This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

AM1808 MDIO

Other Parts Discussed in Thread: AM1808

Can the AM1808's MDIO pin work with a 1K pullup to 3.3V?  It will need to sink 3.3mA.

  • Hi Inderjit,

    From the table in section 4.4 of the Datasheet, the I/O pins can sink up to 300uA if the internal pulls are configured correctly (opposing pulldown).

     

    If you don't mind me asking, what is the purpose of sinking so much current?

  • I'm using the KSZ8051 from Micrel:

    http://www.micrel.com/_PDF/Ethernet/datasheets/ksz8051mnl-rnl.pdf

    P. 10 states that the pullup needs to be 1K, which Micrel has confirmed.  This seems to be a very low value for a pullup, and 300uA isn't going to be enough.

    Inderjit

  • It turns out that Micrel is following the IEEE 802.3 specification, which states the the MDIO pin must be capable of sinking up to 3.8mA.  The MDIO pullup in the IEEE spec is 1.5K, which is meant for several PHYs.  There are no specifications for the total capacitance on the MDIO line, the rise time requirement, or the number of PHYs supported, but the spec does say that the MDIO transceiver input capacitance shall not exceed 10pF.

    This webpage has some additional information:  http://www.interfacebus.com/MDIO_Interface_Description.html

    It says a maximum of 32 devices, and combined with the maximum input capacitance of 10pF, the total capacitance would be 320pF.  For two devices (CPU and PHY), the total capacitance is less than 32pF (10pF*2 for inputs + 5pF for traces and pads).  Keeping RC constant, a 10K pullup *should* work, if the IEEE spec really is for 32 devices, each with an input capacitance of 10pF (please verify).

    Inderjit

  • Does the AM1808 MDIO driver (davinci_mdio.c) scan all MDIO addresses all the time?  Is it possible to scan only one, if the MDIO clock (MDC) is slowed down?