Specifications of reed relays, which are used for current switching in ATE and other applications are explained, including carry current, lifetime, minimum switch capacity, hot switching, operating speed and thermoelectric switching.

BY KEVIN MALLETT, Pickering Electronics, Clacton-on-Sea, Essex, U.K.

Reed relays, which use an electromagnet to control one or more reed switches without requiring an armature, are used for instrumentation and automatic test equipment (ATE), high voltage switching, low thermal EMF, direct drive from CMOS, RF switching and other specialized applications.

Reed relays are deceptively simple devices in principle. They contain a reed switch, a coil for creating a magnetic field, an optional diode for handling back EMF from the coil, a package and a method of connecting to the reed switch and the coil to outside of the package. The reed switch is itself a simple device in principle and relatively low cost to manufacture thanks to modern manufacturing technology.

The reed switch has two shaped metal blades made of a ferromagnetic material (roughly 50:50 nickel iron) and glass envelope that serves to both hold the metal blades in place and to provide a hermetic seal that prevents any contaminants entering the critical contact areas inside the glass envelope. Most (but not all) reed switches have open contacts in their normal state.

If a magnetic field is applied along the axis of the reed blades the field is intensified in the reed blades because of their ferromagnetic nature, the open contacts of the reed blades are attracted to each other and the blades deflect to close the gap. With enough applied field the blades make contact and electrical contact is made.

The only movable part in the reed switch is the deflection of the blades, there are no pivot points or materials trying to slide past each other. The reed switch is considered to have no moving parts, and that means there are no parts that mechanically wear. The contact area is enclosed in a hermetically sealed envelope with inert gasses, or in the case of high voltage switches a vacuum, so the switch area is sealed against external contamination. This gives the reed switch an exceptionally long mechanical life

Inevitably in practice the issues are a little more complicated. The ferromagnetic material is not a good conductor and in particular the material does not make a good switch contact. So the reed blades have to have a precious metal cover in the contact area, the precious metal may not stick to the blade material very well so an underlying metal barrier may be required to ensure good adherence. Some types of reed relay use mercury wetted contacts, consequently reed relays that use plated contacts are often referred to as “dry” reed relays. The metals can be added by selective plating or by sputtering processes. Where the reed blade passes through the glass envelope any plating (in many cases there may be none) requires controlling to avoid adversely affecting the glass to metal hermetic seal. Outside the glass seal the reed blades have to be suitably finished to allow them to be soldered or welded into the reed relay package, usually requiring a different plating finish to that used inside the glass envelope.

The materials used for the precious metal contact areas inside the glass envelope have a significant impact on the reed switch (and therefore the relay) characteristics. Some materials have excellent contact resistance stability; others resist the mechanical erosion that occurs during hot switch events. Commonly used materials are ruthenium, rhodium and iridium– all of which are in the relatively rare platinum precious metal group. Tungsten is often used for high power or high voltage reed switches due to its high melting point. The material for the contact is chosen to best suit the target performance – bearing in mind the material chosen can also have a significant impact on manufacturing cost. Sealed in a long, narrow glass tube, the contacts are protected from corrosion, and are usually plated with silver, which has very low resistivity but is prone to corrosion when exposed, rather than corrosion-resistant but more resistive gold as used in the exposed contacts of high quality relays. The glass envelope may contain multiple reed switches or multiple reed switches can be inserted into a single bobbin and actuate simultaneously. As the moving parts are small and lightweight, reed relays can switch much faster than relays with armatures. They are mechanically simple, making for reliability and long life.

This article reviews and explains common specifications used for reed relays (FIGURE 1).

FIGURE 1. Reed relays are used for current switching in ATE and other applications.

Carry current

Carry current is the current that the reed relay can support through its contact without long term damage. The life of the relay should be indefinite under this condition though some reed relays may also have a pulse current rating which can be applied to the relay without damage.

