Arrays of energy=dispersive x-ray detector arrays are heavily used in biological research at synchrotron radiation facilities, particularly for absorption spectroscopy in fluorescence mode on dilute samples because an individual semiconductor spectrometer typically becomes pileup limited at 40-80,000 counts/sec. The present research proposes to increase the processing rate of the individual detectors by a factor of 10 or more by using the fact that x-rays from storage ring sources do not arrive randomly but rather only at times separated by fixed intervals. Therefore, if a photon can be completely processed in a shorter time, pileup in the detector can be completely eliminated. Because these processing times are shorter by 10 than those currently used, considerable innovation will be required to retain acceptable energy resolution. In this effort we will attempt to determine whether this goal can be reached by a combination of time variant filtering, detector redesign to minimize capacitance, and application of recent advances in the design of CCD amplifiers. Success would lead to detectors capable of counting a million counts/sec, pileup free, which would increase the productivity of these experiments by factors of 10-20.
