2011-01-19 11:47 PM
RMII or MII?
2011-05-17 05:22 AM
Faster clock, less pins?
2011-05-17 05:22 AM
2011-05-17 05:22 AM
There is no difference in functionality. I decided to use RMII because of less signals to route (restricted board space). Minor drawback is the higher clock of 50MHz, instead of 25MHz for MII. The only thing you have to take care about is that the PHY is correctly configured. And if you use an OS you should make sure that the driver can support RMII.
BTW: Nice to hear that the F207s are already sampling. Or are you just preparing a layout for using them?2011-05-17 05:22 AM
already in production.
thanks2011-05-17 05:22 AM
2011-05-17 05:22 AM
I never heared of a 25Mhz RMII clock mode. The only thing I know is that it is possible with some PHYs (LAN8720A) to use it in RMII mode with an external 25Mhz crystal and the PHY will generate the 50MHz clock. This clock then must be routed back to the MAC.
But I don't see how it is possible to transfer 100MBit via 2 data lines with a clock less than 50MHz. At least if the RMII standard is used. Anyway, routing the 50Mhz is not the big deal, simplest solution is to keep the PCB traces short. In my case I even did not use termination resistors (max trace length <7cm).2011-05-17 05:22 AM
2011-05-17 05:22 AM
Well it does say than it can generate the 50 MHz, and the part does contain a PLL, so it could pretty much synthesis anything it wants.
2011-05-17 05:22 AM
The ''25Mhz RMII clock mode'' is exactly what I said in my previous post. The LAN8720A datasheet calls this ''REF_CLK Out mode''. And at the same time says, that this mode is not fully compliant to RMII spec and timing specs have to be checked carefully to make sure it will work with the MAC. If you take a look at Page 51 you will see minor differences in timing for the Micrel KSZ8051, too.
Therefore I really would not recommend this mode for use with the STM32.