Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: Serial PMA reference clock?


I think this is an implementation issue. One implementation
may use a modified version of the XGMII TX_CLK as the PMA
reference clock. Another might choose to perform idle
insertion and deletion in the PCS to compensate for clock
differences between TX_CLK and the PMA reference clock.
Another might use a 66-bit internal serdes device and
would use a different means of converting from TX_CLK to
serial clock.

Putting something like this into the standard would imply
that one implementation is preferred over another. From
outside the device, any implementation is capable of
working and should be transparent to the test setup.


Justin Gaither wrote:
> Everyone,
>         What is the rate of the PMA reference clock?  Where does it come from?
> Is it locked to the same source as the MAC clock?
> I am assumeing the reference for the PMA is 644 or 322 Mhz, and that it
> is locked to the same source as the MAC XGMII TX clock.  So this would
> mean a frequency synthesizing PLL is required in the 10GBase-R PCS.
> Shouldnt we have specifications associated with this PLL?
> justin
> --
> Justin Gaither                       Phone: 512-306-7292  x529
> RocketChips a Division of Xilinx     Fax:   512-306-7293
> 500 N. Capital of TX Hwy.
> Bldg 3                         email: jgaither@xxxxxxxxxxxxxxx
> Austin, TX 78746               WWW:

Benjamin Brown
2 Commerce Park West
Suite 104 
Bedford NH 03110
603-641-9837 - Work
603-491-0296 - Cell
603-626-7455 - Fax
603-798-4115 - Home Office