1. Overview Relays are widely used products, including household appliances such as air conditioners, color TVs, refrigerators, and washing machines; they are also used in industrial automation control and instrumentation. Among electronic components, relays are generally considered one of the least reliable. In overall system reliability design, relays, potentiometers, adjustable inductors, and variable capacitors are recommended to be used sparingly or not at all. However, due to the unique electrical and physical characteristics of relays in control circuits—their high insulation resistance in the off-state and low conduction resistance in the on-state—no other electronic component can compare. Coupled with their high standardization, versatility, and ability to simplify circuits, relays remain widely used. With the rapid development of technology, the use of relays in programmable communication equipment is further increasing. Therefore, ensuring the reliability of relays to meet the reliability requirements of the entire system has become a focus of attention. The reliability of electronic components consists of two parts: the inherent reliability of the component and the reliability of the component during use. Inherent reliability is the foundation of component reliability, primarily relying on effective control by component manufacturers in design and manufacturing to ensure that manufactured components meet the required reliability levels. Usage reliability, on the other hand, focuses on how to guarantee and improve component reliability to meet the reliability requirements of the entire system. Without high-reliability components, it's impossible to manufacture highly reliable electronic equipment; therefore, the inherent reliability of components is the basis of overall system reliability. However, even with high-reliability components, a highly reliable system is not guaranteed. This is where usage reliability comes in. Usage reliability refers to utilizing reliability design techniques based on the characteristics of various components. These techniques include appropriate component selection, derating design, tolerance and drift design, vibration resistance design, thermal design, dustproof design, radiation resistance design, electromagnetic compatibility design, ergonomic design, and maintenance design to maximize the inherent reliability of components and achieve the overall system reliability requirements. According to statistics from relevant departments analyzing the causes of system failures, over 40% of failures are due to inappropriate component selection. With the continuous improvement of component manufacturing technology, while the inherent reliability of components has been significantly improved, operational reliability has become particularly important. Furthermore, as complete systems become increasingly functional and use more numerous components, the reliability requirements are also becoming increasingly stringent. Therefore, operational reliability has received increasing attention from the scientific community and has developed into a new discipline—human engineering. Because relays are mechatronic components, composed of electromagnetic and mechanical transmission parts, they are much more complex than other electronic components. In addition, some assembly and adjustments during manufacturing are done manually, resulting in lower product consistency and reliability. However, with the implementation of certain preventative measures during use, satisfactory results can still be achieved. Failure analysis of failed relays revealed that failures due to usage account for more than 30%. This analysis shows that, besides inherent quality issues, improper use is a major reason for low relay reliability. Therefore, we will focus on measures to improve relay reliability during use. There are many types of relays; this study will focus on the operational reliability of electromagnetic relays, which are currently widely used. Faced with a wide variety of relay products, how to rationally select and correctly use them is a practical problem that system developers and designers must pay close attention to and prioritize solving. To achieve rational selection and correct use, it is essential to fully study and analyze the actual operating conditions and technical parameter requirements of the system, and, in accordance with the "value engineering principle," appropriately propose the technical performance requirements that the selected relay products must meet. In the reliability design of the entire system, the rational selection of components is required. Component selection and control is a task that requires multidisciplinary knowledge and should generally be completed jointly by component engineers, reliability designers, overall and circuit designers, and failure analysts. First, the importance of the entire system, reliability requirements, environmental conditions, and cost requirements must be comprehensively considered and selected. Specifically, the following factors can be analyzed and studied one by one to confirm the required level and value range. The following aspects must be emphasized during selection: 2. Relay Application Environmental Conditions Climatic stress factors mainly refer to temperature, humidity, atmospheric pressure (altitude), coastal atmosphere (salt spray corrosion), sand and dust pollution, chemical gases, and electromagnetic interference. Considering the system's widespread applicability across various industries and natural environments nationwide, and taking into account its crucial requirement for reliable operation year-round, key components of the system must utilize fully sealed relays (metal-sealed or plastic-encapsulated, with metal-sealed products being superior to plastic-encapsulated products). Only fully sealed relays possess excellent long-term resistance to harsh environments, good electrical contact stability, reliability, and stable load switching capability (unaffected by external climate). 