DP83822IF: waveform of the idle 0x55 code Question

Thank you. With your support, we have tested the PMD waveform.

But there are two questions:

1) the configuration in the test is to set an [en] to 0 (that is, RX [D3] is pulled high in mode 2 or Mode3), turn off the auto negotiation function. After the auto negotiation function is turned off, although PMD waveform can be obtained, TwinCAT cannot scan to the EtherCAT slave station (et1100 is connected with dp83822).

2) we compared the waveform of Panasonic A6 servo driver (PHY chip is ksz8041nl). We would like to know if dp83822if can be configured with the same coding waveform? If so, how do we configure dp83822?

Thank your for your relpy.

I am the questioner's partner.I'd like to repeat the question for clearer expression.We are now developping a servo driver control board with EtherCAT communication, using DP83822IF as the physical layer device to connect with ET1100. The PHY connection on the schematic refers to the  TI document "Ethercat Interface for High Performance MCU Schematic (Rev.A)".  The actual connection is shown in Figure 1.

                                                                                                   Figure 1

Then we use Keysight DSOV084A oscilloscope to do the physical layer conformance test  of 100BASE-TX standard.
But we captured waveform shown in fig. 2.  This waveform cannot be tested.

The first question is this the 10BASE-Te waveform?  According to the schematic , the connection has already started the auto-negotiation function. Why it can't  automatically negotiate to 100M full duplex Mode?

Fig. 2 captured waveforms in auto-negotiation mode

According to the description on Datasheet, we configured PHY chip forced to 100BASE-TX,Full-Duplex mode, and auto-negotiation function is turned off.  The specific Strap resistor configuration is shown in Table 1.

                                                Table 1 Strap Resistor Configuration

                                                                                 Fig. 4 waveform captured in forced mode

Then the waveform captured can passed the test , but there is a problem, EtherCAT master (TwinCAT) could not scan to the slave (ET1100).
This is our second question: Why turn off auto-negotiation cause EtherCAT master can not to recognize slave?

Setting AN_EN to 0 (i.e. RX_D3 in MODE 2 or MODE3) ,turn off the auto-negotiation function and pass the physical layer consistency test, but the master cannot recognize the slave .
Setting AN_EN to 1 (i.e. RX_D3 is in MODE 1 or MODE4) ,turn on the auto-negotiation function. the master can recognize the slave ,but cannot pass the physical layer consistency test.

So,what we are looking for help is how to configure PHY to enable EtherCAT slave to be recognized normally and pass the consistency test of physical layer at the same time. 

We compared the waveform of Panasonic MBDLN25BE servo driver (PHY chip is KSZ8041NL). According to your reply, this is the 10BASE-T waveform, but it passed the consistency test of 100BASE-TX.

                                                          Figure 5 Comparison of Test Waveforms of Panasonic A6 Servo Driver