Switched Capacitor Thermal Sensor Using N-Well Diodes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional thermal sensors in high computing power devices face challenges such as complexity, high cost, and accuracy issues due to degradation of PNP components and limited options in FinFET technologies, requiring costly trimming and multi-point calibration.
Innovation Solution
A switched capacitor thermal sensor using an N-well-to-substrate junction as a sensing element, employing capacitive bias and digitally controlled transistors to achieve high accuracy and low power operation, with a simple two-terminal diode design that is robust against process scaling and defects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional PNP-based temperature sensors are used, then temperature sensing function is achieved, but measurement precision deteriorates due to PNP component degradation and requires costly multi-point trimming
Solution Approach 1:
The patent replaces expensive, fragile PNP components with robust, inexpensive N-well diodes that are inherently available in CMOS processes. This substitution eliminates the need for costly multi-point trimming while maintaining measurement accuracy, directly resolving the contradiction between measurement precision and ease of manufacture
Solution Approach 2:
The patent changes the fundamental sensing parameter from PNP transistor characteristics (which degrade) to N-well diode characteristics (which are stable). By measuring voltage across the N-well diode at different temperatures and using these calibrated values for interpolation, the system achieves high precision without complex trimming procedures
2Measurement precision
If classic sensor circuits with bandgap reference and sigma-delta ADC are used, then measurement precision is improved, but device complexity increases significantly
Solution Approach 1:
The patent extracts and eliminates the complex bandgap reference and sigma-delta ADC circuits from the temperature sensor design. By using a simpler voltage measurement approach across the N-well diode with direct digital-to-analog converter control, the system maintains measurement precision while dramatically reducing circuit complexity
Solution Approach 2:
Instead of using complex analog circuits, the patent employs a digital approach where temperature is determined by comparing measured voltages against pre-stored calibration values in memory. This digital copying and comparison method simplifies the circuit while maintaining accuracy
3Reliability
If PNP components are used in FinFET technology, then temperature sensing is achieved, but reliability deteriorates due to micro-defects in junctions and limited fabrication options
Solution Approach 1:
The patent replaces unreliable PNP components with robust N-well diodes that are naturally formed in CMOS and FinFET processes. These N-well diodes are less susceptible to micro-defects and fabrication variations, significantly improving reliability while maintaining full compatibility with modern FinFET technology
Solution Approach 2:
The patent changes the sensing mechanism from PNP transistor operation (which is sensitive to junction defects) to N-well diode operation (which is more tolerant of process variations). This parameter change enables reliable temperature sensing in FinFET technology where PNP components are problematic
4Measurement precision
If multi-point trimming is performed to achieve measurement accuracy, then measurement precision is improved, but loss of time and productivity decrease
Solution Approach 1:
The patent performs the calibration action in advance by storing pre-measured N-well diode voltage values at different temperatures in memory during manufacturing. During operation, the system simply retrieves and compares these pre-stored values, eliminating the need for time-consuming multi-point trimming procedures and significantly improving productivity
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The solution provides high accuracy with zero or one-point trimming, low power consumption, and reduced area and complexity, making it suitable for FinFET technologies and small grid multi-spot temperature sensing.
Implementation Method 1
A switched capacitor thermal sensor uses an N-well-to-substrate junction, available in all CMOS and FinFET process technology nodes, as a thermal sensing element
Implementation Method 2
The diode is forward biased by a capacitor that is charged to a predefined voltage level and then discharged through the diode. A switching signal switches the capacitor between a charging state and the discharged state through the diode. The capacitor has a time constant that depends on a dynamic resistance of the diode
Data Source
AI summary
An apparatus comprises: a first circuitry to charge first and second capacitors to a predetermined voltage level; a second circuitry to discharge the first capacitor through a diode at a first time; a third circuitry to discharge the second capacitor through the diode at a second time, wherein the second time is greater than the first time; a comparator to compare a first voltage of the first capacitor with a second voltage of the second capacitor; and logic to adjust a scaling factor applied to the second voltage according to an output of the comparator.