SAX J2103.5+4545 has been continuously monitored for ∼900 d by the Rossi X‐ray Timing Explorer (RXTE) since its outburst in 2002 July. Using these observations and previous archival RXTE observations of SAX J2103.5+4545, we refined the binary orbital parameters and find the new orbital period as P= (12.665 36 ± 0.000 88) d and the eccentricity as 0.4055 ± 0.0032. With these new orbital parameters, we constructed the pulse frequency and pulse frequency derivative histories of the pulsar and confirmed the correlation between X‐ray flux and pulse frequency derivative presented by Baykal, Stark & Swank. We constructed the power spectra for the fluctuations of pulse frequency derivatives and found that the power‐law index of the noise spectra is 2.13 ± 0.6. The power‐law index is consistent with random walk in pulse frequency derivative and is the steepest among the HMXRBs. X‐ray spectra analysis confirmed the inverse correlation trend between power‐law index and X‐ray flux found by Baykal et al.

XMM-Newton observed SAX J2103.5+4545 on 2003 January 6, while the Rossi X-Ray Timing Explorer (RXTE ) was also monitoring the source. Using the RXTE Proportional Counter Array data set between 2002 December 3 and 2003 January 29, the spin period and average spin-up rate during the XMM-Newton observations were found to be 354:7940 AE 0:0008 s and (7:4 AE 0:9) ; 10 À13 Hz s À1 , respectively. In the power spectrum of the 0.9-11 keV EPIC PN light curve, we found quasi-periodic oscillations (QPOs) around 0.044 Hz (22.7 s) with an rms fractional amplitude of $6.6%. We interpreted this QPO feature as the Keplerian motion of inhomogeneities through the inner disk. In the X-ray spectrum, in addition to the power-law component with high-energy cutoff and the $6.4 keV fluorescent iron emission line, we discovered a soft component consistent with blackbody emission with kT $ 1:9 keV. The pulse phase spectroscopy of the source revealed that the blackbody flux peaked at the peak of the pulse with an emission radius of $0.3 km, suggesting the polar cap on the neutron star surface as the source of the blackbody emission. The flux of the iron emission line at $6.42 keV was shown to peak at the off-pulse phase, supporting the idea that this feature arises from fluorescent emission of the circumstellar material around the neutron star rather than the hot region in the vicinity of the neutron star polar cap.

We present timing and spectral analysis of RXTE‐PCA (Proportional Counter Array) observations of the accretion powered pulsar 4U 1907+09 between 2007 June and 2008 August. 4U 1907+09 had been in a spin‐down episode with a spin‐down rate of −3.54 × 10−14 Hz s−1 before 1999. From RXTE observations after 2001 March, the source showed a ∼60 per cent decrease in spin‐down magnitude, and INTEGRAL observations after 2003 March showed that source started to spin‐up. We found that the source recently entered into a new spin‐down episode with a spin‐down rate of −3.59 × 10−14 Hz s−1. This spin‐down rate is pretty close to the previous long‐term spin‐down rate of the source measured before 1999. From the spectral analysis, we showed that hydrogen column density varies with the orbital phase.

We analysed RXTE archival observations of 4U 1907+09 between 1996 February 17 and 2002 March 6. The pulse timing analysis showed that the source stayed at almost constant period around 1998 August and then started to spin-down at a rate of (−1.887 ∓ 0.042) × 10 −14 Hz s −1 which is ∼0.60 times lower than the long-term (∼15 yr) spin-down rate. Our pulsefrequency measurements for the first time resolved significant spin-down rate variations since the discovery of the source. We also presented orbital phase resolved X-ray spectra during two stable spin-down episodes during March. The source has been known to have two orbitally locked flares. We found that X-ray flux and spectral parameters except hydrogen column density agreed with each other during the flares. We interpreted the similar values of X-ray fluxes as an indication of the fact that the source accretes not only via transient retrograde accretion disc but also via the stellar wind of the companion, so that the variation of the accretion rate from the disc does not cause significant variation in the observed X-ray flux. Lack of significant change in spectral parameters except hydrogen column density was interpreted as a sign of the fact that the change in the spin-down rate of the source was not accompanied by a significant variation in the accretion geometry.

We present analysis of RXTE-PCA observations of GX 1+4 between March 3, 2001 and January 31, 2003 together with the CGRO-BATSE Xray flux and frequency derivative time series between 1991 and 1999. From the timing analysis of RXTE-PCA observations, we are able to phase connect pulse arrival times of the source within two different time intervals and obtain corresponding timing solutions. Using these pulse arrival times, we contribute to long term pulse frequency history of the source. We look for episodic correlations and anti-correlations between torque and X-ray luminosity using CGRO-BATSE X-ray flux and frequency derivative time series and find that correlation state of GX 1+4 seems to change on ∼ 100-200 days long intervals. We estimate torque noise of the source and observe flickering noise ( f −1 ). We achieve to measure the longest observed timescale for a noise process among accretion powered X-ray pulsars by extending the noise estimate for a time scale ranging from 31 days to 44 years. Spectral analysis of individual RXTE-PCA observations indicates a significant correlation between iron line flux and unabsorbed X-ray flux. Pulse phase resolved spectra of the source indicate a broadening of iron line complex at the bin corresponding to the pulse minimum.