I'm confused about definitions of PHY and Ethernet transeiver. They ain't the same, as I assume. Transeiver as I understand, is rather a sort of converter from one media to another, for example 100base-tx 1000base-fx. On the other hand, both PHY and transeiver deal with a physical layer, then what exactly makes them different?
| Transceiver simply mean "transmitter-receiver." A transceiver operates | in the lower half of the PHY layer, called PMD (physical medium | dependent). It's the device that connects your box to the physical LAN | cable, if you will.
In other words, transceiver is a part of PHY layer? Is it valid (as per IEEE standard) to separate transceiver into a distinct product, what is the rationale for this? For instance (not to be considered as advertisement),
According to its description it's rather a PHY chip, they call it transceiver.
Transceiver simply mean "transmitter-receiver." A transceiver operates in the lower half of the PHY layer, called PMD (physical medium dependent). It's the device that connects your box to the physical LAN cable, if you will.
What you describe to connect 100BASE-TX to 1000BASE-FX, for example, is a product usually called Ethernet bridge, or layer 2 switch. As the name implies, layer 2 switches operate above the PHY layer, and in principle, leaving aside the predictable grumbling and flaming, they can also tie together different flavors of IEEE 802 LANs.
Ethernet has been around for decades now and the technology and terminology has changed.
A transceiver was originally used between the AUI interface and the cable - co-axial (two flavours) or later UTP and fibre. Later the term became used for media convertors between different media eg UTP and fibre (at the same data rate). This term is still in use today.
PHY is a later term for the Physical Medium Dependent layer of the connection. It connected originally to the Mii (I think it was) but now the Mii is never presented outside the equipment.
There is also the question of IEEE "standard" terms and commonly used terms. I will leave such distinctions to Mr. Seifert or of course you can read about it yourself. A lot of the 802.3 standards are (were?) free from the IEEE website and of course you can buy any and all of them:-)
You have chosen a *very* bad example here since you are changing media *and* speed. Typo I guess?
100bse-tx to 100base-fx converter could well be called a transciever.
You need a Bridge (includes switch) or Router to go between different speeds.
Maybe someone knows if "Switch" is in the 802.11d standard now or if it still retains the original (and of course best) "Bridge".
Hmmm - 802.11 to 802.3 surely needs no flames. Now 802.3 to token ringy thingy could be different:-).
If you look at Figure 1-1 of IEEE 802.3-2005, you'll see graphically what they have done over the years. The transceiver is indeed a device that operates in the lower half of the PHY layer, and although not mandatory, the device can be separated physically from the rest of the PHY layer.
So what used to be the MAU in the old days is pretty much exactly replicated in the newest Ethernets, although at the higher speeds, it can no longer be located external to the Ethernet NIC.
So to me, the Broadcom description is valid. They are selling the functional equivalent of a GigE MAU, and oh by the way, it must be installed in the NIC itself.
The IEEE 802 standards available for free online are accessed here: