Although most manufacturers cite the rated lifetime for their switches, they seldom indicate what this figure represents. However, various manufacturers’ specifications offer clues as to how they arrive at this rating.
Electrical lifetime applies only to conductive switch types: mechanical, reed, conductiver rubber, membrane etc. The specifications for such switches typically supply figures for most if not all of the following characteristics:
- Contact bounce time
- Rated load during switching
- Rated load when settled
- Contact resistance
The figures for the above characteristics, when not qualified, indicate the end-of-life characteristics for the switch. That is to say, once the switch has reached the maximum number of operations, it will continue to function, but it is no longer expected to meet the specifications. In particular, contact resistance and bounce time may increase with use, and some manufacturers cite typical or factory-new figures in addition to the end-of-life figures.
It’s worth noting that the lifetime of a mechanical switch is inversely proportional to the electrical load that it endures. The catalogue entry for ITT ETL 18 specifies that the rated lifetime at 5 V 1 mA is 20 million, but at maximum switching power of 3.5 W (28 V at 125 mA), switch lifetime is reduced down to only 1 million cycles. Here it’s clear that rated lifetime indicates damage to the switch from the electrical load. ITT rated ETL 18 for a maximum of 50 mΩ contact resistance when new, and a maximum of 1 Ω contact resistance at rated lifetime. Thus, circuitry that can tolerate contact resistance above 1 Ω may continue to operate normally; Hi-Tek Dovetail Series for example has a contact resistance of 10 Ω at end of life, ten times that of ETL 18. Contact bounce is given as 5 ms maximum regardless of age, and the specifications do not indicate whether it will increase beyond this limit at the nominal lifetime. It will not necessarily do so, if the contact resistance is what defines the rated lifetime.
Alternate action switches across the board have consistently lower rated lifetime than the equivalent momentary switch. The table below provides a number of examples, in approximate order of introduction to the market:
|Series||Type||Momentary lifetime||Alt. action lifetime|
|Fujitsu FES-1/-2||Reed||1 million||100 thousand (10%)|
|Datanetics DC-50||Mechanical||100 million||50 thousand (0.05%)|
|Datanetics DC-60||Mechanical||10–20 million||50 thousand (0.25–0.5%)|
|RAFI RS 76 C||Hall effect||100 million||100 thousand (0.1%)|
|RAFI RS 76 M||Mechanical||10 million||100 thousand (1%)|
|Hi-Tek Dovetail Series||Mechanical||20 million||5 million (25%)|
|Cherry MX||Mechanical||20–50 million||500 thousand (1–2.5%)|
|ITT ETL 18||Mechanical||20 million cycles||50 thousand (0.25%)|
|Alps SKCL||Mechanical||20 million||30 thousand (0.15%)|
Alternate action switches are generally rated below a million cycles, with Hi-Tek Dovetail Series being a curious exception at an unusually high 25% of the lifetime of the momentary switches. Meryl Miller’s original design for Datanetics DC-50 alternate was scrapped due to not meeting the expected lifetime, which is intriguing considering that the production design—shared by DC-50 and DC-60—still only offered a low lifetime of 50 thousand operations, or a mere 0.05% of the rated life of the momentary types. Nonetheless, Alps KCL alternate action was rated even lower, at only 30 thousand cycles, and ITT ETL 18 also only offered 50 thousand cycles for alternate action. Cherry’s alternate action mechanism was unusual in being rotary, which might have played a part in its higher lifetime, although compared to the original 20 million cycle lifetime of Cherry MX, this was still only 2.5% of the lifetime of the linear models. (Cherry MX’s rated lifetime was later increased to 100 million, but by this time the alternate action model was discontinued and a fair comparison cannot be made.)