of ~1.5 in ~5000 secs. Spectral fitting of the high and low count rate segments of the light curve reveal that the simplest and most probable interpretation is that the X-ray spectra are due to disk black body emission with an absorbing hydrogen column density equal to the Galactic value of 2.1 X 10^{21} cm^{-2}. During the variation, the inner disk temperature decreased from ~0.29 to ~0.26 keV while the inner disk radius remained constant at ~6 X 10^8 cm. This translates into a luminosity variation from 3.8 to 2.8 X 10^{39} ergs cm^{-2} sec^{-1} and a black hole mass of ~400 solar masses. More complicated models like assuming intrinsic absorption and/or the addition of a power-law component imply a higher luminosity and a larger black hole mass. Even if the emission is beamed by a factor of ~5, the size of the emitting region would be > 2.7 X 10^8 cm implying a black hole mass > 180 solar masses. Thus, these spectral results provide strong evidence that the mass of the black hole in this source is definitely > 100 solar masses and more probably ~400 solar masses.

Nearly $1.32 \times 10^6$ sec of non-zero count rate data were obtained from five years of {\rm RXTE} observations of the atoll source 4U 1636-53. The data was divided into 10309 segments of 128 sec each. The histogram of the number of segments as a function of count rate shows that the system can naturally be classified into four flux states. For each segment an automated search for kHz QPO was undertaken and the histograms of the maximum Leahy normalised power $P_{max}$ for the four flux states were created. These were fitted by probability distribution functions under the assumption that the intrinsic amplitude of the QPO has a normal distribution. The best fit distribution functions, showed that the fraction $f$ of the time the system exhibits a kHz QPO, decreases from near unity for the lowest flux state ($\approx 1400$ c/s) to zero for the highest one ($> 2550$ c/s) while the average amplitude remains nearly constant. Based on the best fit probability distribution functions, a threshold $P_{thre}$ was defined, such that 95% of the segments with $P_{max} > P_{thre}$ correspond to a real QPO signal. For segments selected by this criterion, the frequency versus count rate plot reproduced the earlier known ``parallel track'' variation with the tracks coinciding roughly with the flux states. It is shown that the gaps between the tracks are not caused by uneven sampling, but rather the QPO phenomenon is absent or weak when the system's flux level is intermediate between two flux states.