# US 11,909,565 B2 · Read eye training

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 11,909,565 B2
- Document ID: US11909565B2
- Status: Granted
- Kind: B2
- Application number: 17/879,611
- Filed: 2022-08-02
- Published: 2023-02-16
- Granted: 2024-02-20
- Earliest claimed: 2021-08-11
- Assignees: Cadence Design Systems, Inc.
- Applicants: Cadence Design Systems, Inc.
- Inventors: Anirudha Shelke; Ashwin S. Madhavakaimal; Kiran Baby
- Web page: https://patentgrep.com/patent/US11909565B2

## CPC

- H04L 25/03057: {with a recursive structure (H04L25/03031 takes precedence)}
- H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M)

## Abstract

Decision feedback equalization (DFE) is used to help reduce inter-symbol interference (ISI) from a data signal received via a band-limited (or otherwise non-ideal) channel. In embodiment, a single-ended receiver trains DFE coefficients and the slicer reference voltage to improve the received eye height. The process for training avoids many whole range sweeps thereby shortening training time. A custom data pattern that includes low-frequency (DC with respect to DFE) and high-frequency (AC with respect to DFE) worst cases is used for training in a closed loop manner. Negative DFE is used to measure the AC height of the data. Positive DFE is used to find the DC height of the data pattern.

## Claims

### Claim 1

A method of calibrating a receiver with decision feedback equalization (DFE), comprising:
measuring, at a first sample timing and by varying a DFE coefficient of the receiver, a first eye height while receiving a first data pattern;
measuring, at the first sample timing by varying the DFE coefficient, a second eye height while receiving a second data pattern; and
based on the first eye height and the second eye height, adjusting the DFE coefficient to a first setting.

### Claim 2 (depends on 1)

The method of claim 1, wherein the first setting provides DFE that is a first average of the first eye height and the second eye height.

### Claim 3 (depends on 1)

The method of claim 1, further comprising:
adjusting a reference voltage of a slicer from a first value used during measurements of the first eye height and the second eye height to a second value;
measuring, at the first sample timing, while using the second value, and by varying the DFE coefficient of the receiver, a third eye height while receiving the first data pattern;
measuring, at the first sample timing, while using the second value, and by varying the DFE coefficient, a fourth eye height while receiving the second data pattern; and
determining whether a first difference between the first eye height and the second eye height is greater than a second difference between the third eye height and the fourth eye height.

### Claim 4 (depends on 3)

The method of claim 3, further comprising: based on the first difference being greater than the second difference, adjusting the reference voltage of the slicer to the first value.

### Claim 5 (depends on 4)

The method of claim 4, wherein, based on the first difference being greater than the second difference, the adjusting of the DFE coefficient to the first setting adjusts the DFE coefficient to provide DFE that is an average of the first eye height and the second eye height.

### Claim 6 (depends on 1)

The method of claim 1, wherein the first data pattern has high-frequency switching with respect to a number of DFE taps of the DFE.

### Claim 7 (depends on 1)

The method of claim 1, wherein the second data pattern has low-frequency switching with respect to a number of DFE taps of the DFE.

### Claim 8

An integrated circuit, comprising:
a receiver circuit having a slicer and decision feedback equalization (DFE) circuitry, the DFE circuitry having a DFE coefficient, the slicer having a reference voltage; and
control circuitry configured to calibrate the DFE coefficient by:
measuring, at a first sample timing and by varying the DFE coefficient, a first eye height while receiving a first data pattern;
measuring, at the first sample timing by varying the DFE coefficient, a second eye height while receiving a second data pattern; and
based on the first eye height and the second eye height, adjusting the DFE coefficient to a first setting.

### Claim 9 (depends on 8)

The integrated circuit of claim 8, wherein the first setting provides DFE equalization that is a first average of the first eye height and the second eye height.

### Claim 10 (depends on 8)

The integrated circuit of claim 8, wherein the control circuitry is further configured to:
adjust a reference voltage of the slicer from a first value used during measurements of the first eye height and the second eye height to a second value;
measure, at the first sample timing, while using the second value, and by varying the DFE coefficient of the receiver circuit, a third eye height while receiving the first data pattern;
measure, at the first sample timing, while using the second value, and by varying the DFE coefficient, a fourth eye height while receiving the second data pattern; and
determine whether a first difference between the first eye height and the second eye height is greater than a second difference between the third eye height and the fourth eye height.

### Claim 11 (depends on 10)

The integrated circuit of claim 10, wherein the control circuitry is further configured to: based on the first difference being greater than the second difference, adjust the reference voltage of the slicer to the first value.

### Claim 12 (depends on 11)

The integrated circuit of claim 11, wherein, based on the first difference being greater than the second difference, adjusting the DFE coefficient to the first setting adjusts the DFE coefficient to provide DFE that is an average of the first eye height and the second eye height.

### Claim 13 (depends on 8)

The integrated circuit of claim 8, wherein the first data pattern has high-frequency switching with respect to a number of DFE taps of the DFE.

### Claim 14 (depends on 8)

The integrated circuit of claim 8, wherein the second data pattern has low-frequency switching with respect to a number of DFE taps of the DFE.

### Claim 15

A method of calibrating a decision feedback equalization (DFE) coefficient of a receiver, comprising:
(a) finding a first sample timing of a slicer that corresponds to a first eye center;
(b) providing a first data pattern and a second data pattern to an input of the receiver;
(c) iteratively varying a reference voltage of the slicer and sweeping a value of the DFE coefficient to measure a plurality of first eye boundaries and a plurality of second eye boundaries corresponding to a plurality of reference voltage values;
(d) based on the plurality of first eye boundaries and the plurality of second eye boundaries, determining a reference voltage value that maximizes a difference between a corresponding first eye boundary and second eye boundary;
(e) while the receiver is using the reference voltage value that maximizes the difference between the corresponding first eye boundary and second eye boundary, finding a second sample timing of the slicer that corresponds to a second eye center; and
(f) repeating steps (b) through (e) for a number of iterations.

### Claim 16 (depends on 15)

The method of claim 15, wherein the first data pattern has high-frequency switching with respect to a number of DFE taps of the DFE.

### Claim 17 (depends on 15)

The method of claim 15, wherein the second data pattern has low-frequency switching with respect to a number of DFE taps of the DFE.

### Claim 18 (depends on 15)

The method of claim 15, further comprising:
before finding the second sample timing, adjusting a value of the DFE coefficient to a first setting the DFE coefficient to provide DFE that is an average of the corresponding first eye boundary and second eye boundary.

### Claim 19 (depends on 18)

The method of claim 18, further comprising:
before finding the first sample timing, providing a third data pattern to the input of the receiver.

### Claim 20 (depends on 19)

The method of claim 19, further comprising:
before finding the first sample timing, finding a first value of the DFE coefficient to be used when finding the first sample timing.