10 Misconceptions About Chain Reliability, Have You Been Hired?

In fact, many of us are wrong about general empirical things, and once this is used in design, one can imagine reliability of product. Therefore, it is said that “when choosing components of a scheme, first demonstrate its impossibility and carefully discuss feasibility; The competition in electronic design is not about whose design is better, but about whose design makes fewer mistakes.”

Myth 1: Product failure = product unreliable

Sometimes, problem of product is not a research and development problem. There have been cases where equipment for middle and highly developed regions of China has been exported to Colombia because it has been well used in China, but there have been frequent failures. The reason for failure is that height above sea ​​level in developed areas of mainland China is relatively small, so tightness of equipment in high mountain areas is questioned, and leakage rate increases as pressure difference between inside and outside of equipment increases. When project was approved, only low height was considered, so their design was not a problem. Your boss asked for it. Whoever decided to export this model to Colombia is culprit. If you care about research and development The boss of company participated in decision making without raising objections. He is simply biggest sinner. After all, sales managers can be forgiven if they do not understand technology when making decisions, and mistakes of a technical vice president are incompetent. Product reliability is “the ability to perform specified functions for a specified time and under specified conditions.” Readers should carefully study this definition, study things to find out, and see who can gain more knowledge when they can fit this definition. The conditions in area of ​​use often exceed specified conditions, and this excess may be implicit.

Myth 2: transition process = steady state process

The introduction to "Curve Affecting Product Reliability and Social Harmony" is a good illustration of content of this graph.

Myth 3: Downgrading is easy, it's not a problem

Everyone can derank how to draw, everyone can, but not everyone can survive by drawing. For more information, see "Derating of Electronic Products" article on this blog. Here is just a summary:

1. Devices with same functions but different processes have different derating factors;2. Adjustable devices and fixed value devices have different derating factors;3. Various load. , different derating factors;4. Different derating factors for wires of same specification in multi-turn and single-turn applications;5. Some parameters cannot be lowered;< br> >6. The derating of junction temperature cannot be ignored.

Misunderstanding 4. That device can be used with confidence

Why is device damage often referred to as "fire"? The reason is that most device failures are due to overheating. Pay attention to two points. First, ambient temperature of device ≠ ambient temperature of whole machine, and environment of device is affected by other devices in chassis Due to effect of heat dissipation, ambient temperature of main unit is higher than ambient temperature of whole machine. Secondly, we can recall third question at beginning of speech, for more details, see section “Common errors in measuring ambient temperature and device load characteristics”.

Myth 5. The reliability of electronics has nothing to do with mechanics and software

Installation, wiring, layout and spraying may affect electrical performance. Electromagnetic compatibility, virtual welding, heat dissipation, vibration and noise, corrosion and grounding are related to structure. Software error prevention, error detection, error correction and error tolerance measures avoid problems with mechanical and electronic defects.

Misunderstanding 6: The device is very simple, it doesn't matter if there is a datasheet

When designing, you must get Datasheet of all devices, then read all graphs, diagrams and parameters on it, and finally relate to these curves in design. The figure below shows current-voltage characteristic of a diode. When designing, it is necessary to carefully confirm static operating point of device in our circuit.

Myth 7: Maintainability has nothing to do with me

What is purpose of working on reliability of electronics? To make money. What do you rely on to make money? Open source and cost reduction. Difficult to increaseincome and easily cut costs. Do not always think about saving on material costs. Material costs are saved and maintenance costs are high. dying early, you will die sooner or later. Why bother?? It is better to die early and be born early. It is best to pay attention to maintainability and save on this part of cost. This is real profit.

Misconception 8: If process control is bad, there are no good masters

Poor process control is not only a problem for craftsmen, but also process of building a value chain. The design engineer's requirements for device, selection of manufacturer by purchasing engineer, and control content of inspection link should be developed for key indicators of device. The detection method should not represent mechanism of component failure and damage, and assembly link should not lead to damage oven, anti-static treatment of manual soldering table, etc.), factory inspection link should check drift part of device parameters, which may lead to product failure, and maintenance link should not present a failure. It can be seen from above that if a problem occurs, two technologists can guarantee it. Therefore, it is concluded that a special approach is to ensure consistency. The prerequisite for consistency is that designer provides sufficient technical information as a matter of priority, and process is based only on design drawings and design documentation to ensure that production reliability is infinitely close to design reliability.

Misconception 9: The relationship between MTBF value and failure rate of one particular machine

MTBF is a macroscopic and statistical concept, while single device failure is a microscopic and concrete concept. Customers love to ask question, “Your product has an MTBF of 10,000 hours, so if I buy yours, will there be any issues within 10,000 hours?” 1 km or 1 kg?

Myth 10. Reliability problems can be fixed with more testing

Because this question can be classified as one of ten most popular myths, its definition is, of course, incorrect. There are three conclusions:1. Some problems cannot be detected by simulation tests at all;2. Test methods = engineering calculation + specification review + simulation test + electronic simulation; >3. The corresponding operating time at low temperatures cannot be calculated from results of the heat resistance test.