Basic Ethernet principles
‘Ethernet’ is used at present in different performance classes, depending on the data transfer rate: 10, 100, 1000 Mbit/s. The 100Mbit FastEthernet exclusively dealt with here as a physical transmission method according to ISO/IEC 8802-3 is known as
- type 100Base-TX
- full duplex, hence collision avoidance according to CSMA/CD is not necessary
- Use of 2 of the 4 possible core pairs: 1-2 and 3-6. A four-core cable is therefore sufficient.
- point-to-point connection between 2 intelligent devices, which dynamically negotiate parts of the connection establishment via the connection ICs
The 100 Mbit/s usable data stream is triple-encoded
- 4-bit/5-bit encoding (ISO9314, for clock recovery) ® 125 Mbit/s gross data stream
- NRZI coding (for frequency reduction, a level change means 1bin) ® max. 62.5 MHz “intermediate frequency”
- MLT-3 coding (for frequency reduction, 3 instead of 2 voltage states) ® max. 31.25 MHz signal frequency on the cable; the actual frequency depends on the data stream and is thus variable
Taking into account the harmonics generated, a connection performance of the total section as per EN50173-1 Class D/Cat5 for signals up to 100 MHz is thus sufficient for FastEthernet (EtherCAT).
Conversely, 1000 Mbit/1 Gbit Ethernet works with a middle signal frequency of 62.25 MHz and needs all 4 core pairs. In principle, a fully assigned Class D section is suitable for transmission; however, since all four core pairs are used in Gbit Ethernet, and these are also used bidirectionally at the same time, it is recommended to obtain section certification in accordance with the tightened limit values (crosstalk, return loss) as per ANSI/TIA/EIA-TSB-6 (TIA Cat. 5e).
Establishment of a connection
The simplest way of diagnosing an Ethernet connection is to observe the link display at both end points: if an Ethernet cable is connected at each end to a device, both terminals begin to synchronize themselves or to maintain synchronization by the continuous transmission/reception of a special bit sequence (the idle symbol). This idle symbol consists of the maximum possible number of level changes, since the ‘1’ is transmitted 5 times – the transmission of a ‘1’ means a level change in the NRZI process.
Hence, due to the constant exchange of idle symbols, an Ethernet device that is not operative has a higher current consumption than later during normal data traffic!