A project aiming at producing more efficient position sensitive gas proportional detectors and readout systems is presented. An area detector with reduced electrode spacing and a spatial resolution of 0.5mm and two time to digital convertors (TDC) based on ASICs were produced. The first TDC, intended for use with linear detectors, relies on time to space conversion, whereas the second one, for area detectors, uses a ring oscillator with a phase locked loop. A parallel readout system for multi-anode detectors aiming at a maximum count rate extensively uses RISC microcontrollers. An electronic simulator of linear detectors built for test purposes and a mechanical chopper used for attenuation of the X-ray beam are also briefly described.
A NIM module for delay line readout of linear gas proportional X-ray or neutron detectors based on time to space conversion is presented. Each of the 16 ASICs in the module contains a delay line with 64 elements each connected to a 24 bit-counter. Readout of the coincidence of signals travelling in opposite directions on the delay line and incrementation of the counters is triggered by the prompt anode signal so that simultaneous events are correctly processed. Transfer of the contents of the 1024 individual counters to a histogramming device at the end of a time frame can be as short as 60μs. With continuous delay lines the total conversion time equals the transit time of the delay line in the detector, whereas with segmented delay lines the total conversion time is around 20ns.
Characterization results are given for an original ASIC allowing continuous acquisition of ionising radiation images in spectroscopic mode. Ionising radiation imaging in general and spectroscopic imaging in particular must primarily be guided by the attempt to decrease statistical noise, which requires detection systems designed to allow very high counting rates. Any source of dead time must therefore be avoided. Thus, the use of on-line corrections of the inevitable dispersion of characteristics between the large number of electronic channels of the detection system, shall be precluded. Without claiming to achieve ultimate noise levels, the work described is focused on how to prevent good individual acquisition channel noise performance from being totally destroyed by the dispersion between channels without introducing dead times. With this goal, we developed an automatic charge amplifier output voltage offset compensation system which operates regardless of the cause of the offset (detector or electronic). The main performances of the system are the following: the input equivalent noise charge is 190e rms (input non connected, peaking time 500ns), the highest gain is 255mV/fC, the peaking time is adjustable between 200ns and 2μs and the power consumption is 10mW per channel. The agreement between experimental data and theoretical simulation results is excellent.
A new partially integrated magnetometer based on continuous-wave proton nuclear magnetic resonance (NMR) has been realized and characterized. Planar coils on a glass substrate of diameter down to 2 mm bonded to a CMOS IC for NMR signal detection and amplification, with cis-polyisoprene samples of volume down to 1 mm3, have been used to show the feasibility of a fully integrated NMR magnetometer. The sensor has been tested at static magnetic fields between 1 and 2 T. A precision better than 5 ppm (at 1.4 T) is achieved.
A voltage reference which yields a voltage proportional to a reference voltage (V/sub REF/) and relative temperature (/spl Delta/T) is presented. The application, typically the interfacing of sensors whose output signal is proportional to V/sub REF/, is described. The principle of the voltage reference is presented as an arithmetic combination of temperature dependent sources and of a reference dependent source. The description of the realization of the main blocks is followed by the simulation results. The expected precision of the temperature measurement is better than 1/spl deg/C, for V/sub REF/ variations of /spl plusmn/10% around its nominal value.