Alexander Teverovsky, QSS Group, Inc./NASA Goddard Operations, 301.286.6216
Alexander.A.Teverovsky.1@gsfc.nasa.gov

Ashok K. Sharma, NASA GSFC, 301.286.6165
Ashok.k.Sharma.1@gsfc.nasa.gov

Micromachined relays combine benefits of solid-state devices, such as low size, weight, power consumption, and time response with conventional electromechanical relays, such as low leakage currents and high radiation hardness. MEMS switching devices do not generate spurious signals at high frequencies and have low insertion losses, high linearity, and broad bandwidth, which is advantageous for developing RF and microwave frequency systems. These features make micromachined relays very attractive for space applications, especially for a new generation of small and nano-satellites.

One of the major reliability concerns in MEMS switches is contact sticking. In this respect, thermally actuated relays have advantages over electrostatic relays. The actuation mechanism in thermally activated devices creates significant mechanical forces during opening and closing, which overwhelms potential adherence forces and micro-welding of metal contacts.

This paper reports results on quality and reliability evaluation of the first micromachined relays commercially available from Cronos. The parts have been characterized in a wide range of temperatures (from -100 °C to +160 °C) and load conditions (voltages from 10 V to 100 V and currents from 0 mA to 200 mA).

Mechanical integrity of the parts has been evaluated by subjecting them to multiple mechanical shocks in the range from 100 G to 1,000 G (Fig.1). Life testing was performed at different load conditions during more than 108 switching cycles. Figures 2 to 4 illustrate some results of the life testing.

To evaluate conditions that cause contact failures, experiments with capacitance discharge through the contacts were performed. The values of capacitance changed from 4.7 mF to 47 mF, the load resistance varied from 3 Ohm to 100 Ohm, and the voltage across the capacitors was incrementally increased from 1 V to 200 V. The results (Fig. 5) could be explained based on a simple energy dissipation model and allowed for estimation of critical energy required to cause contact failures.

It is known that low-pressure conditions can cause failures in conventional electromechanical relays [1]. For the thermally actuated MEMS relays, vacuum was found also to be a detrimental environment. All observed failures were caused by overheating of the polysilicon heaters (see Figures 6 and 7) and were due to a reduction of heat dissipation in the actuator under vacuum conditions.

Typical failure mechanisms associated with different test conditions (see Figures 7 through 9), as well as the processing and manufacturing defects, are discussed. The detailed test results will be reported and posted on the NEPP Web site when they have been completed.

[1]. A. Teverovsky, Relay Failures Specific to Space Applications, ISTFA’99, Proceedings from the 25th International Symposium for Testing and Failure Analysis, 1999, Santa Clara, CA, pp. 285-292

Fig. 1. Mechanical shock test results. No failures after 10 shocks 400G.	Fig. 2. Contact resistance variation during life test cycling. No load conditions.
Fig. 3. Intermittent failures at 60V/2mA during 100 pulses step test.	Fig. 4. Resistive load life test at different contact voltages.
Fig. 5. Capacitance discharge step test. Marks indicate experimental data, solid lines – calculations.	Fig. 6. Kinetics of heating currents at normal conditions and during failure in vacuum.

Fig. 7. Failure of the heater during vacuum testing.	Fig. 9. Failed contacts after life test cycling at 60V, 10 mA.	Fig. 8. Cracks in polysilicon after mechanical shock testing.