Once a relatively small market the requirement for low cost, high specification RF switching relays is now one of the largest expanding Global markets. A manufacturer of commercial relays twenty years ago would have offered just the occasional ¡¥military¡¦ based RF switching relay usually designed to a MIL 5757 specification that, in reality, was too expensive to consider for use in the majority of high frequency applications. Alternatively, the ubiquitous coax cable based relay that typically matched the size of a large transformer was a possibility ¡V of course providing the equipment was of a suitable size. Unfortunately, this was not an incentive for designers to miniaturise circuits.
However, the necessity to utilise ever higher frequencies from the 1GHz cellular telephone systems through to frequencies above 30GHz has forced relay manufacturers to design and produce relays capable of efficient switching at previously unconsidered frequencies. As such manufacturers have researched and developed a wide range of both commercial and specialised switching relays and modules.
A good example of this change is the set top digital TV box ( STB ) market where RF switching relays are utilised in distribution, amplification and modem networks - an application that did not exist 20 years ago - but now requires some millions of relays per year world-wide. With the globalisation of interactive digital TV this is a rapidly expanding market.
To meet the growing demands for high frequency switching relays at a commercial cost the leading manufactures, such as Panasonic, have invested in an ongoing research program aimed at extending the ranges of relay product offered. In addition to this expansion a total new area of individual and multiplexed units have been developed to allow modular design of multi-channel switching banks for both test and redundancy use.
The design of an RF switching relay requires inherent features not usually associated with standard electromechanical devices. In earlier designs, to avoid excessive cost, non-shielded designs were utilised. These consist of a basic relay design very similar to standard relays however consideration is given to selecting the most suitable contact material, designing the contact faces and associated areas to offer the lowest capacitance value and utilising plastics and moulding outlines that allow the best options for avoiding leakage across secondary conduction paths.
The above methods are still suitable for frequencies upto 500 MHz. However above this frequency the losses via the unshielded conduction areas become too large. The shielded relay differs from the standard relay construction by encapsulating the contact areas with a metal surround. This shield is then grounded in the application circuit. This is the same principle as is used in RF carrying coax cable. As an indication the insertion losses of an unshielded relay at 1 GHz will be in the order of 1dB attenuation. Under the same conditions a shielded relay offers less than 0.1dB attenuation.
In addition to shielding the contact construction also differs from a conventional relay type. Should standard ¡¥dry¡¦ relay contacts be required to switch at RF problems arise due to the inconsistent contact resistance. Typically this will be around 100mOhms for a standard contact but under the life cycle conditions can rise to several Ohms. Although the actual contact resistance can be considered as not too much of a problem - specifications allow upto 100 Ohms under some conditions ¡V a variable contact resistance definitely does cause circuit instability. To avoid this problem many RF based relay contacts are lubricated during the manufacturing process. Although this initially can result in a higher contact resistance it does ensure that this initial resistance is stable during the life of the relay.
The terminology in use in the specification of RF switching relays is similar to that in other areas of relays utilised in circuit design but with the additional parameters:
„h Cross-talk attenuation is the ratio of an unintentionally coupled signal value to the original. Due to capacitance RF signals can pass from one circuit to an adjacent one, even with the contacts open. The larger the value of the cross-talk attenuation the better.
„h Insertion loss is the ratio of input to output power. The insertion loss arises from losses due to line inductances, resistances, capacitive shunts, and mismatching. Insertion loss should be as small as possible.
„h Matching attenuation characterises the ratio of the incoming RF power to the reflected power. The larger this ratio, the better the matching of the relay to the source of the RF signal, and the less the RF power will be reflected.
„h Voltage standing wave ratio ( VSWR ) results from the voltage maximum and minimum if, due to mismatching, transmitted and reflected signal superimpose. If there is no reflection then the VSWR is 1.
„h Characteristic impedance should be matched to the application circuit. Relays with a characteristic impedance of 75 Ohms are suitable for use in video based circuits whilst those of 50 Ohm are more suited to measurement technology.
Although in the past relay manufacturers have lagged behind the market requirements with the introduction of new RF relays the future is bright with the ever increasing technical performance being matched by the ever decreasing price levels. The realisation that the future of broadcasting transmission and cellphone networks are only achievable in the digital, not analogue, domain has resulted in a wide range of relay and switch components capable of offering designers the optimum choice when considering their circuit parameters.