The carry current is determined primarily by the contact resistance of the relay and the heat sinking to the environment. As the current increases the temperature of the reed blades increases until it reaches a temperature where the material is no longer ferromagnetic (Curie Temperature). Once that temperature is reached the relay contacts may open since the blades no longer respond to the magnetic field. The blade temper- ature is clearly dependent upon the current and relay path resistance – the normal assumption is that this is a square law (with current) relationship. In reality, the temperature rise is significantly more than a square law since the metallic resistance also increases with temperature, the magnetic field drops with temperature because of coil resistance rise and the mechanical properties of the blade can change. Consequently like all relays, exceeding the rating can result in a type of thermal runaway.

The packaging of the reed switch has a significant impact on the temperature rise, a lead frame tends to conduct heat to the outside world while the plastic encapsulation materials insulate it. The packaged reed relay will always have a lower current rating than that of the reed switch because manufacturers quote the rating with the reed switch directly exposed (no coil, no plastic packaging). The coil power will also add to the heating effect. Consequently Pickering Electronics always de-rates the reed relay ratings to ensure that the relay switch remains within its design limits.

There is also another subtle effect that occurs as the carry current increases – the signal creates its own magnetic field that twists the blades and therefore can modulate the contact resistance. The blade twisting may start to see a contact resistance rise as the blade contact area reduces or changes.

Care must be taken not to exceed the relays ratings and pulse ratings should take account of the square law relationship between current and temperature.

It becomes difficult to manufacture reed relays with a carry current of greater than 2A because the contact area has to be increased and that tends to make the bladed stiffer and require a higher magnetic field strength to operate them.

Lifetime

The lifetime of reed relays is critically dependent on the load conditions the reed switch encounters. For reed relays which are instrument grade the mechanical lifetime is much greater than 1 billion operations – they are mechanically simple devices that rely purely on the deflection of a blade to operate and there are conse- quently fewer wear out mechanisms.

The blade contact area though stills wears as they are opened and closed. If the signal load when the blade closes or opens is low then the wear out is very slow, as the load increases and hot switching (interruption or closure of a signal live carrying significant current or voltage) occurs higher temperatures are generated at the contact interface and this makes the materials more prone to wear. DC signals can also result in the migration of metal from one contact to another and without regular polarity reversal eventually the underlying contact materials are exposed with their poorer conduction characteristics. Hot switching can also create a temporary plasma in the contact area with high local temperatures, rapid operation of a relay under load can start to raise the contacts temperature to an extent where premature wear out can occur. The life an instrument grade reed relay can vary by three orders of magnitude according to the load conditions, perhaps 5 billion operations under no or light load to 5 million operations at a heavy load.

Minimum switch capacity

Some types of relay have a minimum switch capacity, if the relay is closed on a very low level signal (current or voltage) oxide or debris on the relay contacts can remain at the interface and cause a higher than expected resistance, or even an open circuit. This tends not to be the case with reed relays because the precious metal contacts are sealed in a hermetic glass envelope containing inert gas. Minimum switch capacity tends to be a characteristic of higher power mechanical (EMR) relays.

Hot switching

Hot switching occurs whenever a relay contact is opened or closed with a signal (current and voltage) is present. As the contacts move apart or close an arc can be created which transfers material from one contact to another, or simply redistributing the material. As the contact plating is damaged the resistance will eventually start to rise until the relay is no longer fit for the intended application.

For reed relays hot switching tests are always conducted into resistive loads. The hot switch capacity of a reed relay is typically quoted at a current/voltage that results in the number of operations that the relay will support around 10million operations. The data sheet specifies a hot switch current (the limiting factor at low voltages), a hot switch voltage (limiting factor at low current) and a power (from the product of the open contact voltage and the closed contact current).

Operating speed

The operate time is the time from when the relay coil is energized or de-energized to when the contact reaches a stable position.

