Why does your component fail for no reason?
Oct 16, 2021
Sometimes the device is "dying," sometimes it is stressful but not obvious.
The "end of life" of a device is a cumulative decay effect originating from physical or chemical changes.
Everyone knows that electrolytic capacitors and certain types of film capacitors are "mortal" because of the combined action of trace impurities (oxygen, etc.) and electrical pressure, and chemical reactions occur in their dielectrics.
Integrated circuit structures follow Moore's Law and become smaller and smaller, and the migration of dopants at normal operating temperatures leads to an increasing risk of device failure within decades (instead of the original hundreds of years). In addition, the fatigue caused by magnetostriction can cause mechanical fatigue of the inductor, which is a well-known effect. Certain types of resistive materials will slowly oxidize in the air, and when the air becomes more humid, the oxidation rate will increase. Likewise, no one would expect the battery to be effective forever.
Therefore, when selecting a device, it is necessary to understand its structure and possible aging-related failure mechanisms; even if the device is used under ideal conditions, these mechanisms may have an impact. This article will not discuss failure mechanisms in detail, but most reputable manufacturers pay attention to the aging of their products, and are generally familiar with working life and potential failure mechanisms. Many system manufacturers provide relevant information on the safe working life of their products and their restriction mechanisms.
However, under proper working conditions, the life expectancy of most electronic devices can reach decades or even longer, but some will still fail prematurely. The reason is often the pressure of not being noticed.
In this "Uncommon Question Answers" column, we keep reminding readers: A useful saying to quote Murphy's Law is "The laws of physics don't fail to work just because you didn't pay attention to it." Many stress mechanisms are easily overlooked.
Anyone who designs electronic products for use in marine environments will consider salt spray and humidity-this is justified, because they are terrible! In fact, many electronic devices may encounter chemical challenges that are not so terrible, but may still cause harm.
Human (and animal) breath contains moisture and is slightly acidic. Kitchens and other household environments contain all kinds of mildly corrosive fumes, such as bleach, disinfectant, various cooking fumes, oils and alcohol. All of these fumes are not very harmful, but we should not take it for granted that we The circuit will be "safe for life" under the condition of being well protected. Designers must consider the environmental challenges that the circuit will encounter, and when economically feasible, they should design to minimize any potential hazards.
Electrostatic damage (ESD) is a stress mechanism, and warnings related to this are the most common, but we often turn a blind eye.
When PCBs are produced, the factory will take adequate measures to eliminate ESD in the manufacturing process, but after delivery, many PCBs are used in systems that do not have adequate protection against ESD caused by general operations. It is not difficult to do adequate protection, but it will increase a little cost, so it is often overlooked. (Maybe because of the economic downturn). In the most extreme cases of normal use, assessing what ESD protection is required for system electronics and considering how to achieve it should be part of all designs.
Another factor is overpressure. Few people require semiconductors or capacitors to be safe even if they suffer major overvoltages, but it is common for large-value resistors to encounter voltages far greater than the absolute maximum values listed in the data sheet. The problem is that although its resistance is high enough to not become hot, a tiny arc may be generated inside, causing it to drift slowly and deviate from the specification, and eventually short-circuit. Large wire-wound resistors usually have a breakdown voltage of hundreds of volts. Therefore, this problem was not common in the past, but nowadays, small surface mount resistors are widely used, and their breakdown voltage may be lower than 30 V, which is quite susceptible to overvoltage.
Large currents can also cause problems. Everyone is familiar with an ordinary fuse—it is a piece of wire. If an excessive current flows through it, it will heat up and fuse, thereby preventing short circuits in the power supply and other similar problems. However, if there is a very high current density in a very small conductor, the conductor may not become very hot, but it may eventually fail.
The reason is the so-called electromigration 3 (sometimes also called ion mobility). That is, the momentum transfer between the conducting electrons and the diffusing metal atoms causes the ions in the conductor to gradually move, causing the material transport effect. This makes the thin conductors carrying large direct currents become thinner and thinner over time and eventually fail.
But some parts will fail like a fuse, that is, blown, such as wires or conductive traces on semiconductor chips. A common cause of this phenomenon caused by high current is that the capacitor charging current is too large. Consider a 1 µF capacitor with an ESR of 1 Ω. If you connect it to a 110 V, 60 Hz AC power source, about 41 mA of AC current will flow through it. But if the voltage is at the maximum value (110√2 = 155.6 V) when connected to the AC power supply, only the ESR will limit the current, and the peak current will reach 155.6 A, although its duration is less than 1 μs, which is enough to damage many small signals Semiconductor device.
Repeated surges may damage the capacitor itself, especially electrolytic capacitors. This is a particularly common failure mechanism in cheap low-voltage switching power supplies ("wall adapters") used to charge small electronic devices. If it is inserted at the wrong time of an AC cycle, the rectifier and capacitor will carry a very large inrush current. If this happens many times, the device may eventually be damaged. Using a small resistor in series with the rectifier can limit this inrush current and minimize the problem.
If we are lucky, ESD or overvoltage/overcurrent events will immediately damage the device, so it is easy to know where the problem is. But it is more common that the damage caused by pressure causes the device to fail, and the pressure that caused the failure in the first place has long since disappeared. It is very difficult or even impossible to diagnose the cause of such failures.
No matter what circuit is designed, it is necessary to consider the working life and failure mechanism of the device used, as well as whether there are any potential problems or stress sources that can cause damage to the device under the most extreme use conditions allowed. Any such issues should be considered and minimized in the final design as much as possible.