Another thought on 10baseT vs 10Base2: was Re: Microvaxen bits etc.

I just had to stick in my 2.1 cents worth (inflation, you know...) on this issue. I will *never again* use 10base2 for plant wiring for the simple reason that 10base2 (and 10base5) provide a ready path for "ground loops" between all the equipment sharing the cable. It's true that the "grounds" are "supposed" to be isolated (to the tune of several Kvolts) from the computer connection, but, believe me: "sh*t happens." And the results can be dangerous. 10baseT is designed to be isolated at every connection point in the "star" configuration (high-isolation transformer coupling,) so there is a much less likely path for the "big ugly voltage" from one piece of defective equipment that just happens to be shorted to the Ethernet 10base2 cable. You don't get killed when you reach behind your computer and grab the Ethernet cable and suddenly find it has 480vac on it due to some stupid short in a control cabinet. If the 10baseT cable is shorted, the hub has a high potential, but it never makes it out to any of the other cables plugged into the hub. Both of these scenarios assume correctly isolated cable-to-interface connections. It's just there is nothing isolating the bad stuff from you when you touch the cable shield. The necessary hub on 10baseT keeps the bad stuff from getting past the first short - assuming you *do* use a hub... And making 10baseT connections (with the proper 50US$ tool) is MUCH FASTER than making coax connections, if speed is important. Of course, the better (best?) approach is 10baseF, but it isn't readily available for us with home installations (at least not at the price *I* want to pay.) Gary