LucyDuke
staring at the clock
What is it that causes hardware to fail? Why do all my CD drives always crap out completely for no apparent reason?
Hardware failures?
- Thread starter LucyDuke
- Start date
Genocidicbunny
Bug squasher
Who knows. Components that have mechanical parts can have them fail, such as your cd drive or a Hard Drive. Electronic components can fail from overheating or some other type of damage. Mainly though, you see problems because of excessive heat.
Stuff wears out.
Optical drives are one of the components I see die most often, along with memory, even in "high-end" computers.
Optical drives, like many other devices (hard drives, notably), are speced for power-on hours of life, meaning they'll die eventually even if you don't actually use them.
If you want to prolong the life of your drives, your best bet would be to get an external usb optical drive, and to just plug it in when you need to use it. However, it's probably not worth the effort, with internal drives running around $30.
Optical drives are one of the components I see die most often, along with memory, even in "high-end" computers.
Optical drives, like many other devices (hard drives, notably), are speced for power-on hours of life, meaning they'll die eventually even if you don't actually use them.
If you want to prolong the life of your drives, your best bet would be to get an external usb optical drive, and to just plug it in when you need to use it. However, it's probably not worth the effort, with internal drives running around $30.
aimeeandbeatles
watermelon
- Joined
- Apr 5, 2007
- Messages
- 20,112
I'm pretty sure that electronics are made to fail after the warranty runs out, so you have to buy a new one. Just a strange idea.
LucyDuke
staring at the clock
What I'm curious about is the mechanics of the failure. Does a plastic thing snap, or does a fuse blow, or what? I figure if I kick the damn thing I know what went wrong, but what actually goes bad to make the thing bust?
Aimee, it's fun to think conspiracy theory but I don't see how something could be easily designed to break after a certain period of time. After a certain number of uses, sure, but not after "three years and one day".
Aimee, it's fun to think conspiracy theory but I don't see how something could be easily designed to break after a certain period of time. After a certain number of uses, sure, but not after "three years and one day".
In every mechanical device, there is friction. Sometimes this friction slowly 'erodes' the rubbing parts, until one or both are thin enough to actually break apart. In something with very fine tolerances like a CD drive, it is sufficient that they wear enough to no longer allow parts to be correctly aligned.
Another enemy is heat. Friction produces heat. So does electric current. When we talk about electronics, we *always* have heat issues. Heat causes the metals (in particular) to fatigue, and become more brittle, or conversely, soften. This can result in obvious mechanical areas, like simple friction. As conductors fatigue, the begin to impede the flow of electricity more and more, which results in more heat. Eventually a circuit will no longer conduct the electricity applied, and the unit 'fails'.
These are just some of the ways that our toys go bad.
As for conspiracy theories, engineering and analysis can determine the "mean time to failure" for a given device. So if they have a CD play, for instance, that has a 90% survival rate to 36 months, but only a 10% survival rate at 40 months, you will give a warranty for 36 months, knowing that those customers will be back between 3-4 years for a new player.
Another enemy is heat. Friction produces heat. So does electric current. When we talk about electronics, we *always* have heat issues. Heat causes the metals (in particular) to fatigue, and become more brittle, or conversely, soften. This can result in obvious mechanical areas, like simple friction. As conductors fatigue, the begin to impede the flow of electricity more and more, which results in more heat. Eventually a circuit will no longer conduct the electricity applied, and the unit 'fails'.
These are just some of the ways that our toys go bad.
As for conspiracy theories, engineering and analysis can determine the "mean time to failure" for a given device. So if they have a CD play, for instance, that has a 90% survival rate to 36 months, but only a 10% survival rate at 40 months, you will give a warranty for 36 months, knowing that those customers will be back between 3-4 years for a new player.
LucyDuke
staring at the clock
Awesome answer, Padma. Thank you. 
So heat is the problem with non-mechanical things like circuits and stuff? 'Cause I can understand mechanical failures, that'll work just like anything, like a car or a washing machine or something. I just couldn't imagine how something with no moving parts would go about breaking.
Doesn't the time-to-failure depend on how much you use the thing? If you use it x hours a day every day it'll burn out twice as fast as half that, right?
I suppose it would be ideal if you busted your thing just before the warranty ran out, and it was replaced rather than repaired. You could double its lifetime.
So heat is the problem with non-mechanical things like circuits and stuff? 'Cause I can understand mechanical failures, that'll work just like anything, like a car or a washing machine or something. I just couldn't imagine how something with no moving parts would go about breaking.
Doesn't the time-to-failure depend on how much you use the thing? If you use it x hours a day every day it'll burn out twice as fast as half that, right?
I suppose it would be ideal if you busted your thing just before the warranty ran out, and it was replaced rather than repaired. You could double its lifetime.
Thank you!
That was a very broad, high-level description of some of the problems. The details can get very interesting, but you have to be a real electronics geek to follow some of it.