2.1 The Influence of Temperature on Relays Relays are heat-sensitive components. High temperatures accelerate the aging of internal plastic and insulating materials, cause contact oxidation and corrosion, hinder arc extinguishing, and deteriorate electrical parameters, reducing reliability. Therefore, the design must ensure relays are not placed near heat-generating components and provide good ventilation and heat dissipation. Although relays are heat-sensitive, excessively low temperatures (such as -55℃ in military aviation conditions) cannot be ignored. Low temperatures exacerbate contact adhesion, causing condensation on the contact surface and the formation of an ice film on the armature surface, preventing normal contact switching, especially severe in low-power relays. Tests have shown that for some domestically produced low-power relays manufactured according to national standards, although the specified operating temperature is -55℃, in reality, the relays cannot perform normal switching under this condition. It is recommended to allow sufficient margin when selecting relays. For critical military electronic systems, it is recommended to choose products meeting national military standards. 2.2 The Effect of Low Atmospheric Pressure on Relays: Under low atmospheric pressure, the relay's heat dissipation deteriorates, the coil temperature rises, causing changes in the relay's given pull-in and release parameters, affecting the relay's normal operation. Low atmospheric pressure can also reduce the relay's insulation resistance, making it difficult to extinguish the arc at the contacts, easily causing the contacts to melt, affecting the relay's reliability. For harsh operating environments, it is recommended to use a fully sealed system. 2.3 The Effect of Mechanical Stress on Relays: This mainly refers to stress factors such as vibration, impact, and collision. For control systems, the main considerations are resistance to seismic stress and mechanical stress; therefore, small intermediate relays with a balanced armature mechanism are recommended. Electromagnetic relays use cantilever structures with low natural frequencies. Vibration and shock can cause resonance, leading to a drop in relay contact pressure. This can result in momentary disconnection or contact chatter, and in severe cases, structural damage. The movable armature can also malfunction, affecting the relay's reliability. It is recommended to incorporate vibration-damping measures in the design to prevent resonance. 2.4 Insulation Withstand Voltage: The exposed insulators at the leads of non-sealed or sealed relays are subject to long-term dust and moisture contamination, leading to a decrease in insulation strength. Under overvoltage conditions during inductive load switching, this can cause insulation breakdown failure. Due to the inherent insulation characteristics of relays, the following technical characteristics must be considered when selecting a relay: 2.4.1 Sufficient creepage distance: generally >3mm (operating AC 220V); 2.4.2 Sufficient insulation strength: >AC 2000V (operating AC 220V) between conductors without electrical connection, >AC 1000V between contacts in the same group; 2.4.3 Sufficient load capacity: DC 220V inductive; 5～40ms, >50W; 2.4.4 Long-term ability to withstand climatic stress: coil anti-mold breakage, long-term stable and reliable insulation dielectric level. 3. Excitation Coil Input Parameters These mainly refer to factors such as over-excitation, under-excitation, isolation between low-voltage excitation and high-voltage (220V) output, the influence of temperature changes, long-distance wired excitation, and electromagnetic interference. These are all factors that must be carefully considered to ensure reliable system operation. Excitation according to the excitation amount specified by the relay is a necessary condition to ensure its reliable and stable operation. The technical specifications for relays generally specify the operating voltage, pull-in voltage, and release voltage for the coil. To ensure the normal operation of the relay, the circuit connection must ensure that these three voltages meet the values ​​specified in the technical specifications under all circumstances. Otherwise, the relay will not function correctly. 3.1 Regarding Series Power Supply Excitation Methods: Many users employ a series voltage divider power supply method to apply excitation to the relay coil, driving the relay to operate. This excitation method is generally not recommended. The relay's pull-in time mainly depends on the circuit's time constant T, and T = L/R. When a series resistor R1 supplies power to the relay coil, R = R1 + R2, then L/R2 > L/(R1 + R2). Obviously, connecting R1 in series reduces T, accelerating the relay's pull-in time. Especially when R1 >> R2, and the voltage is very high, the pull-in time will be significantly reduced. Excessive speed of moving parts will increase the impact, collision, and rebound during engagement, thereby increasing contact bounce, accelerating mechanical wear, and reducing the contact's load capacity and mechanical life. Therefore, series power supply excitation alters the normal operating state specified in the original relay design and is generally not advisable. It can only be used when contact bounce or mechanical wear does not pose a significant threat to actual use, and a faster operating speed is particularly needed. 