Output Driver Partial Bit De-Emphasis for Signal Integrity

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional de-emphasis techniques in digital systems cause signal integrity issues due to asymmetric and full-bit period de-emphasis, leading to degraded signal quality and limited bandwidth.

Innovation Solution

The implementation of partial bit de-emphasis, which allows for adjustable de-emphasis time and amplitude, using a de-emphasis control circuit with a transition detector, variable delay circuit, and leg control circuits to manage the pull-up and pull-down circuits of the output driver, enabling symmetric de-emphasis and improved signal integrity.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of energy

If conventional de-emphasis is applied to compensate for channel loss, then signal amplitude is reduced to counteract capacitive filtering, but signal integrity deteriorates due to asymmetric and full-bit period de-emphasis

Engineering Contradiction:
Improvesignal amplitude lossVSAvoidsignal integrity
Core Design Contradiction:
Loss of energyVSReliability

Solution Approach 1:

The patent applies partial de-emphasis by controlling the pull-up and pull-down circuits to operate for only a portion of the bit period rather than the full bit period. This partial action reduces the de-emphasis effect on signal amplitude while maintaining enough compensation for capacitive filtering, thereby improving signal integrity without completely sacrificing amplitude compensation

Inventive Principle:
Principle #16Partial or excessive action

Solution Approach 2:

The patent dynamically adjusts the de-emphasis characteristics by varying the duration of pull-up and pull-down circuit operation within the bit period. The de-emphasis time is made variable rather than fixed, allowing optimization between amplitude compensation and signal integrity maintenance based on channel conditions

Inventive Principle:
Principle #15Dynamics

2Strength

If full rail-to-rail swing is provided at every bit transition, then signal amplitude is maximized, but subsequent repeated bits are过度 de-emphasized causing amplitude to drop below desired levels

Engineering Contradiction:
Improvesignal amplitudeVSAvoidamplitude control precision
Core Design Contradiction:
StrengthVSManufacturing precision

Solution Approach 1:

The patent uses partial bit period de-emphasis to prevent过度 de-emphasis of repeated bits. By limiting the de-emphasis duration to a portion of the bit period rather than the full period, the signal maintains adequate amplitude for repeated bits while still providing sufficient compensation for capacitive effects on transition bits

Inventive Principle:
Principle #16Partial or excessive action

Solution Approach 2:

The patent employs a control circuit that monitors bit transitions and adjusts the de-emphasis application accordingly. The leg control circuits receive control signals based on detected transitions and adjust the pull-up/pull-down circuit operation to achieve desired amplitude levels for both transition and repeated bits

Inventive Principle:
Principle #23Feedback

3Extent of automation

If delay chain is used to implement de-emphasis by sequentially turning on/off pull-up and pull-down circuits, then de-emphasis function is achieved, but output impedance varies dynamically degrading signal integrity

Engineering Contradiction:
Improveautomatic de-emphasis controlVSAvoidsignal integrity
Core Design Contradiction:
Extent of automationVSReliability

Solution Approach 1:

The patent merges the pull-up and pull-down circuit control into a unified structure where both circuits are controlled simultaneously by coordinated control signals. This simultaneous control maintains constant output impedance while achieving de-emphasis, unlike sequential control that causes impedance variations

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The patent maintains homogeneous (constant) output impedance during de-emphasis operation by ensuring both pull-up and pull-down circuits operate in a coordinated manner. The controlled variation of both circuits together preserves impedance homogeneity, preventing signal integrity degradation

Inventive Principle:
Principle #33Homogeneity

4Device complexity

If asymmetric de-emphasis is applied (strengthening only pull-up or only pull-down), then implementation is simplified, but signal integrity is degraded due to imbalance

Engineering Contradiction:
Improvecontrol circuit complexityVSAvoidsignal integrity
Core Design Contradiction:
Device complexityVSReliability

Solution Approach 1:

The patent intentionally introduces symmetric asymmetry by applying equal but opposite de-emphasis to both pull-up and pull-down circuits. This symmetric de-emphasis approach balances the effects on both voltage transitions, improving signal integrity while maintaining controlled circuit operation

Inventive Principle:
Principle #4Asymmetry

Solution Approach 2:

The patent uses counterbalancing control where the pull-up and pull-down circuits are controlled in a coordinated, opposite manner. When one circuit is activated for de-emphasis, the other is deactivated or activated correspondingly to maintain balance, preventing the signal integrity issues caused by asymmetric de-emphasis

Inventive Principle:
Principle #8Anti-weight (Counterweight)

Data Source

PatentUS10340913B2Apparatuses and methods for partial bit de-emphasis
Publication Date: 2019.07.02 MICRON TECHNOLOGY INC
  • US10340913B2 patent drawing
  • US10340913B2 patent drawing
  • US10340913B2 patent drawing

AI summary

Apparatuses and methods for partial bit de-emphasis are provided. An example apparatus includes an output driver and control circuit. The output driver includes a pull-up circuit including one or more pull-up legs, and a pull-down circuit including one or more pull-down legs. The control circuit may be coupled to the output driver and configured to receive an input signal having a first logical value and a second logical value, and in response to determining the logical transition has occurred from the second logic value to the first logic value, cause the pull-up circuit and pull-down circuit respectively to enter a first state for a duration of a first portion of a bit period and to enter a second state for a duration of a second portion of the bit period preceding the first portion.