RMII with DP83848

Can you help me understand the setup?  Are you saying that you used the ST MB786 as a reference design for your own board and you are beginning bring up checks of your new board?  

The DP83848C wiring shown in your schematic is generally correct, but I do have some questions / recommendations:

1. Your schematic shows a PAD_THERMAL pin on the DP83848C that is grounded, but the DP83848C does not have a ground pad.  What is this connection in the layout?
2. Your schematic also shows that the TXD_3 pin is shorted to ground.  Please sever this connection.  Functional pins should not be directly connected to power or ground.  

I do not know why the board would reset based on a cable connection.  I would suggest two options for evaluating this condition:

1. Check your power supplies, starting with the 3.3V supply of the PHY.  When the PHY links, it will begin to draw more power.  If your power supply is marginal, the additional power consumption could be causing it to droop.
2. Evaluate the impact of the PHY interrupt on the ST MCU.  If the PHY issues an interrupt when it links, how does the MCU respond?

Regards,

Patrick

Thank you for your answer.

Yes that's it, I'm inspiring myself of the demo board to make my own card.

Concerning the thermal pad, in the datasheet (page 16) the description of thermal pad is called GNDPAD, that's the reason why it's grounded. Do you think I need to remove this connection ?

Concerning TXD_3, I need to put a 2k2 pull down resistor is that what you mean ?

I will check the two points that you mentionned. Thank you.

best regards,

Fabrice

It seems that the problem comes from my voltage regulator. I saw at the oscilloscope that the 3V3 voltage goes down to 1V that can explain the reset.
I have checked by putting another 3V3 supply and it seems that now, I have not any reset, so it can be a start of explanation.

So now, I follow this document (www.ti.com/.../snla076a.pdf) which explain how the pin necessary for RMII communcation must be connected.
So, I have add 2k2 resistor for RX_DV and TX2/TX3. RX_DV is connected to VDD by a 2k2 resistor, but when I measure the voltage, it's 0V.

On the topic of the thermal pad, the datasheet (Literature Number: SNLS266) covers four related part numbers. The DP83848VYB and the DP83848YB have a higher temperature rating and are packaged in an HLQFP package with a thermal pad. The DP83848C and the DP83848I have a lower temperature rating and are packaged in an LQFP package without a thermal pad.

The 2k2 resistor strap on RX_DV will be sampled by the device during reset (during power up or when the RESET_N pin is asserted low). During reset, RX_DV will be configured as an input. After reset, RX_DV will be configured as an output. If you assert the RESET_N pin and measure the voltage on RX_DV, you should see it pull up.

Patrick

Thank you for your reply. My 3V3 voltage is supplied by a TPS5401. I have used the webench to configure it. With this configuration, as I mentionned before, my card was resetting.

Now, I have just changed the Cout (47µF -> 1000 µF), and my board is not resetting anymore. Moreover, I have measured the current provided by the regulator and I'm around 200mA (the regulator can provide up to 500mA), so I don't really understand why my board reset. Moreover, the TPS5401 is really hot, around 60°C (I have used an IR thermometer to check it), and I don't know if it's normal or not ?

Thank you in advance,

The DP83848 can typically draw 80-90mA depending on the mode of operation.  With an MCU and other components, a total current consumption of 200mA sounds reasonable.

For your questions on the TPS5401, I would suggest that you post a question in the  Non-Isolated DC/DC Forum:

https://e2e.ti.com/support/power_management/non-isolated_dcdc/

Aside from your questions on the regulator, do you have any remaining open questions on the DP83848 functionality?  Are you now able to link with another device? 

Patrick

For the moment, I cannot link it with another device. I try just to make a ping on it but without success. I'm following my investigations.

Fabrice,

That is good news. I will mark this as closed. If any new questions arise during your evaluation, you can open a new post and we will address them there.

Patrick