Reed relays: from voiceband calls to GHz ATE, now facing MEMS disruption part 1

Reed relays have evolved from their audio roots to handle GHz signals; MEMS technology may push them aside.

Electromechanical relays are like dinosaurs that haven’t gotten the news that they are supposed to be extinct. Instead, they beget new generations, get designed into new applications, and even take over new territory. Countless millions of these relays are used in new designs to solve often otherwise intractable problems from DC to RF, switching low-level signals and power, while addressing conflicting objectives.

For designers who are comfortable with the solid-state, no-moving-parts world of semiconductors, the electromechanical relay (EMR) may seem like a holdover from an era long past. That’s not true at all, as EMRs have distinct advantages over solid-state relays (SSRs) in many applications. However, EMRs do face a new form of solid-state competition, albeit with micro-sized moving parts, in some applications, in the form of MEMS-based devices.

Start with basics

The classic EMR is credited to Sir Michael Faraday and is the oldest “electrical” component (along with the basic on-off switch). It functions by energizing a primary-side coil, which pulls in an armature, causing the contacts to close (or open, depending on the design, as shown in Figure 1). It’s the representation envisioned by most people who know anything about relays, and is used in basic electricity training courses as a very tangible circuit element. The first electricity-based digital computers used thousands of these in the first half of the 20^th century to form logic gates and computing elements.

Figure 1. (left) In the EMR, when no current is flowing in the electromagnet coil, the armature is pulled up by the spring and its common (COM) contact connects to the normally closed (NC) contact; (right); when a voltage is applied to the electromagnet coil the current flowing in the coil produces magnetic energy in the iron core which pulls the armature down, and the COM contact switches from NC to normally open (NO). (Image: Glolab Corporation)

EMRs have many virtues, including:

The relay contacts create a basic switch closure, and current through it can be AC or DC, independent of the coil drive; the contact resistance is in the sub-milliohm range, so the voltage drop across the contacts is very close to zero; the open-contact resistance is an air gap and therefore in the multi-megaohm range with near-zero leakage current.
The EMR is a completely passive device, which has implications for ruggedness and reliability. It is both electrically and mechanically robust, resisting spikes, transients, and EMI.
The relay can be a multipole, multiple-contact device, with more than one NO or NC contact pair; three, four, or even more independent NO and NC contacts are available, with double-pole/double-throw (DPDT) being the most common.
The multiple contacts do not have to carry the same type and rating of loads, which is another benefit; some contacts can be rated for low-level signals, while others can be rated for power.
Relays can be designed for coil currents as low as 10 or 20 mA or as high as tens of amps, with contacts rated to handle currents ranging from a few tens of mA to several orders of magnitude greater.
EMR contacts are largely signal “agnostic,” as long as the signals being handled are within the voltage and current maximum ratings; further, it is irrelevant whether it’s a power signal, data signal, or a mix across multiple contacts. Further, the load does not have to be well-known or defined; it just has to be within the design limits.
The relay is very easy to troubleshoot; all that is needed is an ohmmeter to measure the unpowered coil continuity and contact resistance, and a simple AC or DC power source to energize the coil.

The classic electromechanical relay has served the industry well and continues to do so, with many millions sold every year. Nonetheless, it was not suitable for one class of application: switching audio-band telephone signals in a central office (CO). In that situation, the copper loops of the two parties involved in the conversation needed to be connected to each other. This was accomplished via an X-Y matrix with one caller connected directly to another caller if they were both in the same CO, or one caller connected to a “trunk line” to another CO for calls behind that CO.

A classic EMR would be too large, costly, and power hungry, and would be exposed to the dirt, dust, and other environmental conditions of the CO. The resultant lack of long-term reliability would pose a maintenance problem, contrary to the Bell Telephone system’s target of a 40-year operating lifetime – can you even imagine that goal now? Even modest contact degradation over the years of use, due to the ambient conditions, would affect signal quality and reliability.

Even as the original relay-based telephony switching matrix advanced to the highly sophisticated crossbar switch design, the exposed contacts of that arrangement lacked the desired long-term reliability.

The next section examines the reed relay and its application in ATE systems.