Trimming Circuits for Nonlinear Temperature Drift Compensation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing temperature trimming methods for electronic devices only eliminate the linear component of temperature drift, failing to compensate for nonlinear fluctuations, which limits the accuracy of parameters such as voltage references to around 5 ppm/°C or worse.
Innovation Solution
A method that adjusts output parameters at multiple temperatures using both linear and nonlinear temperature-dependent variables, specifically trimming the nonlinear component to zero at various temperatures to maintain parameter stability across a wide temperature range, employing techniques like Taylor polynomial and exponential function approximations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional single-temperature trimming is performed, then the linear temperature drift is eliminated, but nonlinear temperature fluctuations remain uncompensated
Solution Approach 1:
The temperature drift compensation is segmented into two independent components: linear temperature dependence and nonlinear temperature dependence. The linear component is eliminated by trimming the temperature-dependent variable to zero at a first temperature, while the nonlinear component is compensated by trimming at multiple temperatures (first, second, and third temperatures). This segmentation allows each component to be addressed with appropriate trimming strategies, resolving the contradiction between eliminating linear drift and compensating for nonlinear fluctuations.
2Ease of manufacture
If trimming is performed at only one temperature, then the process is simple, but the output parameter cannot maintain stability across wide temperature ranges
Solution Approach 1:
The trimming process performs preliminary actions at multiple temperatures before final operation. First, the temperature-dependent variable is trimmed to zero at a first temperature (e.g., 100°C). Then, additional trimming is performed at a second temperature (e.g., 20°C) and a third temperature (between the first and second temperatures) to compensate for nonlinear effects. These preliminary trimming actions at different temperatures prepare the system to maintain stability across the full operating temperature range, resolving the contradiction between process simplicity and parameter stability.
3Measurement precision
If multi-temperature trimming with nonlinear compensation is implemented, then output parameter accuracy is significantly improved, but the trimming process complexity increases
Solution Approach 1:
The trimming process extends from a single temperature point into the temperature dimension by performing trimming at multiple temperatures (first, second, and third temperatures). This dimensional expansion allows the system to capture and compensate for nonlinear temperature dependencies that cannot be addressed at a single temperature point. The additional temperature dimension provides the necessary information to calculate and apply nonlinear compensation, resolving the contradiction between accuracy improvement and process complexity by systematically utilizing temperature as an additional degree of freedom.
Data Source
AI summary
Methods and circuits for adjusting the output parameter of a device wherein the output parameter is temperature dependent are disclosed herein. An example of a method includes: adjusting the output parameter to a target level at a first temperature; adjusting a linear temperature-dependent variable related to the output parameter to zero at the first temperature; adjusting a nonlinear temperature-dependent variable related to the output parameter to zero at the first temperature; adjusting the output parameter to the target level at a second temperature using the linear-dependent variable; adjusting the nonlinear temperature-dependent variable to zero at the second temperature; and adjusting the output parameter to the target level at a third temperature by adjusting the nonlinear variable.


