Hello , I am currently working on S32K3X8EVB-Q289 for ethernet implementation.
I was going through the hardware user manual and came across the section "rework for rmii"
As per this table , i made the necessary changes in the board.
However , after making the changes lwip baremetal example was not working . Ping test was not working as well. Moreover , I have another eval board without making the above mentioned hardware changes. That one works fine .
What could be the reason that the first board with hardware changes fails.? Why is this rework for rmii needed if lwip example can work even without these changes ?
Hi @PavelL
You mentioned that I have to change the GMAC pin mapping if I am reworking the eval board for rmii.
It was XMII initially. Should I change it to XGMII ? Is that what you mean? If not what should I change exactly?
Hello @ssattva1 ,
I double checked pin mapping of example lwip_baremetal_s32k358 together with schematics of S32K3X8EVB-Q289 and ADTJA1101.
In my opinion:
- No SW changes are necessary - RX_ER is not assigned in Gmac to pins, RX_DV (PTC16) and TX_EN (PTE9) are assigned to pins
- HW changes shall look like that, i.e. the highlights are valid:
Please check it out.
Best regards,
Pavel
Hi @PavelL ,
So this is the pin mapping that i have currently , the default mapping that comes along with the example.
I am flashing this configuration in a board with no rework for rmii . I probe the txctl and tx_en signals , which are at PTB25 and PTE9 respectively. I observed that TXCTL goes high whenever i transmit a frame , but when i observed tx_en from PTE9 I obsrved this
A periodic pulse of period 80ns .
What I infer is , tx_ctl essentially acts as the tx enable signal. Also in schematic Tx_ctl line is populated and tx_en from PTE9 is DNP. Then what is this periodic signal from PTE9 that I am observing.
