Skip to main content

Power over ethernet

More
17 years 11 months ago #18820 by TheBishop
Replied by TheBishop on topic Re: Power over ethernet
The information I have relates to traditional Ethernet over coaxial cable but the principles should still be the same. Here goes:
[code:1]
"The transcievers act as current generators in transmission and as current detectors in reception. Thus [data] signals are propogated on the cable by applying a current. The actual signal placed onto the cable is composed of an a.c. component (the data) and a d.c. offset (the carrrier).
The a.c. component 'carries' the bit pattern forming the transmission. This signal will vary about the steady-state or idle value by +/- 28mA.
The d.c. element of the transmitted signal allows the transciever to detect whether it is receiving a valid transmission or a collision fragment. Only one device will be transmitting if the signal is valid, when the d.c. element present in the signal will be from 37mA to 45mA. A collision occurs when two or more devices are transmitting. Since d.c values add, the total d.c. element present in a collision signal will be greater than 74mA. Thus, whenever the d.c. element in the signal is greater than a threshold value which results in a voltage level on the cable greater than 1.492V to 1.629V (for thin Ethernet the values are 1.404V to 1.581V) a collision must be occuring. The transciever therefore monitors the d.c. voltage component of any received signal. When this exceeds the threshold, the transciever raises the SQE (Signal Quality Error) signal on the AUI cable. This signal is received by the PLS sub-layer which in turn generates a PLS_SIGNAL.indicate (SIGNAL_ERROR) to alert the MAC layer that a collision has been detected"[/code:1]
How's that?
Time to create page: 0.121 seconds