A little context as to why you ask / what did you hear that makes you ask would help.
For any flavor of Ethernet on UTP (like Cat5,5e,6,6a), transmission to more than one other device (ala crossover cable) will use an active component.
In the old coax days (except for Arcnet), it was a linear bus ... with transceivers (thicknet) or no transceivers (thinnet) ... possibly with some interconnecting bridges (another active component).
There once was a passive hub used for an optical Ethernet from Codenoll, where all eight input ports (glass fibers) were bundled and fused to eight output ports (glass fibers). The power output was significantly higher than regular optical Ethernet, but there were some advantages to teh passive design (the US Navy had 'em everywhere...).
Arcnet (a token-passing, tree-based topology) used RG62 coax (same as the old IBM 3270 terminals) and totally passive hubs (in the sense that they were un-powered).
The original flavors of Token Ring used MAUs that didn't have any power of their own, but the hosts that connected to it used a series of phantom voltages to make the MAU's relays work ... so it was passive in that it didn't have it's own power, but it wasn't really, because without the voltages supplied by the hosts/clients, the Ring would not come up.
So, what do you mean RE: "Active" versus "Passive?"
Let us know