3.2 The use of series-connected relay coils: Using multiple relay coils in series and then exciting them with a DC 220V power supply must be used with caution. 3.2.1 For relays of the same type and specifications, since the impedances (including DC resistance and transient reactance) of each coil are roughly the same and the difference is small, using a series voltage divider excitation method is generally not a problem. Practice has proven this to be feasible. 3.2.2 For relay products of different types or specifications, due to the inconsistent impedance of different relay coils, and the significant difference in impedance depending on the instantaneous inductive reactance, the difference in excitation voltage (determined by the instantaneous voltage division ratio) across each relay coil at the moment of series excitation will inevitably be large. This will inevitably result in some relays being in an overvoltage excitation state while others are in an undervoltage excitation state. The switching sequence and speed of each relay contact will undergo fundamental changes, inevitably leading to reversed action order, speed, and unreliable switching. Therefore, series voltage division excitation is not suitable for relay coils of different types and specifications. 3.3 Regarding the parallel use of relay coils in complex control circuits, it is common to use two (or more) relay coils of different types (such as contactor K1, small sensitive relay K2) in parallel. In this case, practical problems may occur such as delayed release of K1, decreased contact arc-breaking capability, repeated reverse excitation of K2, and contact malfunction. In DC control circuits, the magnetic energy stored in the K1 and K2 coils may differ significantly. When the coil power supply is lost, the energy stored in K1 (high magnetic energy) will be discharged through the coil of K2 (low magnetic energy), generating a reverse current. This results in a prolonged release time for K1, a slower arc-breaking speed at the contacts, and a prolonged arc-burning time between contacts; K2 has a short release time and is subsequently excited by the reverse discharge current, even experiencing a momentary re-engagement after release, leading to malfunctions. In practical applications, care should be taken to avoid these unreliable phenomena caused by insufficient research. 4. Contact Output (Switching Circuit) Parameters These mainly refer to the nature of the contact load, such as lamp load, capacitive load, motor load, inductor, contactor coil, choke coil load, resistive load, etc.; and the contact load values ​​(open-circuit voltage value, closed-circuit current value), such as low-level load, dry circuit load, small current load, large current load, etc. Selecting a suitable relay based on the load nature and load capacity of the device driven by the relay is a fundamental condition for the reliable operation of the relay. The failure or unreliability of the relay mainly refers to whether the contacts can complete the specified switching circuit function. If the actual load capacity being switched exceeds the switching load capacity specified by the selected relay, the relay cannot operate reliably. 4.1 Regarding the load of the contacts: Relay contact failure is the core cause of relay failure. When the actual load voltage switched by the contacts is less than the arcing voltage and the current is less than 1A, especially under medium current (test standard is DC 28V, 0.1A), low level (10-30mV, 10-50μA), or dry circuit (meaning the relay contacts close first, then connect the millivolt/microampere level load), the failure mechanism and failure mode of the contacts during actual operation are completely different from the actual rated power load being switched. Precisely to meet the different requirements of different loads, different products have different design, manufacturing processes, testing, and experimental requirements. Therefore, when actually selecting relay products, one must not mistakenly assume that the relay contacts are suitable for all loads from zero to the specified rated value, nor should one assume that the smaller the actual load passing through the contacts is compared to the rated load specified in the product standard, the more reliable the relay. For example, a contact that can reliably switch a 220V, 10A load may not be able to reliably switch a 10mA load, and it should not be used to switch low-level or dry circuit loads. Therefore, for medium current, low-level, dry circuit loads, it is recommended to choose a fully sealed product with a metal casing and excellent contact reliability. Generally, in reliability design, derating is the most effective measure to improve reliability. For other components, if other factors such as cost and size are not considered, the greater the derating, the higher the reliability. However, relays are different from other components. It is not that the lower the load stress applied to the contacts, the more reliable they are. This is mainly determined by the contact failure mechanism. When the contact current reaches 100mA, the arcing effect of the contacts is significantly weakened. The carbonaceous material precipitated by the contacts under high temperature conditions cannot be burned off by the arc and deposits on the contact surface, increasing the contact resistance and affecting contact reliability. When the contact load is below 10mA or below 50mV, the contact reliability is significantly reduced because the voltage cannot break down the film resistance on the contact surface, resulting in low-level failure. Especially under high-temperature conditions, contact oxidation is accelerated, making