ENHANCING THE ACCURACY OF MEASUREMENT CHANNELS IN AUTOMATION SYSTEMS BASED ON SWITCHED-CAPACITOR DELTA SIGMA ADCS

Abstract

This article investigates the critical problem of improving the conversion accuracy of primary analog signals, specifically pressure and temperature, within industrial automation systems. In modern industrial applications, such as the alcohol production industry, the overall accuracy of an automated control system is fundamentally limited by the precision of its primary sensors. Upgrading entire facilities with high-end, ultra-precise instruments is frequently economically unviable, creating a strong demand for cost-effective, mid-range sensors that can deliver superior accuracy without a proportional increase in manufacturing costs.

Traditional analog-to-digital converter (ADC) architectures typically rely on integrated continuous-time resistive networks for voltage division and feedback. However, these integrated resistors exhibit significant inherent drawbacks, including high susceptibility to temperature drift, substantial manufacturing tolerances, and elevated thermal noise levels. These physical limitations directly degrade the overall accuracy of measurement systems, often capping the performance of standard analog transmitters at a 0.5% error rate.

To comprehensively address this hardware limitation, this research justifies and implements the application of the Switched-Capacitor (SC) method within a 16-bit Delta-Sigma (ΔΣ) ADC architecture. By replacing conventional resistors with capacitors that are periodically switched by precise clock signals, the electrical charge is transferred with exceptional accuracy.

To rigorously validate the proposed architectural solution, an advanced mathematical model was developed in the MATLAB environment. The simulation incorporated realistic industrial constraints and physical imperfections, including a standard 4-20 mA current loop input converted to a 1-5 V full-scale range, random thermal fluctuations reflecting the actual workshop environment, additive hardware noise at a 0.2% level, and specific two-phase clocking dynamics at a 100 kHz sampling rate. A comprehensive statistical Monte Carlo analysis comprising 2000 independent iterations was performed to evaluate the robustness and repeatability of both the conventional resistive and the proposed SC architectures under identical dynamic conditions.

The findings theoretically confirm that implementing switched-capacitor techniques allows upgrading standard, cost-effective industrial sensors to a high-precision measurement class (with an error margin of 0.2% or better) using standard CMOS technologies, thereby paving the way for highly reliable automation systems in the food industries.