Method of CTLE estimation using channel step-response for transmitter link equalization test

Patent metadata

Field	Value
Publication number	US-12339298-B2
Application number	US-202218059297-A
Country	US
Kind code	B2
Filing date	Nov 28, 2022
Priority date	Dec 29, 2021
Publication date	Jun 24, 2025
Grant date	Jun 24, 2025

Abstract

A method of automatically selecting a continuous time linear equalization (CTLE) filter includes capturing a response waveform for a channel of a communication link of a device under test (DUT), generating a set of candidate CTLEs, and automatically selecting the CTLE from the set of candidate CTLEs using the response waveform. A test and measurement instrument has a user interface, a port to allow the instrument to connect to a device under test (DUT), and one or more processors configured to execute code to cause the one or more processors to: generate a set of CTLE candidates; capture a response waveform for the channel; and automatically select the CTLE from the set of candidate CTLEs using the response waveform.

First claim

We claim:

1. A method of automatically selecting a continuous time linear equalization (CTLE) filter for a communication link equalization, comprising:
capturing a signal transmitted over the communication link;
receiving, through a user interface, a user-defined granularity for direct current (DC) gain steps for one or more analog-to-digital converters (ADC);
generating a set of candidate CTLEs by varying the direct current gain across a range of DC values as the user-defined granularity for the one or more ADCs;
using the one or more ADCs to generate a response waveform for a channel in a communications link for each of the candidate CTLEs;
automatically selecting the CTLE from the set of candidate CTLEs using the response waveform; and
applying the CTLE to protocol traffic to equalize the channel on the communication link to improve traffic on the channel.

2. The method as claimed in claim 1 , wherein automatically selecting the CTLE comprises:
determining a time-domain impulse response for each candidate CTLE;
computing a channel impulse response from the response waveform;
using the channel impulse response with the time-domain impulse response of each candidate CTLE to produce an equalized impulse response for each candidate CTLE;
generating a set of step waveforms and frequency responses from the equalized impulse responses for each candidate CTLE;
reducing a number of the candidate CTLEs in the set; and
selecting the CTLE that has steady state values most closely aligned with the original channel response.

3. The method as claimed in claim 2 , wherein determining the time-domain impulse response for each candidate CTLE comprises applying an inverse Fast Fourier Transform to each candidate CTLE.

4. The method as claimed in claim 2 , wherein using the channel impulse response and the impulse response of each candidate CTLE comprises convolving the channel impulse response with the time-domain impulse response for each candidate CTLE.

5. The method as claimed in claim 2 , wherein generating the set of step waveforms from the equalized impulse responses comprises re-creating the step waveform for each candidate CTLE from the equalized impulse response for that candidate CTLE.

6. The method as claimed in claim 2 , wherein reducing the number of candidate CTLEs in the set comprises:
removing candidate CTLEs from the set that cause over-equalization; and
keeping only the candidate CTLEs from the candidate CTLEs remaining in the set that have controlled undershoot behavior.

7. The method as claimed in claim 2 , wherein selecting the CTLE that has steady state values most closely aligned with the original channel response comprises:
finding a steady-state value for the frequency responses for each candidate CTLE;
determining a maximum steady-state value from the steady-state values of the candidate CTLEs;
determining a normalized difference between the maximum steady-state value and the steady-state value for each candidate CTLE; and
selecting the candidate CTLE with the minimum analog-to-digital converter gain.

8. The method as claimed in claim 1 , wherein capturing the response waveform includes acquiring a step response waveform of the channel and normalizing the step response waveform by selecting only samples from the step response waveform for analysis that capture transient and steady state portions of the step response waveform, and normalizing the y-axis to fall within the range [0, 1].

9. The method as claimed in claim 1 , further comprising:
starting a transmitter communications link equalization test for a device under test (DUT); and
initiating link training for the DUT.

10. A test and measurement instrument, comprising:
a user interface;
a port to allow the instrument to connect to a device generating a signal on a communication link;
one or more analog-to-digital converters (ADC) to receive the signal and convert the signal to a digitized waveform; and
one or more processors configured to execute code to cause the one or more processors to:
receive, through the user interface, a user-defined granularity for a direct current (DC) gain;
generate a set of continuous time linear equalization filter (CTLE) candidates by varying the DC gain across a range of DC values at the user-defined granularity for the one or more ADCs;
capture a response waveform for the channel;
automatically select the CTLE from the set of candidate CTLEs using the response waveform for the channel;
apply the CTLE to protocol traffic on the communication link to equalize the channel and improve the protocol traffic.

11. The instrument as claimed in claim 10 , wherein the code to cause the one or more processors to automatically select the CTLE from the set of candidate CTLEs causes one or more processors to:
determine a time-domain impulse response for each candidate CTLE;
compute a channel impulse response from the response waveform of the channel;
use the channel impulse response with the time-domain impulse response of each candidate CTLE to produce an equalized impulse response for each candidate CTLE;
generate a set of step waveforms and frequency responses from the equalized impulse responses for each candidate CTLE;
reduce a number of the candidate CTLEs in the set; and
select the candidate CTLE that has steady state values most closely aligned with the channel response as the selected CTLE.

12. The instrument as claimed in claim 11 , wherein the code to cause the one or more processors to determine the time domain impulse response for each candidate CTLE comprises code to cause the one or more processors to apply an inverse Fast Fourier Transform to each candidate CTLE.

13. The instrument as claimed in claim 11 , wherein the code to cause the one or more processors to use the channel impulse response and the impulse response of each candidate CTLE comprises code to cause the one or more processors to convolve the channel impulse response with the time-domain impulse response for each candidate CTLE.

14. The instrument as claimed in claim 11 , wherein the code to cause the one or more processors to generate the set of step waveforms from the equalized impulse responses comprises code to cause the one or more processors to re-create the step waveform for each candidate CTLE from the equalized impulse response for that candidate CTLE.

15. The instrument as claimed in claim 11 , wherein the code to cause the one or more processors to generate the frequency response for each candidate CTLE comprises code to cause the one or more processors to apply a Fast Fourier Transform to the equalized impulse response for each candidate CTLE.

16. The instrument as claimed in claim 11 , wherein the code to cause the one or more processors to reduce a number of the candidate CTLEs in the set comprises code to cause the one or more processors to:
remove candidate CTLEs in the set that cause over-equalization; and
keep only the candidate CTLEs that have controlled undershoot behavior.

17. The instrument as claimed in claim 10 , wherein the code to cause the one or more processors to capture the response waveform of the channel includes code to cause the one or more processors to capture a step response waveform of the channel and normalize the step response waveform of the channel.

18. The instrument as claimed in claim 10 , wherein the code to cause the one or more processors to select the candidate CTLE that has steady state values most closely aligned with the channel response comprises code to cause the one or more processors to:

Assignees

• Tektronix Inc

Inventors

• Ghose Subhankar
• Dash Ankit
• Bouse David M
• Joshi Pradeep

Classifications

• G01R19/2509

  Details concerning sampling, digitizing or waveform capturing · CPC title

• G01R13/34 Primary

  Circuits for representing a single waveform by sampling, e.g. for very high frequencies · CPC title

• H04L25/03885 Primary

  adaptive · CPC title