Multi-Parameter Sensor Circuit for Process-Insensitive Temperature Sensing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Sensor circuits in integrated circuits that rely on a single physical parameter for temperature measurement are prone to inaccuracies due to process variations and mechanical stress, leading to unreliable data.
Innovation Solution
A model-based approach is employed using a control circuit to generate a linear regression model that correlates multiple observable physical parameters with the operating characteristic, such as temperature, to improve accuracy by accounting for variations and stress effects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single physical parameter is used for temperature measurement, then the sensor circuit structure is simple, but the measurement accuracy deteriorates due to process variations and mechanical stress
Solution Approach 1:
The temperature measurement function is segmented into multiple independent physical parameter measurements (voltage, current, resistance, frequency) rather than relying on a single parameter. Each parameter is measured separately by dedicated sensor circuits, and the results are combined through a linear regression model to achieve accurate temperature measurement while compensating for process variations and mechanical stress effects
Solution Approach 2:
The patent uses a composite measurement approach by combining multiple physical parameter measurements (voltage, current, resistance, frequency) into a unified temperature measurement system. The linear regression model integrates these diverse parameter measurements to produce a composite temperature value that is more accurate than any single parameter measurement alone
2Reliability
If multiple physical parameters are measured to account for process variations and stress, then the measurement accuracy improves, but the device complexity increases
Solution Approach 1:
The sensor system is designed with multi-functionality where multiple sensor circuits measure different physical parameters (voltage, current, resistance, frequency) that all relate to temperature measurement. This universal approach allows the system to gather comprehensive data about the operating conditions and process variations, improving reliability while the parameters can be measured within the integrated circuit structure
3Measurement precision
If a linear regression model is generated during calibration mode, then the temperature measurement accuracy improves, but the processing time and complexity increase
Solution Approach 1:
The linear regression model is generated in advance during a calibration mode before normal operation. During calibration, the system collects data from multiple physical parameter measurements at known temperatures and pre-computes the regression coefficients. This preliminary action ensures that during normal mission mode, the system can quickly apply the pre-established model without time-consuming calculations, thus improving measurement accuracy while minimizing processing time loss
Data Source
AI summary
A sensor system included in an integrated circuit includes multiple sensor circuits and a control circuit. Using characterization data, a model may be generated that defines a relationship between measurable parameters of the integrated circuit and an operating characteristic of the integrated circuit. The control circuit can combine, using a function included in the model, data from the multiple sensor circuits to determine a value of the operating characteristic that is more accurate than a sensor circuit configured to measure a single parameter of the integrated circuit that varies with the operating characteristic.