Mercurial switching [Microwave Bytes]

On an interest scale of one to ten, I suspect electrical relays would barely score a one for most radio frequency (RF) designers. The relay in electrical language is an electromagnetically controlled switch and, as such, has little direct use for the average microwave engineer, and, indeed, some may not even be familiar with the term. But it turns out that I have had some experiences with certain types of relays that did actually once get me quite excited. Basically, a relay is a switch that can easily handle many amps and hundreds of volts, and that can be actuated by a control current that flows in an electromagnet. The control current is thus completely isolated from the circuit containing the main switch. Until maybe the late 1960s, most telephone exchanges were almost totally dependent on this device, along with a more extroverted cousin, the uniselector. But trouble lay ahead in the form of the transistor. Someone had the crazy notion that a transistor can be made to perform the same function as a relay, inasmuch as the current flowing between two of its three terminals could be used to control a much larger current flowing between a different pairing of terminals. I will not attempt to trace the ensuing history of electronic telephone exchanges, since I can´t say I know much about it and it´s probably rather boring. The main user benefit of replacing the electromechanical exchanges was faster call routing, but there were plenty of downsides that took quite a long time to sort out.