Repeaters

A common problem in the networking world is that of weakening electical signals.  Electrical signals traveling through wires (such as copper wires used in most networks), weaken due to the wire's electrical resistance.  This effect limits the lengths of the cable that can be used.  A repeater will overcome this limit, when there is a need to connect two computers at a larger distance.
A repeater is connected to two cable segments.  Any electrical signal reaching the repeater from one segment, will be amplified and retransmitted to the other segment.

The number of repeaters that can be chained together is limited.  This is because most network types assume a maximum segment length and propogation time.  For instance, each Ethernet network type imposes a maximum time limit for each electrical signal to reach the entire network.  Using repeaters slows the signal's propogation, and thus the number of repeaters should be limited.

The upside of repeaters is that they are very simple to manufacture and therefore cheap.


The downside of repeaters, which are dumb electrical devices, is that they do not "understand" the nature of the electrical signals they are amplifying.  A repeater may encounter 3 types of electrical input signals:

1. Proparly formatted data that needs to be be repeated to the repeater's other segment. (i.e. when the data's target is on the repeater's other segment)
2. Proparly formatted data that does not need to be repeated to the repeater's other segment. (i.e. when the data's target is not on the repeater's other segment)
3. Improparly formatted data or noise that, of course, should not be repeated.

The graphics and text (with some minor editing) on this page were written by my former student in 1997 at Tel Aviv University.