A brief discussion on the reliability design of electronic products

    1 Introduction

Reliability refers to the ability of a product to complete specified functions under specified conditions and within a specified time. Any product, whether it is mechanical, electronic, or mechatronics, has a certain degree of reliability. Product reliability is closely related to experimentation, design, and product maintenance.

There are many indicators to measure reliability, and the common ones are as follows: (1) Reliability R(t), which is the probability of completing the specified function under specified conditions and within the specified time, also known as the mean time between failures (MTBF); (2) ) Mean maintenance time MTTR refers to the time required for a product to recover its specified functions from the discovery of a fault; (3) Failure rate λ (t) refers to the probability of product failure after the product is used under specified usage conditions until time t; in addition There is also validity A(t) and so on. The reliability changes of products generally follow certain rules. Its characteristic curve is shown in Figure 1. Because it is shaped like a bathtub, it is usually called the "bathtub curve". In the early stages of experimentation and design, the early failure rate is high due to errors in product design and manufacturing, imperfect software, and insufficient component selection. Through design modifications, process improvements, aging components, and complete machine testing, the product It enters a stable accidental failure period; after a normal period of use, the product enters a wear and tear period due to component wear, machine aging, maintenance and other reasons. This is why the reliability characteristic curve has a "bathtub curve" shape.

Reliability design involves probability theory, Boolean algebra, graph theory, set theory, optimization theory, etc. This article will discuss the reliability design technology of electronic products.

The reliability design of electronic products needs to pay attention to the following basic accuracy:

●The product structure and circuit should be as simple as possible.

●Try to use mature structures and typical circuits.

●The structure should be simplified, building blocks, and plug-ins.

●If new circuits are used, attention should be paid to standardization.

●When adopting new technologies, we must pay full attention to inheritance.

●Use digital circuits as much as possible.

●Use integrated circuits as much as possible.

●Logic circuits should be designed to be simplified.

●Performance indicators and reliability indicators should be considered comprehensively.

●Traditional techniques and customary operating methods should be adopted as much as possible.

●New reliability design technologies should be continuously adopted.

In electronic products, commonly used reliability design technologies include component derating design, redundant design, thermal design, electromagnetic compatibility design, maintainability design, drift design, fault tolerance design and fault weakening design, etc. Some also include Software reliability design. These main design technologies are introduced below.

2 Reliability design technology

2.1 Derating design

The so-called derating design is a design technology that allows components to be used in a stress state lower than the rated value. In order to improve the reliability of components and extend the life of the product, the working stress (such as electrical, thermal, mechanical stress, etc.) imposed on the device must be consciously reduced. The conditions for derating and the amount of derating must be comprehensively determined. , to ensure that the circuit can work reliably and maintain its required performance. Derating measures also have different regulations depending on the type of component. For example, resistor derating is to reduce the ratio of its use power to rated power; capacitor derating is to make the operating voltage lower than the rated voltage; semiconductor discrete device derating is to Keep power dissipation below rated values; contact components must reduce tension, torque, temperature and other limitations associated with the particular application.

Derating of electronic components usually has an optimal derating range. Within this range, changes in the working stress of the component have a significant impact on its failure rate. The design is easy to implement and does not require the weight of the equipment. It pays too much in terms of size and cost. Therefore, the appropriate derating level should be determined based on the specific application of the component. Because if the derating is not enough, the failure rate of the components will be relatively large and the reliability requirements will not be met; on the contrary, excessive derating will make the design of the equipment difficult and will cost a lot in terms of weight, volume, and cost of the equipment. The cost may also lead to an unnecessary increase in the number of components, which will in turn reduce the reliability of the equipment.

The levels of derating are divided into three levels, namely Level I derating, Level II derating and Level III derating.

Level I derating is the maximum derating. Greater derating beyond it will limit the reliability growth of components and make the design difficult to implement. Level I derating is applicable to the following situations: equipment failure will seriously endanger the life and safety of personnel, may cause significant economic losses, lead to the failure of work tasks, cannot be repaired after failure or is economically uneconomical, etc.

