sargon wrote:
On Friday, 4-November-2005 12:36, pelibali wrote:
Hi,
I found several unofficial values, how long can be the longest RJ45 network cable. Few sites listed about 12 meters, but I saw 20 meters and also 300ft as well; later being in fact around 90 meters
End-to-end distance from an Ethernet transmitter to an Ethernet receiver cannot exceed 100 meters (328 feet). This is due to the timing of the electrical signal on the cables.
The cable lenth has never been limited by timing. It is limited by signal degradation, which is why 10base5, 10base2 and 10baseT all have different maximum cable lengths. If timing were the issue, they'd all have similar maximum lengths, as they have similar velocity factors. If you examine the details and the reason for the timing issues, you'll realize that the maximum network distance is much greater. The timing issue is for collision detection. The ethernet standard required collision detection within 512 bit times. At 10 Mb, that equals 51.2 uS (round to 50, for convenience). So, in that 50 uS, a signal has to be able to make a round trip to the furthest point on the network, which means that the furthes point can only be 25 uS away. At the speed of light 300,000 KM/s or 300M/uS, that point is 7.5 KM away. In cables, the velocity factor is in the vicinity of 70% (light in a vacuum has a velocity factor of 100%), so the maximum distance would be around 5 KM.. That distance is far greater than can be reached with any copper cable used for ethernet. It can be reached with fibre. Also, with switched hubs and full duplex NICs, where there is no possibility of a collision, there is no maximum (realistic) network distance, though you still have to worry about signal degradation, which limits you to 100 M in twisted pair cable.