A reed relay is a type of relay that uses an electromagnet to control one or more . The contacts are of magnetic material and the electromagnet acts directly on them without requiring an armature to move them. Sealed in a long, narrow glass tube, the contacts are protected from corrosion. The glass envelope may contain multiple reed switches or multiple reed switches can be inserted into a single bobbin and actuate simultaneously. Reed switches have been manufactured since the 1930s.
Compared with armature-based relays, reed relays can switch much faster, as the moving parts are small and lightweight, although switch bounce is still present. Also, they require less operating power and have lower contact capacitance. Their current handling capacity is limited but, with appropriate contact materials, they are suitable for "dry" switching applications. They are mechanically simple, making for reliability and long life.
Such an electrically latching reed relay requires continuous power to maintain state, unlike magnetically , such as ferreed (ferrite and reed relay) or the later remreed (remanent reed relay).
Each input of the array had, besides the two talk wires, a P lead for controlling the crosspoints on that level. Two coils on each crosspoint were wired in series with all the others on that level, to the P lead. Each output of the array also had a P lead with two coils on each crosspoint of that output level. The two windings controlled by the same level were unequal, and were wound around opposite ends of the reed, in opposing polarity. When a pulse passed through the crosspoints of a level, the two ends of each reed were magnetized north to north or south to south, thus repelled each other and opened the crosspoint in all except the selected crosspoint.
The selected crosspoint had current passing through both its input P lead and its output P lead, thus through all four windings. On each end of the ferreed, the windings provided by the two different P leads were opposed to each other, and the greater one predominated when both were energized. This being the input P lead at one end of the ferreed, and the output P lead at the other end, the two ends of that particular ferreed were magnetized north to south, hence attracted each other and closed the contact. Current was applied by the pulser only to set up the connection. The P leads remained dry and the crosspoint remained closed until such time as another connection was made involving one of the levels.
Because the individual crosspoints were more expensive than those of , while the control circuitry was cheaper, reed arrays usually had fewer crosspoints and were more numerous. This required them to be arranged in more stages. Thus, while a telephone call in a typical crossbar exchange like 5XB passed through four switches, a call in a reed system such as 1ESS typically passed through eight.
In the later 1AESS, the reeds were of remanent magnetic material. This "Remreed" design allowed further reduction in size and power consumption. A "grid" of 1024 2-wire crosspoints, arranged as two stages of eight 8×8 switches, was permanently packaged in a box. Despite the sealed contacts, plating with silver rather than with precious metals resulted in reed arrays being less reliable than crossbar switches. When one crosspoint failed, the grid box was quickly replaced as a unit, and either repaired at a local workbench or shipped to a repair shop.
Stromberg-Carlson made the similar ESC system, whose reeds were called crossreed.
Reed relays were extensively used in the British TXE family of telephone exchanges.
|
|