Abstract

Measuring short time intervals is a common and pressing problem in various areas of science and industry. In particular, repeated short intervals are an informative parameter for defect thickness measurement systems in nuclear physics, medicine, electronics, electrical engineering, etc. Method of direct comparison with the quantizing impulses period is commonly used for their measurement. The main disadvantage of this method is the minimum duration limitation of the measured time intervals with quantization tolerance, which depends on the maximum frequency of the quantizing impulses located at 600 MHz. Another problem related to this issue is restrictions of the application of the modern microcontroller components, which mainly operate at much lower frequency. Most microcontrollers use clock crystal oscillators with frequency output up to 50 MHz. One of the possible solutions for this problem is to use a special method of measuring the time interval, such as statistical one. The purpose of this paper is to research and design time intervals statistical meter on the basis of inexpensive microcontrollers and consumer components of lowspeed performance. The pulse duration is determined by the probability of the measured pulses coincidence and quantity pulses in the statistical method. This method allows to measure the time intervals which are much shorter than the quantizing impulses period and to use the low speed microcontrollers. Microcontrollers Atmel, ATmega and ATtiny with clock frequency of 16 MHz and 9,8 MHz were used in this project to build the time intervals statistical meter. The functional diagram of the device was developed. Large part of the functional blocks is implemented in software. This allows changing the software algorithm using programming methods to display debugging information and to correct errors. Such issues as methodological error depended on the number of measured time intervals (sample size) and the instrumental error, which associated with limited duration of quantizing impulse, were considered in this paper. The research of the time intervals statistical meter indicated that with the sample size of 10000 intervals and quantizing impulse duration of 0,01 μs, the measurement error in the range 1...10 μs does not exceed 0,6 %, while quantizing impulse duration is 0,05 μs - 3 %, which corresponds the preliminary calculations. The following research will be aimed at improving accuracy of the time intervals statistical meter by optimizing the sample size and the quantizing impulses duration.

Authors and Affiliations