Q & A on Ethernet and Transmission Errors Q: > Yes, attenuation is a minor design issue -- for a given transmitter power > level and type of cable, the signal will only reach a certain > distance before becoming too weak to receive reliably. However, if > attenuation were a major limitation issue, it could be overcome by > more stringent cable requirements or increased transmitter power levels. True. But you can seldom reduce the probability of bit errors (which depends on the so called signal-to-noise ratio) by increasing the signal power level. The reason is that often many wires (e.g. twisted pairs) are collected very close to each other for long distances in the same cable. This causes "cross-talk" between the wires due to cross-induction. If the cable is long, the cross-talk is the main noise source. If everyone increases their power level with, say, a factor 10, then the noise level will also be increased by a factor 10 (if the cable is long). Thus, the signal/noise qoutient is unchanged, and the bit error probability is the same. A: Actually, when I wrote the answer I was thinking of the distance limitation on the original 10Base-5 coaxial cable. You are correct that on a 10Base-T cable, crosstalk will be a problem at higher power lavels.