TheAbandonwareGuy wrote:

If anyones wondering how and where i ran across this I've been assimilating knowledge of 486 PC's at a pretty inconcievable rate […]
Show full quote

If anyones wondering how and where i ran across this I've been assimilating knowledge of 486 PC's at a pretty inconcievable rate lately because I'm trying to get enough knowledge to repair a 486/SX 25 motherboard in my AST Advantage Pro. This particular segment of text comes from Aubrey Pilgrims "Build your own 486 PC and save a bundle, 2nd Edition" courtesy of Openlibrary.org.

Is this true? I've never heard this theory before, certainly not from someone who i presume knows his way around a circuit board. I can think of very few circuit boards that have lasted aslong as i (hopefully) will last. Has something changed in circuit board design since then that just obsoleted this theory? Capacitors and boards wear out regardless of how little or much voltage you push through them based on my understanding, where as based on my interpretation of Aubrey's statement logically my graphics card should never die within my lifetime so long as it survives 168 hours (Yeah... way past that now 🤣) and i never overvolt it. I would think capacitors or something would fail? Or does this statement apply only to the PCB and not the chips, capacitors, and resistors attached there of?

I don't know, I just figured I'd share this here as it seems somewhat interesting. Am i just interpreting this wrong or is the guy who wrote this misinformed/full of shit?

electro migration will eventually kill ever SMID ever manufactured no matter what. Weather it takes 30 minutes or 30 million years depends 🤣

https://en.wikipedia.org/wiki/File:Bathtub_curve.svg

Pretty standard stuff in electrical and electronic engineering.

"Burn in and Infant Mortality" sounds like the name of a death metal band.

Seems to me capacitor plague might not have been a thing when that was written.
https://en.wikipedia.org/wiki/Capacitor_plague

https://en.wikipedia.org/wiki/Bathtub_curve

Reality begs to differ 😀

First the article writes about semiconductors, then 'electronic components'. Especially some of the latter have a finite life. Like electrolytic caps - the liquid escapes along or through the rubber seal. Then there's the heat-up-cool-down cycles. It puts stress on the materials and can finally crack them. Things made from oil may eventually decay. Even in good enviroments corrosion never stops to 100%.

I used to repair photocopiers, and saw a bulletin about a defect in a composite power supply.

The bulletin said that some power supplies left the factory with capacitors that had did not pass a burn in procedure. It could lead to a failure on the exposure lamp.

So the burn in procedure is not a legend, or at least some companies take it seriously enough to do it at factor y.

TheAbandonwareGuy wrote:

http://puu.sh/qaIyW/06b7786cde.jpg I would think capacitors or something would fail? Or does this statement apply only to the PC […]
Show full quote

I would think capacitors or something would fail? Or does this statement apply only to the PCB and not the chips, capacitors, and resistors attached there of?

Well, it does talk about purely semiconductors, but then moves on to complete circuits, which is not the same thing.

Semiconductors are things like chips, transistors, diodes (LEDs too), etc. Electronic circuits and devices contain also other components such as resistors, capacitors, inductors etc, even the printed circuit board itself, and these are not semiconductors.

Everything wears out, it's just a thing if it wears out enough to matter much, or if it affects another part of the circuit.
Everything is also manufactured with certain lifetime in mind, so usually better and more expensive components are used in longer lasting devices.

For example, light emitting diodes wear out. LED bulbs, LCD backligts, etc. The brightness will decrease over time, and the normal specs is to say a LEDs lifetime is about 20 thousand hours after which the brightness has dropped to 50% of the original.

Now, a LED wearing out won't cause anything to break by itself, the TV or monitor backlight just gets dimmer. But when these LEDs or circuits are powered from a switch mode power supply, when a capacitor starts to wear out and approach the end of its usable lifetime, it can cause an increase in supply output voltage. Increased supply voltage causes increase in currents flowing and increased power losses cause temperatures to rise. Overvoltage/overcurrent/overtemperature will wear out semiconductors faster than usual and when the voltage/current/temperature gets to a maximum for some IC chip or LED then it will break right away.

So yes, either things break quickly after manufacturing because of unexpected component defects, or after the expected lifetime of the product.

WTF does infant mortality have to do with CPUs dying? 😜 Was the heading software-translated from another language?

mr_bigmouth_502 wrote:

WTF does infant mortality have to do with CPUs dying? 😜 Was the heading software-translated from another language?

Well if the brand-new tech. is your newborn baby... it's an analogy.

Capacitors, and maybe some other passive components, are clearly not meant by this definition. Electrolytics have a very finite lifetime, very typically 1000 hours at 85C. (Hours double with every 10C cooler they're run, so at 35C it's 32000 hours, or less than 4 years if always on. Sitting in storage at 25C, 7-8 years.)

Same with transformers and chokes. Temperature spec and lifetime.

one of the fastest semiconductor 'death' are known to be the SSD.

That's due to the limited write cycles of flash memory.

I remember ads in Computer Shopper used to mention "burn-in" as part of the deal - usually with barebones systems.