Level II derating means that when components are derated within this range, the reliability of the equipment increases dramatically, and the equipment design is easier to implement than Level I derating. Level II derating is applicable to situations where equipment replacement will degrade the working level or require unreasonable maintenance costs.

Level III derating means that when components are derated within this range, the reliability of the equipment will increase the most and the difficulty in equipment design will be minimized. It is suitable for equipment failure that has little impact on the completion of work tasks and does not endanger the work tasks. Complete or quickly fixable.

2.2 Thermal design

As the density of electronic components used in modern electronic equipment becomes higher and higher, this will cause thermal coupling between components through conduction, radiation and convection. Therefore, thermal stress has become one of the most important factors affecting the failure rate of electronic components. For some circuits, reliability depends almost entirely on the thermal environment. Therefore, in order to achieve the expected reliability purpose, the temperature of the components must be reduced to the lowest level that is actually achievable. Some data show that for every 10°C increase in ambient temperature, the life of components is reduced by approximately 1/2. This is the famous "10℃ rule". Thermal design includes three types of technologies: heat dissipation, radiator installation and refrigeration. Here the author mainly talks about heat dissipation technology. Methods commonly used in applications:

The first is the conductive heat dissipation method. Materials with high thermal conductivity can be used to manufacture heat transfer elements, or contact thermal resistance can be reduced and the heat transfer path can be shortened as much as possible.

The second is convection heat dissipation. There are two methods of convection heat dissipation: natural convection heat dissipation and forced convection heat dissipation. The following points should be noted for natural convection heat dissipation:

●Excess space must be left when designing printed boards and components;

●When arranging components, attention should be paid to the reasonable distribution of the temperature field;

●Pay full attention to the principle of applying chimney wind;

●Increase the contact area with the convection medium.

The forced convection cooling method can use a fan (such as a fan on a computer) or a dual-input push-pull method (such as a push-pull method with a heat exchanger).

The third method is to use thermal radiation characteristics, which can be done by increasing the roughness of the surface of the heating element, increasing the ambient temperature difference around the radiator, or increasing the area of ​​the radiator surface.

In thermal design, the most commonly used method is to add a heat sink. The purpose is to control the temperature of the semiconductor, especially the junction temperature Tj, so that it is lower than the maximum junction temperature TjMAX of the semiconductor device, thereby improving the reliability of the semiconductor device. The equivalent thermal circuit diagram when the semiconductor device and the heat sink are installed and working together is shown in Figure 2. The meaning of each parameter in the figure is as follows:

RTj—internal thermal resistance of semiconductor device, ℃/W;

Tj—semiconductor device junction temperature, ℃;

Tc—semiconductor device case temperature, °C;

Tf—radiator temperature, ℃;

Ta—ambient temperature, ℃;

Pc—power used by semiconductor devices, W.

According to Figure 2, the thermal resistance RTf of the heat sink should be:

RTf=(RTj-Ta)/Pc-RTj-RTc

Radiator thermal resistance RTf is the main basis for selecting a radiator. Tj and RTj are parameters provided by semiconductor devices, Pc is a parameter required by the design, and RTc can be found in thermal design professional books. The following introduces the selection of radiators.

(1) Selection of natural cooling radiator

First, calculate the total thermal resistance RT and the thermal resistance RTf of the radiator according to the following formula, that is:

RT=(Tjmax-Ta)/Pc

RTf=RT-RTj-RT.

After calculating RT and RTf, the radiator can be selected based on RTf and Pc. When selecting, according to the selected heat dissipation RTf and Pc curves, find out the known Pc on the abscissa, and then find out the thermal resistance R'Tf of the radiator corresponding to Pc.

Just choose a reasonable radiator according to the principle of R'Tf≤RTf.

(2) Selection of forced air cooling radiator

When selecting a forced air-cooled radiator, the appropriate radiator and wind speed should be selected based on the thermal resistance RTf and wind speed υ of the radiator.