One thing I didn't mention above is that in modern, solid-state devices ("no moving parts"), the 'wires' and 'insulators' have widths measured in nanometers. A nanometer is 0.000000001 meters, or 25.4 million nanometers to the inch. Even though the smallest are currently 45 nm wide, the 'big' ones, at ~100nm, are still plenty small. It's easy to see how a speck of dust, or (more likely, since they tend to be sealed against dust) a little overheating can cause opens or even shorts in the circuits, = failure.
Some devices, even when 'off', draw small amounts of power to maintain 'state'. Newer devices, which don't need the constant trickle of current, theoretically can last longer. In lab tests, once they know the mean time to failure, they extrapolate based on what they consider to be 'normal usage', and they calculate their warranty period based on these guesses.
That was a very broad, high-level description of some of the problems. The details can get very interesting, but you have to be a real electronics geek to follow some of it.
One thing I didn't mention above is that in modern, solid-state devices ("no moving parts"
), the 'wires' and 'insulators' have widths measured in nanometers. A nanometer is 0.000000001 meters, or 25.4 million nanometers to the inch. Even though the smallest are currently 45 nm wide, the 'big' ones, at ~100nm, are still plenty small. It's easy to see how a speck of dust, or (more likely, since they tend to be sealed against dust) a little overheating can cause opens or even shorts in the circuits, = failure.Some devices, even when 'off', draw small amounts of power to maintain 'state'. Newer devices, which don't need the constant trickle of current, theoretically can last longer. In lab tests, once they know the mean time to failure, they extrapolate based on what they consider to be 'normal usage', and they calculate their warranty period based on these guesses.
Everything (at least, all known materials) degrades in some way with usage. Think about something like an airliner - obviously you have engines and electronics and all that jazz, but what about the actual skin of the plane? It's just metal. No movement, no current through it, it's just... there. Yet every time that airliner flies, the fuselage is pressurized so that you can survive at high altitudes. We're talking about a tiny amount of pressure here - about 7 psi. Yet multiply that by many tens of thousands of pressurizations and depressurizations, and the metal will begin to weaken and develop faults, to the point where there have been a few catastrophic in-flight failures due to lax inspections/maintenance.
The same idea applies to basically every man-made object.
LucyDuke
staring at the clock
Well yeah, I know stuff does wear out, I just can't always figure out how. That's why I was askin'!
Heat does cause the failure of solid state circuits. That is the reason that your CPU has a heat sink and fan, chipsets normally have heat sinks with cooling fins and fans on high performance systems, memory sticks now often have heat sinks and sometimes fans, and graphics cards now often have fans.
GoodGame
Red, White, & Blue, baby!
- Joined
- Dec 17, 2004
- Messages
- 13,725
Actually a engineer can use statistics and calculus to easily determine the probable life for product. Getting it accurate to a day would probably take a decade or so of making the product, but all things would be possible with a large enough sample. Planned product obselescence/failure is a valid market strategy.
I agree with Zelig+Aimee statements above.
I agree with Zelig+Aimee statements above.
J-man
Deity
Brian Boru
Prince
Not really. All products undergo constant design and production method changes, so the item made last month is different from the item made today. Maybe not by much, but after a year or three, it'll probably be significantly different. By then it's probably being phased out
But the general principle of what you say is certainly valid, just not very applicable in practice with all the guesses and assumptions required.
I agree re heat being the #1 enemy, that's why I leave the side door of my PC off permanently. People with overclocked [ie pushed beyond spec] high-end systems often have room fans blowing into the case, and fancy water cooling these days.
I've had good luck with drives and memory--only one CD ever failed on me. But this decade alone, I've had 3 power supplies and 4 motherboard failures--one of which took a hard drive with it. Backup is good
Genocidicbunny
Bug squasher
Hopefully you blow out the dust every week because leaving the door open just makes the pc a dust magnet.
That is a Very Bad Idea. You may see your overall temp go down according to mobo/CPU sensors, but you end up creating stagnant bubbles of air in parts of the case that are unable to circulate, and these bubbles will significantly hotter than the case air would be under normal circumstances. It's extremely bad for the components around them. PC cases are designed to promote even airflow and circulation throughout the entire chassis. Let it do what it was meant to do.
Brian Boru
Prince
No, only a small increase in dust. I clean it every time I clean my monitor, every month or two. It helps a lot to have the PC about 30" off the ground.
In theory I agree with you. In practice though, every PC is different. My default ventilation is poor, what with HD cages and IDE ribbons etc, so the airflow is probably considerably worse than the optimum case design.Speedo said:
I doubt the dead spots can get too much hotter than the rest of the case, since it's open and temp tends to normalize around the ambient room temp.
No doubt a good case with a proper fan or water system would be better, but I'm neither expert nor interested enough to pursue that. I've had my case open for years, after checking it out on some overclocking forums--which largely said "it depends."
Similar threads
