BBXRT consisted of two identical, coaligned telescopes (labeled A and B) with cooled Si(Li) detectors. The telescope mirrors were conical approximations to the Wolter Type I geometry (VanSpeybroeck and Chase (1972)). This allowed lightweight X-ray-reflecting foils to be used which could be nested as shown in Figure 1 to maximize the effective area at the expense of position resolution. Each BBXRT telescope had 118 coaxially nested mirrors.
The calibration of the instruments is discussed in detail by Weaver et al. (1995). Each segmented detector consisted of a central pixel (denoted A0 or B0), with a field of view 4 arcmin in diameter, surrounded by four outer pixels giving the detector a total field of view 17 arcmin in diameter. The point spread function was such that about 62% of the events from a point source fell on the central pixel for an on-axis pointing. The energy resolution was about 150 eV FWHM at 6 keV in the central pixels.
Photon arrival times were recorded with a resolution of 63 s. There were five photon quality flags recorded with each event. These used anticoincidence techniques to reject spurious events which were probably not due to X-ray photons. Events which were detected simultaneously with events in the same or other pixels were labeled with the pulse-pulse and pixel-pixel flags respectively. Events which had a deposited charge greater than about 14 keV were flagged as very large events (VLE). Firings of the opto-feedback circuits caused events to be labeled with the LED flag. The final anticoincidence flag was against triggering of the charged particle guard detector.