2.3 Redundant design

Redundant design is a design technology that uses one or more identical units (systems) to form a parallel connection. When one of the units fails, the other units can still make the system work normally. Redundancy is divided into hot redundancy reserve and cold redundancy reserve according to characteristics; according to the degree of redundancy, there are double redundancy, triple redundancy and multiple redundancy; the scope of safety redundancy is divided into component redundancy and component redundancy. Redundancy, subsystem redundancy and system redundancy. This design technology is usually used in more important situations that require higher safety and economy, such as boiler control systems, program-controlled exchange systems, aircraft control systems, etc.

2.4 Electromagnetic compatibility design

Electromagnetic compatibility design is also environmental resistance design. First of all, we must understand what electromagnetic compatibility problems are. Electromagnetic compatibility problems can be divided into two categories: one is when electronic circuits, equipment, and systems fail to meet expected technical indicators due to mutual interference or external interference during operation; Another type of electromagnetic compatibility problem is that although the equipment is not directly affected by interference, it cannot pass the national electromagnetic compatibility standard. For example, computer equipment produces exceeding the limit value specified in the electromagnetic emission standard, or the electromagnetic sensitivity or electrostatic sensitivity reaches Not required. In order to make equipment or systems achieve electromagnetic compatibility, technologies such as printed circuit board design, shielded chassis, power line filtering, signal line filtering, grounding, and cable design are usually used. When designing and laying out printed circuit boards, you should pay attention to the following points:

●Circuit connections at all levels should be shortened as much as possible to reduce parasitic coupling as much as possible, especially for high-frequency circuits;

●High-frequency lines should try to avoid arranging wires in parallel to reduce parasitic coupling, and they should not be bundled together like low-frequency circuits;

●When designing circuits at all levels, try to arrange them in the order of the schematic diagram to avoid cross-arrangement of circuits at all levels;

●The components of each level of circuit should be as close as possible to the transistors and electron tubes of each level of circuit, and should not be distributed too far away. Try to make each level of circuit form its own loop;

●One-point grounding or nearby grounding should be used at all levels to prevent interference caused by ground current loops. High-current ground wires and ground wires of Qin current loops should be set up separately to prevent large currents from flowing into the public ground wire and causing strong coupling. interference;

●Components that generate strong electromagnetic fields and components that are sensitive to electromagnetic fields should be arranged vertically, away from each other or shielded to prevent and reduce mutual inductance coupling;

●The ground wire in a strong magnetic field should not form a closed loop to avoid interference caused by ground loop current;

●The power supply wire should be close to the ground wire (of the power supply) and arranged in parallel to increase the power filtering effect.

2.5 Drift design technology

The main reasons for drift are the tolerance between the standard parameter values ​​of components and the actual values, the impact of changes in environmental conditions on component performance, or the degradation of component performance caused by use in harsh environments.

If the component parameter value drifts beyond its design parameter range, the equipment or system will not be able to complete the specified function. Drift design is a design method that writes characteristic equations based on circuit principles at the design stage, and then collects the distribution parameters of components to calculate their drift range so that the drift results are within the design range to ensure normal use of the equipment.

2.6 Interconnect reliability design

Since there are connectors in most electronic products, in order to reduce the failure rate of these connecting parts, it is necessary to design interconnect reliability. Commonly used methods include:

●Pay attention to the selection of connectors. Printed circuit boards should use large boards or multi-layer boards as much as possible to reduce the number of connecting points:

●Minimize the number of pluggable points to improve reliability, and redundant designs can be used for important components;

●When two plugs face each other at the same time, one should be fixed and the other floating to ensure alignment and insertion;

●Adopt mechanical fixing method;

●For components that are frequently plugged and unplugged, it is best to design single-sided wiring;

●The connection space should be divided in an orderly manner;

●Feeders and ground wires should be installed concealed.

In addition, in the reliability design of electronic products, maintainability design technology, software reliability design technology, mechanical parts reliability design technology, fail-safe design technology and some new reliability design technologies are sometimes used. Due to space limitations, this article will not introduce them one by one.