| ??? 10/31/09 21:40 Read: times |
#170294 - to PHY or not to PHY Responding to: ???'s previous message |
Richard Erlacher said:
The typical "MII" (media independent interface) to the typical PHY uses about 26 pins. If the PHY is on board, it uses just the transmit/receive lines that actually go to the RJ45, presumably with integrated magnetics. That saves board area, and lots of hassle! Sure it increases the value of the chip a lot. I've had a quick glance at the datasheet, and the PHY is definitively on-chip. However, at one point, the datasheet hints about an "external PHY mode", but there's nothing more on it. I suspect one of the deliberately undocumented TEST pins might make the chip to bypass the internal PHY, outputting MII onto the general-purpose IO pins perhaps. Not that it is very useful, most probably used in factory testing only. "Richard" said:
Of course, if you damage the PHY, the whole MCU is in the toilet, but "nobody" attempts to repair boards at the component level any longer. Well, a separte PHY would be most probably an SMD part as well, so it's not any simpler to replace that one than the whole MCU - and the price of the replacement W7100 ($6-7) is not that bad. But, given the isolation provided by an average ethernet transformer, I would be surprised if a damaged PHY would not mean a major damage to the rest of the circuitry, too... JW |
| Topic | Author | Date |
| Fancy a '51 with TCP/IP? | 01/01/70 00:00 | |
| Dead ? | 01/01/70 00:00 | |
| Shame about the support? | 01/01/70 00:00 | |
| waiting | 01/01/70 00:00 | |
| DIY? | 01/01/70 00:00 | |
| it's still more | 01/01/70 00:00 | |
| Cortex-M3 comparison | 01/01/70 00:00 | |
| PHY on-chip | 01/01/70 00:00 | |
| it saves I/O's! | 01/01/70 00:00 | |
| RMII | 01/01/70 00:00 | |
| to PHY or not to PHY | 01/01/70 00:00 | |
| It's not that easy to damage the PHY | 01/01/70 00:00 | |
| Reason for off-chip PHY | 01/01/70 00:00 | |
| different PHY? | 01/01/70 00:00 | |
| Wisdom of youth | 01/01/70 00:00 | |
Wiznet | 01/01/70 00:00 |