Transmitter Impedance Adaptation for Serial Interface Training
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High-speed serial interfaces face challenges in ensuring reliable data transmission due to variations in circuit design, component manufacture, and environmental conditions, leading to suboptimal transmitter and receiver equalization mechanisms that require extensive iterative processes for channel loss compensation.
Innovation Solution
A system-aware transmitter adaptation method that adjusts impedance settings and compensation values based on bit error rate (BER) and receiver eye pattern characteristics, using a channel output module and channel management module to determine and record optimal impedance settings for subsequent iterations, optimizing data transmission.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If iterative equalization processes are used to optimize channel loss compensation, then data transmission reliability is improved, but link training time increases significantly
Solution Approach 1:
The patent applies preliminary action by pre-calculating and storing equalization compensation values in lookup tables before actual data transmission. The transmitter and receiver devices populate these tables with pre-computed compensation values based on anticipated channel conditions, allowing the system to quickly retrieve and apply appropriate compensation without performing iterative calculations during link training. This pre-computation approach significantly reduces link training time while maintaining reliable data transmission through accurate equalization.
Solution Approach 2:
The patent employs parameter changes by transforming the complex iterative equalization optimization problem into a simpler parameter selection problem. Instead of iteratively adjusting multiple equalization parameters simultaneously, the system pre-computes optimal compensation values for different channel conditions and stores them in lookup tables. During operation, the system simply selects the appropriate pre-computed parameters based on measured channel characteristics, achieving reliable equalization with minimal training time.
2Productivity
If best-effort equalization calibration is used for quick setup, then link training speed is improved, but compensation accuracy deteriorates
Solution Approach 1:
The patent applies preliminary action by pre-calculating and storing equalization compensation values in lookup tables before actual data transmission. The transmitter and receiver devices populate these tables with pre-computed compensation values based on anticipated channel conditions, allowing the system to quickly retrieve and apply appropriate compensation without performing iterative calculations during link training. This pre-computation approach significantly reduces link training time while maintaining reliable data transmission through accurate equalization.
Solution Approach 2:
The patent substitutes the mechanical iterative adjustment process with a lookup table-based retrieval system. Instead of using iterative algorithms that mechanically adjust equalization parameters through multiple measurement and adjustment cycles, the system replaces this process with direct lookup of pre-computed compensation values. This substitution eliminates the time-consuming iterative mechanism while providing accurate compensation through pre-calculated optimal values.
Data Source
AI summary
A high-speed serial interface includes a transmitter having an output module with settings that select an output impedances of the output module and a tuning value for the output impedance, and a receiver having a plurality of compensation modules each to provide a selectable level of equalization to a data bitstream from the transmitter, and a control module that directs the transmitter to successively select each of the tuning values, that directs the compensation modules, for each tuning value, to successively select each of the levels of equalization, that evaluates an indication of a performance level of the receiver for each of the successively selected levels of equalization and for each of the tuning values, and that selects a particular tuning value based upon the indications of the performance level of the receiver.


