July 1993, Volume 2, Number 2
Sensitivity is one of the fundamental parameters of the Nuclear Magnetic Resonance experiment. It depends particularly on the coil and impedance matching of the resonance circuit of the probe to the input amplifier. However, sensitivity is also influenced by the noise of the receiver of the spectrometer, and the insertion loss of the matching and switching circuits as well as of passive fillers inserted in front of the input preamplifier manifests itself markedly. In this paper, a simple method of measuring the noise figure feasible in the NMR laboratory is described. Using non-standard probes and coils, the measurement of the noise figure enables one to estimate to what degree the detection system itself participates in the signal/noise ratio deterioration. The method can be used also for measurements of other radio-frequency signal processing systems.
This paper describes the design of a Nonius Flash Analog to Digital Convertor for video-applications. The conversion is split up in two stages: the MSB and the LSB conversion stage. In the LSB conversion stage, the reference-voltage for the comparators is shifted up and down as a nonius of a vernier gauge, according to the MSB conversion obtained in the first half clock cycle. The MSB's and LSB's are simultaneously available at the output to make a 8-bit representation of the analog signal. Using the subranging mechanism, the number of comparators is reduced from 256 to 32 and the Nonius approach results in a very simple configuration for the subranging switches. The design has been simulated with HSPICE in ES2 1. 5 CMOS technology.
A new circuit of frequency dependent negative resistor (FDNR), derived using an almost linear s-z transformation in a four phase switched capacitor (SC) network is presented.
In this paper a novel approach to the speech-and-channel coding system is proposed. The main idea of this approach is splitting an input speech signal into its integer and fractional parts and their different coding to fulfil the demand of bit-rate-reduction and unequal error protection, respectively. The experiments with the fractional part of the signal verify this hypothesis and the results of the average SNR of the input and output speech signal support a practical implementation of the system.
The contribution summarizes the results of the mean code acquisition time calculations of multiple - dwell serial search detector with fixed false-alarm penalty. The repeated calculations enable to choose optimum values of the synchronizer parameters.