Talk:Failure of electronic components
|WikiProject Electronics||(Rated C-class)|
|WikiProject Guild of Copy Editors|
Need some good photos of failed electronic components
Photos would would help readers recognize what some failed components look like. I'd contribute some, but I'm not up to speed on uploading and processing photos for the Web. Surely, somebody has some good "mugshots" of faulty components? Reify-tech (talk) 00:52, 2 May 2011 (UTC)
Different kinds of failures
As a retired Reliabilty Engineer, I would point out that classic Failure Modes and Effects Analysis clearly distinguishes between HOW components, subsystems and systems fail, the WAYS in which those failures have on the operation of systems, equipment, users, missions and impact to services and the like. This is called the failure MODE. Looking forward in the system to the impact/effect of the failure. This article is not discussing this, but discussing the CAUSES of the failures. This is/was in my day called FAILURE or FAULT ANALYSIS and was in effect looking at the issue in the other direction - towards the source of the problem i.e the CAUSE not the IMPACT. As the article is dealing with CAUSE, the term FAILURE MODE ANALYSIS is not appropriate and is in effect confusing for young engineers trying to learn the discipline.
This Wiki article is very useful as there is a lot of good stuff in it, but it should be rewritten in the context of what I have said above. — Preceding unsigned comment added by 18.104.22.168 (talk) 03:08, 9 November 2011 (UTC)
- That still has mode in the title, which was the root of the original objection. Without mode, Failure of electronic components still says the same thing. SpinningSpark 16:57, 17 October 2015 (UTC)
- Unfortunately contents in Wikipedia regarding Failure of electronic systems is missing. Andries (talk) 17:50, 17 October 2015 (UTC)
List of 'Failure Modes' by type and frequency, possibly with a FLOW-CHART indicating how one can pinpoint failure type
The number of failures listed within the article seems significant. However, what is NOT indicated is some type of generic process (which would work in the VAST majority of cases) for determining WHICH failure has afflicted a particular machine. Further, the type, frequency and PROBABILITY of a particular failure mode would be good to know. Arguably, if designers are doing their jobs well, many of these failure modes should be unlikely (especially if they stick to industry and manufacturing guidelines, etc...). Presumably, one would look at the SYMPTOMS exhibited by a machine in order to find out which failure mode has occurred. There are the complex issues of what happens to a machine when MORE THAN ONE failure mode has afflicted it - how would one pinpoint the failure modes in these cases? The key point is to get an idea, at least for 'modern' machines (manufactured in past 20 years, say) of the probability of a certain failure mode. ASavantDude (talk) 11:12, 21 August 2015 (UTC)
- These are difficult questions and the answers differ wildly per machine and even model of a machine. I think it would be overambitious to try to answer these questions in Wikipedia.
- I know from experience that capacitors have a high failure rate and not just elcos but also ceramic capacitors. Andries (talk) 19:07, 14 November 2015 (UTC)
Thermal runaway mythbusting
Thermal runaway was a problem associated with early germanium transistors and diodes, which unlike modern silicon components exhibited significant leakage current even when off. This leakage current was heavily temperature dependent, thus once a germanium device exceeded its rated temperature the buildup of leakage current would accelerate until device failure occurred.
Bipolar silicon power transistors may have a small temperature coefficient of conduction, but they do not exhibit this accelerating leakage effect. Thermally-related failures typically arise from 'secondary breakdown' due to the circuit designer not understanding the principles of safe operating area. Or, failure may arise through a simple inability to calculate the actual die temperature for any given heatsink temperature and power loading -the actual chip temperature under full load being way outside of its ratings even though the heatsink does not appear excessively hot. Either way, since the mode of failure is not understood, it is ascribed to 'thermal runaway.'
With MOSFETS, a different situation applies. The usual cause of failure here is that of an excessive gate-to-source voltage being applied at some point in the operating cycle, due to poor driver circuit design.
There is a case where thermal runaway is exhibited in modern electronics, and that is with lithium batteries. Abuse of a lithium cell, typically shorting or overcharging, may cause the cell to overheat, resulting in an internal insulation failure. The internal short may then cause the cell to continue to self-heat until fire occurs. The heat released by this cell may cause other cells in the battery to also ignite. I'm not sure if this qualifies as an electronic failure, but worth mentioning as it's an important safety consideration.
Whilst thermal runaway is exhibited in other situations, notably in chemistry as above, in the field of electronics it is more often an excuse for, "It broke and I'm not sure why."