RTC Calibration Using Derived Fast and Slow Clock Signals
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Real-time clocks (RTC) face significant frequency errors due to initial and operating condition variations, which are not acceptable in many applications, especially when powered by batteries, requiring careful implementation to minimize supply current and avoid power and area increases from additional circuitry adjustments.
Innovation Solution
A calibrated RTC system generates faster and slower clock signals from an uncompensated clock, using edge detectors and dividers to distribute adjustments and reduce frequency errors by selectively replacing the uncompensated clock with these signals, generating a calibration strobe and window trigger to detect and correct errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If additional faster and slower clocks are introduced to adjust the uncompensated clock, then clock frequency accuracy is improved, but power consumption and die area increase
Solution Approach 1:
Instead of using separate additional clock generators, the patent creates copies of the uncompensated clock signal by generating fast and slow clock signals from the same source using edge detectors and dividers. This approach provides the necessary frequency variation for calibration while avoiding the power and area overhead of independent clock generators
Solution Approach 2:
The uncompensated clock signal serves multiple functions: it is the primary clock source and also the basis for generating both the fast and slow calibration clock signals. This multi-functionality eliminates the need for separate dedicated calibration clock sources, reducing overall power consumption and die area
2Measurement precision
If additional faster and slower clocks and circuitry are introduced to adjust the uncompensated clock, then clock frequency accuracy is improved, but die area increases
Solution Approach 1:
The patent generates fast and slow clock signals as copies/derivatives of the uncompensated clock signal using simple edge detection and division circuits, rather than implementing separate full clock generator circuits. This significantly reduces the die area required for calibration functionality
Solution Approach 2:
The calibration functionality is merged with the existing clock circuitry by using the uncompensated clock signal itself as the source for generating calibration clocks. The switching circuit integrates the selection of different clock sources, combining the uncompensated clock and derived fast/slow clocks into a unified calibration system that minimizes additional area
3Adaptability or versatility
If large adjustments are done in widely spaced, short bursts to correct clock error, then calibration capability is provided, but jitter is introduced in the compensated clock
Solution Approach 1:
The patent implements periodic calibration using window triggers that activate at regular intervals to perform frequency error detection and correction. This periodic action distributes the calibration burden evenly over time, preventing large sudden adjustments that would cause jitter while maintaining accurate calibration capability
Solution Approach 2:
The system performs preliminary frequency error detection using window triggers before executing clock adjustments. This allows the calibration logic to determine the appropriate correction amount in advance and apply it smoothly, avoiding reactive large-burst adjustments that would introduce jitter into the compensated clock signal
4Measurement precision
If digital circuitry is synchronized with the higher speed uncompensated clock, then timing precision is improved, but required current increases
Solution Approach 1:
The patent dynamically selects between different clock sources (uncompensated, fast, or slow) based on calibration needs and operational conditions. During normal operation, the system can use the lower frequency compensated clock to reduce current consumption, while still maintaining timing precision when needed by switching to the uncompensated clock for critical timing operations
Data Source
AI summary
Real-time clock calibration is accomplished by generating a fast clock signal and a slow clock signal from an uncompensated clock signal; selectively, momentarily, replacing the uncompensated clock signal with the fast and slow clock signal to generate a compensated clock signal; generating from the compensated clock signal a calibration strobe and window trigger; responding to the window trigger to detect any uncompensated clock signal frequency error and responding to the calibration strobe to selectively, momentarily, replace the uncompensated clock signal with the fast or slow clock signal to reduce the clock signal frequency error.