Relay contact materials – does it matter?

Many relay users source a standard product, and with it, the standard offering with respect to contact material. More often than not they are perfectly happy, never have a problem and don’t give alternative materials a second thought. For some applications however, access to alternative contact materials can be essential.

Power switching
Power switching up to 50A is generally possible with standard industrial relays, whilst higher currents are usually the province of contactors. The principle contact materials used for relays with nominal contact ratings within the range 5 to 50A are most commonly silver nickel, silver cadmium oxide and silver tin oxide.

Silver nickel has been around almost forever. The relatively small nickel content (ten percent) is primarily to harden the silver mechanically, and to increase the resistance to electrical erosion of the contact faces, therefore making it that much more robust under heavier electrical load. It is well suited for resistive loads at the full nominal current rating of the contact, and for other loads where the load current is less. It is an economical and good performing general purpose material and quite often the standard material for many power relays.

Silver cadmium oxide has been popular for perhaps 50 years, particularly for its good performance when switching inductive and motor loads. Contact material erosion is lessened and in particular the material has an improved resistance to contact welding under conditions of short-term, high peak, in-rush currents that result from switching large contactor coils, incandescent lamps and small motors.

Unfortunately, even though virtually all experts agree that the cadmium content is so small and so well bound into the bulk silver that it presents no environmental hazard, its use has for some time been limited by the ‘RoHS’ European Directive 2002/95/EC. In its first edition, cadmium was prohibited completely - but a further revision allowed its use in electrical contacts. And the so-called ‘RoHS II’ 2011/65/EU still permits this, but establishes a deadline (unless any further revision in the next months) of July 2016 for general use and July 2024 for industrial monitoring and control instrumentation. For this reason, and keeping in mind that such a directive is not applied at all in several markets (e.g. automotive, or extra-EU countries such as US or BRICS), Finder will maintain availability of relay versions with silver cadmium oxide – for the foreseeable future.

Silver tin oxide is a more recent innovation, and like AgCdO, is produced by a powder/sintering process - unlike AgNi, which is a true alloy. The incredibly fine grinding of tin oxide into sub-micron particles, the even dispersion of this within the powdered silver, and the final high pressure forming to make the contact, is a procedure that requires the most meticulous process control. In the early days of AgSnO2 the quality control, and therefore the performance, of these sintered materials was not always as consistent as it needed to be. However, today the high performance of AgSnO2 can be relied on, and nowhere more so than in the handling of large peak inrush currents primarily caused by power factor correction capacitors associated with fluorescent and other gas discharge lamps, and also the input circuitry associated with modern energy saving lamps, CFL or LED. 

The fundamental problem with switching into capacitors is that there is virtually no intentional current limiting in circuit. Instantaneous currents are therefore limited only by source and line impedance, and will be in the order of several hundred, if not a few thousand amperes. Similar peak currents occur when powering-up switched mode power supplies and variable speed inverter drives. Not surprisingly, contact welding has historically been a major problem in these applications, but with careful evaluation of the application against known performance of the relay under such conditions, it is often possible to predict the likely improvement that a change to AgSnO2 will bring. However, the relay manufacturer will need an extensive bank of performance data and the ability to experimentally test and evaluate in a repeatable manner, and in a short time scale.

Reliable switching at low power levels
At the other end of the current scale contacts eroding away, or welding together are not the concern – it is that the contacts make a reliable and low resistance connection.

Quite simply, the lower the voltage and current being switched, the more difficult it is for the contact surfaces to produce a good connection. This is well understood by relay manufacturers and great care is taken to ensure that there is adequate pressure between the contacts and that minimum levels of cleanliness are maintained during manufacture.

Nevertheless, there are choices that the user can, or indeed should, exercise when it comes to the subject of contact materials. Broadly speaking, one should avoid using relays with power contact materials, as the characteristics that made them good power switchers tend to work against reliable low-level switching. However, occasionally there arises the need to switch both power and low-level circuits - then the only realistic option would be to select a power relay with the apparent anomaly of having gold plated contacts. ‘Anomaly’ since it makes little sense to gold plate power contacts, as gold is expensive and would simply be burnt off under power switching conditions. ‘Apparent’ because we know that there will be the odd occasion when this solves a mixed switching application with reliability at both ends of the scale. There is, however, a very important aspect to this. The gold must be plated to a significant thickness – avoiding any suggestion of using a gold flash that is typically of the order of 0.2 micron.

This is not just because such a thin coating will mechanically wear through within a few thousand operations - so do not be mistaken into thinking that because the relay only operates in your application once a month all will be well – it won’t! For reliable low level switching a gold plate will be excellent, but a gold flash is likely to be worse than bare silver. The reason for this is a very interesting mix of physics and chemistry at play – but unfortunately the in-depth explanation is beyond the scope of this article.

Backtracking a little, what do we mean by low level switching? Typically a 16A power relay from Finder’s range of relays has a specified minimum switching load of 10V / 10mA / 1,000mW, which for a relay specifically designed to switch loads up to 4kW, is not bad. The specification means that all three minimum values should be met.

A 7A medium power relay with AgNi contacts has a minimum switching specification of 5V / 5mA / 300mW. This relay is also available with gold-plated contacts when the revised values become 5V / 2mA / 50mW.

If a much lower voltage must be reliably switched, consider two contacts in parallel. This dramatically lowers the minimum switching load - two parallel gold contacts make it possible to handle loads down to 0.1V / 1mA / 1mW. It may be useful to appreciate that statistically the unreliability of two contacts in parallel is equal to the unreliability of the single contact raised to the power of two. So, just to illustrate the maths, a one percent unreliable switching circuit becomes 0.01% unreliable - i.e. a 100x improvement in reliability. And for three contacts in parallel the unreliability would be raised to the power of three – a 10,000x improvement in reliability.