SLLA613 august   2023 TUSB1146

 

  1.   1
  2.   Abstract
  3.   Trademarks
  4. 1Introduction
  5. 2Test Description
    1. 2.1 Equalization Configuration
    2. 2.2 Eye Test
    3. 2.3 BER Test
    4. 2.4 Alternative BER Test
  6. 3Test Results
  7. 4Summary
  8. 5References

2 Test Description

The following tests demonstrate that AEQ performs equivalently, or better, than static EQ across various channel lengths and jitter levels. The result of changing from EQ to AEQ is apparent when selecting EQ values without performing jitter tolerance (JTOL) testing and tuning. This test measures Bit Error Ratio (BER) and eye diagram parameters against a set of channel board lengths for each EQ configuration. The test is first conducted with clean signal, followed by a JTOL test.

There are four test configurations:

  1. No Redriver - No redriver included in the signal path. This test provides a baseline.

  2. Static EQ - This test provides a preset constant EQ value, regardless of the connected channel characteristics. The preset can be changed by overriding the default EQ value and setting it manually.

  3. Fast AEQ - Fast AEQ must be enabled in the registers. Upon cable connection, this configuration determines if the connected channel is considered long or short and chooses the preset EQ setting for either case, as determined by the TUSB1146 algorithm. This preset can be changed in the registers.

  4. Full AEQ - Full AEQ must be enabled in the registers. Upon cable connection, this configuration checks every EQ value and chooses the optimal setting for the channel characteristics, as determined by the TUSB1146 algorithm. The Full AEQ algorithm takes longer to run than Fast AEQ.

To configure the EQ through registers, the TUSB1146 must be in the I²C mode. Please see the register map on the TUSB1146 data sheet for details about register functions and locations, as well as details about Fast versus Full AEQ. Use the TUSB1146 EVM user's guide for physical setup assistance.

Note:

The tests and test results in this application note can not be used in place of official compliance results and part capabilities. These tests are conducted solely to compare both methods of equalization, not measure their capabilities in application. Attempting to reproduce these results in test or application can yield different outcomes depending on quality or length of cables used, quality of the receiver measuring the signal, and other similar variables.

Figure 2-1 and Figure 2-2 show that the tests in this application note use a TUSB1146 breakout board allowing for direct connection to TUSB1146 pins. If this apparatus is unavailable, this document provides an alternative test that utilizes the TUSB1146EVM, producing JTOL results for comparing EQ configurations. This test description can be found in Alternative Test Procedure.