low-level failures more severe. Therefore, loads above 10 mA and below 50 mV are considered low-level loads. While the load stress on a relay should not be too low, the load stress specified in the technical specifications is the maximum rated value of the contacts, a parameter that should never be exceeded under any circumstances. Exceeding this value during use can, at best, shorten the lifespan and reduce reliability, and at worst, burn out the contacts, causing failure. This is mainly because the arcing generated when the relay contacts operate under heavy loads causes the contacts to melt, forming unevenness on the contact surface, resulting in mechanical seizing that prevents separation. The greater the contact load, the larger the arcing, and the greater the likelihood of contact burnout. From the above analysis, it is clear that appropriate derating remains an effective measure to improve relay reliability. For proper contact load use, under normal circumstances, the load should be designed to be above 100 mA, and below 80% of the rated load value given in the technical specifications for better reliability. It is worth noting that the rated load value of the relay contacts is given under resistive load conditions. When the load used is inductive, capacitive, or lamp load, it can generate a surge current of 10 times. Therefore, if it is not a resistive load, it should generally be converted when using it. Load conversion: Resistive load current, Inductive load current, Motor load current, Lamp load current 100% 30% 20% 15% 4.2 Regarding capacitive loads, the relay contacts, as self-holding contacts for switching capacitive load circuits, are prone to contact sticking and failure to release. This is because the charging and discharging process of the capacitor is similar to the capacitor energy storage spot welding process. Further analysis and experiments show that after giving a 22μF capacitor a sufficient DC 220V voltage, and then energizing the relay to directly short-circuit discharge it, the pure silver contacts can produce a soldering failure phenomenon within 10 times. Theoretically, the discharge current of the capacitor is: i=（UT）eT where U—is the voltage across the capacitor; R—is the resistance of the discharge circuit; T—is the time constant. Since R is approximately equal to the contact resistance of the contact point and approaches zero, at the instant of discharge, i = U/R; i is very large, meaning that all the energy stored in the capacitor is discharged through the contact in a very short time, directly causing spot welding failure. Therefore, long transmission lines, filters for eliminating electromagnetic interference, power supplies, etc., are all highly capacitive. Relays used for this type of load should be selected based on the characteristics of the equipment. 4.3 Regarding the parallel use of relay contacts 4.3.1 Power cannot be increased by connecting contacts in parallel. Sometimes, when one set of contacts cannot meet the power requirements of the circuit, two or more sets of contacts are connected in parallel to ensure the power requirements of the circuit. However, because there is a small time difference when the relay contacts operate (generally, the operating time of two sets of contacts differs by 0.1 ms to 0.2 ms), the first set of contacts to connect will bear all the power and switch under overstress conditions. It is easily burned out by the arc formed by the large current and fails. Therefore, when using relays, power cannot be increased by connecting contacts in parallel. 4.3.2 Generally, parallel connection of relay contacts is not used to improve reliability: In reliability design, redundancy can improve reliability. Some designers subjectively intend to improve the reliability of control circuits by using parallel connection of relay contacts, taking advantage of the principle of redundancy. However, the function of a general control circuit is to control the circuit by using the mutual switching action of contacts. If parallel connection of contacts is used, although the reliability of connection is improved, the reliability of disconnection is reduced. Therefore, it is not advisable to use parallel connection to improve reliability for general relay-controlled switching circuits. Only for special requirements, such as circuits that can complete the specified function with a single connection or disconnection (such as launching a satellite, where only the relay contacts are required to connect the rocket's ignition system to complete the mission), can parallel connection of contacts improve reliability. 4.4 Correct Connection of Relay Contacts 4.4.1 Normally open contacts should be used as much as possible and normally closed contacts should be used as little as possible. When connecting relay contacts, normally open contacts should be used as much as possible and normally closed contacts should be used as little as possible. This is because normally open contacts bounce back less times when they operate than normally closed contacts. As is well known, contact bounce has a negative impact on circuits and shortens contact life. 4.4.2 Correct connection of changeover contact polarity; the connection of changeover contact polarity has a significant impact on contact life. The correct connection should be that the movable contact is connected to the power supply cathode, and the fixed contact is connected to the power supply anode. The reason is that tests on two different connections show that, under the same load conditions, the arcing time of the contacts is reduced by half when connected with the correct polarity as described above, compared to the opposite polarity connection, thus improving contact life.