[En-Nut-Discussion] Problem with ETH PHY Lan8742

[Uwe Bonnes]

>>>>> "Ole" == Ole Reinhardt <ole.reinhardt at embedded-it.de> writes:

    Ole> ...
    Ole> You mean that on the NUCLEO-F746ZG, the Phy works after power up,
    Ole> but fails after a reset and on the other Nucleo Board it often
    Ole> fails?

No, on the Nucleo144 board, the captured strap values where often wrong. And
on th F767 additional even if the right value was captured, DHCP often did
not succeed, while on the F746 it always succeded with the right values
captured.

    Ole> I do not know these boards, nor their schematic. But pon the LPC
    Ole> boards, that I designed, I connected the reset line of the PHY to a
    Ole> GPIO pin of the CPU. This allowes me, to configure the MII lines as
    Ole> GPIO with pullups set to the right strap values and the reset the
    Ole> phy manually.  This code is added to the board startup code, so
    Ole> that the MII interface is later configured automatically correctly
    Ole> when the ethernet device is initialised.

Luckily on the Nucleo144 boards, system reset NRST is connected to the Phy
reset via a 0 Ohm jumper. Even only removing the jumper and pulling up Phy
reset let the Nucleo Boards always start with the right strap values. The
problem with the F767 not succeedibg DHCP remains, in a test with 30 System
reset only in about half of the starts DHCP succeeded.

The negotiation mode values for the Lan8742 should be settable beside by the
strap values also with register settings. I will try that to save users from
the need to modify the board.
    Ole> ...

Bye
-- 
Uwe Bonnes                bon at elektron.ikp.physik.tu-darmstadt.de

Institut fuer Kernphysik  Schlossgartenstrasse 9  64289 Darmstadt
--------- Tel. 06151 1623569 ------- Fax. 06151 1623305 ---------