For a normally open contact when the coil is energized the current, and therefore the magnetic
field, in the coil rises until the blades start to move closer together until they make contact. The contacts may impact each other sufficiently rapidly that there is bounce where for a short duration the contact is inter- mittently closed then opened. The operate time should be the time from when the relay coil was energized until the contacts are stably closed.

If the coil is driven from a higher than specified coil voltage the closing speed of the relay will be faster, however once the contacts make there may be more contact bounce as they meet with greater force. Overdriving the coil can also increase the release time since the magnetic field takes longer to collapse to the point where the contacts start to open.

For a normally open Form A (SPST) contact the release time is the time from when the coil is de-energised to when the contact is open. This operate time can be dependent on how the reed relay is driven, the presence of a protection diode on the coil will increase the release time. Typically, the release time is around one half the operate time.

Soft and hard weld failures

Operation of reed relays (or EMRs) under high load conditions causes one of the most common failure mechanisms for relays – a failure where the contacts are welded together. By convention these welds are classified as being either soft or hard failures. In the event of hard failure the contacts tend to be welded together and nothing will separate them. This is an easy fault to identify. Soft failures occur where the contacts sick but eventually come apart without any additional assistance. The failure is caused by small areas on the contact welding together, but the weld area is sufficiently small that the reed blades will separate because of their sprung nature. They could spring apart very quickly, or it may take several seconds to spring apart depending on how hard the weld is.

In either case the impact on the user is that the switching function of the relay is impaired and this is likely to have an adverse impact on the user application. So in either case the relay will require replacing since the defect is unlikely to improve with time. The cause of the weld will also need to be investigated and corrected.

Thermoelectric EMF

The cause of thermoelectric voltages is often misunderstood by users, and often misrepresented in articles and on the internet. The effect of thermoelectric EMF’s is to generate a small voltage (measured in microvolts) across the relay terminals when the relay is closed (FIGURE 2).

FIGURE 2. Thermoelectric EMF’s are used to generate a small voltage (measured in microvolts) across the relay terminals when the relay is closed.

The voltage arises whenever a metal wire has a temperature gradient across it (the Seebeck Effect), if one end of the wire is at a different temperature to the other then a voltage will appear which is dependent on the temperature difference and the materials that make up the wire. Reed relays use a mix of metals, and these can have different temperature drops across them which results in a voltage appearing at the relay connection terminals. The voltage is not created at a connection junction. Nickel iron has quite a strong thermoelectric EMF, so designing reed relays with low thermal EMF’s can be a challenge.

The number and type of materials varies according to FIGURE 2. Thermoelectric EMF’s are used to generate a small voltage (measured in microvolts) across the relay terminals when the relay is closed. to the way the reed switch is designed and how it is packaged. If the relay was perfectly symmetric in construction (so the materials used from each contact to the reed switch were the same and the reed itself was perfectly symmetric in all materials and dimensions) and all heat sources in the relay body (primarily due to the coil) then this would be the case. However in reality the symmetry is not perfect so a residual voltage will arise.

Users can also degrade the performance by how they use the relay. When mounted on a PCB if the PCB has a temperature profile across it then that will generate an additional thermal EMF. Relay manufacturers usually assume that the thermal EMF is zero when the relay is first closed since up to that point no heat source exists inside the relay body. However, a temperature profile across the PCB (caused by the presence of other heat sources or forced air cooling) will create a thermal EMF.

Reed relays that have excellent Thermal EMF performance are typically designed to be as symmetric in design as possible and to use highly efficient coils to avoid heating the reed switch. Typically though, this results in a physically larger relay.

Two pole designs often quote the Differential Thermal EMF, this is the voltage generated between the two switches (usually) in a single package.

Assuming the relay design is reasonably symmetrical to a first order the voltage in one switch is the same as the other, so the differential voltage can be much smaller for the relay. Differential and single ended Thermo Electric EMF numbers should not be directly compared or confused with each other.