Coaxial RF relays are used throughout military systems. Systems developers and integrators such as Raytheon Systems Company (St. Petersburg, FL) must rigorously screen defective components from their suppliers in order to avoid system-level failures. Since 1993, the company has tracked failures of coaxial RF relays from two suppliers, identifying the need for design, manufacturing, and test-process changes. To aid in the testing/screening process, an automated low-current contact resistance test set was developed. The test set uses generic test equipment to identify failures (in some cases isolating relay lots with failures exceeding 15 percent). In addition, the system revealed that the contact resistance of failed relays typically exceeded 30,000 Ω, and has learned that three issues continue to impact the reliability of vendor-supplied relays: improper adjustment (lack of overtravel), excessive volatile-material condensation, and contamination.

Raytheon Systems Company (RSC) has developed some simple methodologies for troubleshooting failsafe coaxial relays. The military contractor has used coaxial RF relays in two military applications in particular: a missile Global Positioning System (GPS) antenna and a frequency-agile filter (radio). The missile application uses two relays to switch receive signals only: a null signal for the normally closed-path and antenna-receive signals for the normally open path. The filter/radio application uses nine relays in series that transmit less than 50 W on the normally open side and conduct receive signals on the normally closed side. Both applications operate in dry circuit conditions. The Engineers' Relay Handbook¹ states "that the performance of contacts under dry circuit conditions is affected by the following parameters: contact material; contact force; contact wipe; cleanliness; environment; magnitude of current and voltage." Since dry circuit applications at RSC typically pass microwatts of power through the contacts, contact resistance is an important parameter.

For years, the coaxial relay industry has used contact resistance testing to detect RF anomalies. The cost of a simple tester that detects opens (a few thousand dollars) makes that method of testing far more attractive than the investment in an RF test set for $80,000 or more. However, test results indicated that there was a correlation between receive application VSWR failures and contact resistance.

Figure 1 shows typical examples of failsafe coaxial relays for radio (left) and missile (right) applications. System-level ESS testing usually indicates a symptom of high VSWR for a failed relay, with most failures occurring for tests conducted at a cold temperature (−55°C). Failures can occur at high (+70°C) and room (+25°C) temperatures, however, as well as during temperature transitions (from cold to hot or from hot to cold).

When attempting to duplicate receive failures during RF testing, power levels are a prime concern. Because of this, voltage and current levels must also be taken into consideration when measuring contact resistance using conventional digital multimeters (DMMs). Both types of relays were designed to meet the requirements of MIL-S-3928/15.² Unfortunately, a review of the MIL-S-3928 standard reveals that it does not specify the power levels required during RF testing. Contact resistance is specified, for a maximum of 0.240 Ω. However, the voltage and current levels are not specified for that contact resistance. These shortcomings in MIL-S-3928 required RSC to develop procurement documentation that specifically addressed the needs of a particular application. RSC also took into account that coaxial RF relay suppliers do not always test their products to ensure that they will perform as needed in receive applications

RSC has attempted to simulate the receive application using generic laboratory equipment. The current failsafe coaxial relay test set uses two 6.5-digit DMMs, although it was recognized that test current varies depending on the DMM scale selected. The 300-Ω scale has a 1.7 mA test current compared to 160 µA on the 30,000-Ω scale. Typically, DMM low-range scales (300 to 3000 Ω) exceed 1.0 mA test current. The 30,000-Ω scale was chosen to limit the current and more closely mimic the receive application. When using the 30,000-Ω scale, the 6.5-digit DMMs allow a resolution of 10 mΩ.

The RSC automated test set (Fig. 2) has evolved from testing one part at a time (at cold temperature only) to eight at a time exposed to temperature cycling (+25°C, −55°C, −47°C, +71°C) with each relay actuated 800 times. The missile-application relay incorporates female pins that accept male solder tabs for the contact connections. A male pin-adapter fixture was used to connect to these contacts. Double-shielded cables with SMA (subminiature series A, MIL-C-348) connectors were used to connect between the fixtures and each of the test equipment. The relay test set employs three programmable RF scanners to handle eight relays at a time. The test set measures resistance by the two-wire rather than the (preferred) four-wire method mostly due to lack of reed switches.

The relay test set is capable of measuring each of the four states of the relay: normally closed to common closed (NCCC); normally open to common open (NOCO); normally closed to common open (NCCO); and normally open to common closed (NOCC). To increase the number of measurements during a four-hour test period, modifications were made to the test program to capture only the closed measurements (NCCC and NOCC). After measuring hundreds of relays, it was found that that the normally open data was not very valuable and was not captured.