Keith Jahoda, Scott Barthelmy, Frank Marshall, Arnold Rots
The bright gamma ray burst observed on 1996 September 24 (BATSE trigger 5614, GRB 960924) provides an opportunity to check the absolute timing of the RXTE data collection and analysis tools. We present the PCA lightcurve for this burst, observed in the coincidence rate, and discuss the timing standards and all sources of correction. We demonstrate that the RXTE clock agrees with the CGRO clock to better than 1 second, and anticipate that a careful correlation of the light curves will correlate the clocks to better than a few tens of milliseconds.
After receiving a report of a large BATSE trigger, we collected the Standard1 data from the SOF archives using the xff program with the real time option. RXTE was pointed at the galactic center and performing observations of GRO J1744-28 (public TOO data) at that time. The data were input to SAEXTRCT version 3.6. The output was futher processed by fcalc3.0b to produce a column with time expressed in UT seconds of 1996 September 24. The conversion from TT in the extracted file is accomplished by subtracting 86140802 seconds. We subtract 86140800 seconds to account for September 24 being mission day 997; we subtract 2 additional seconds to account for the leap seconds in July 1994 and January 1996.
Standard1 data accounts for every event passed to the EDS exactly once. The eight columns correspond to good rate from each of the 5 PCUs, the propane rate, the very large event rate, and the remaining rate. The last three rates are summed over the five detectors. The remaining rate is composed of all events which triggered 2 or more lower level discriminators (not including propane and very large events). These are normally considered background events. Figure 1 plots the rate of remaining counts centered on BATSE trigger 5614; the obvious peak corresponds to gamma-rays entering the back or sides of the detectors and creating activity. (There is no discernable increase in the good rates.)
Figure 1: Coincidence rate observed n RXTE PCA during burst GRB960924. The horizontal axis is UT seconds of 1996 September 24. No corrections for spacecraft position have been applied.
Batse light curves indicate that the peak occurs at 42119 seconds, in agreement with this result. Neither result has been corrected for spacecraft position. We anticipate that cross-correlation of the lightcurves will lead to a correlation of CGRO and RXTE time with a precision of better than several tens of milliseconds.