Take the 2-minute tour ×
Super User is a question and answer site for computer enthusiasts and power users. It's 100% free, no registration required.

I've had a few computers of my own break after a few years of use, and I've seen computers owned by friends and family do the same thing. They either slow down (even running the original software they shipped with) or crash more often (kernel panic/BSOD, freezing, etc). All computers eventually fail, and not always in easily explainable ways.

For example, my first computer was a MacBook Pro. After about 3 and a half years of use, it started to freeze a few times a week, then reliably every day. OS X would stop responding; my mouse moved, but absolutely nothing responded. It ended up lasting more than four years, but I was in the habit of saving my work every time I paused in typing, and kept rigorous backups.

I've also seen old family computers running early versions of Windows completely freeze (no mouse movement) just after completing startup. And I've seen a 4 year old 12" PowerBook G4 run like molasses even after having been wiped and reset to factory settings and OS.

What causes this? Do the electronic components break down over time?

share|improve this question
12  
You might want to look into this thing called 'entropy'. –  Shinrai Jan 5 '12 at 21:37
3  
Related: Why does hardware get slower with time?. –  dmckee Jan 6 '12 at 4:15
    
Note that the old Powerbook probably had a bad battery, and that causes it to run slower. –  Daniel R Hicks Jan 6 '12 at 21:49
2  
Cruft forces. –  grawity Jan 6 '12 at 23:21
    
Please make sure that your answers actually are related to malfunctions. It's not a malfunction if a sports car slows down because it pulls a trailer, and neither is a fragmented hard disk or a bloated OS with tons of crapware running. Note that the user even mentions that reinstalling the OS does not help. –  Daniel Beck Jan 11 '12 at 21:16
add comment

10 Answers

up vote 26 down vote accepted

The environment has an effect on your computer.

Temperature changes make electronic components expand and contract, if they do that a lot of times eventually connections will get damaged and even break. This results in different resistance which reroutes the power and causes components to behave in an unexpected different way.

This is why in some data centers whole server rooms are cooled with gas to keep the temperature at a constant level, besides keeping the servers from overheating this also keeps the hardware lifetime.

Humidity changes are also not very good for electronics, this can lead into different conductivity; or in the case of condensation, it could even short circuit an electronic circuit.

This is also why they very tightly seal electronic circuits on a ship, even though the rain and sea can't reach the electronic circuits. That way, the circuits don't suddenly start to malfunction when they are in the middle of the sea. Better to be safe, than to be sorry...

There are other factors like (direct) mechanical or magnetic damage, as the environment is not the only cause of damage. An example is a hard drive running out of its lifetime or the computer being placed near magnetic materials. But components like hard disks are known to not keep on for more than a few years; while not mentioned explicitly when sold, one should keep track of its state as a good habit.

share|improve this answer
5  
Hard drive's wear over time is actually largely influenced by the environmental conditions. One of the main factors that play a keyrole in the magnetic head stack degradation - are inadequate temperature conditions. Humidity also influences regular thermal calibrations inside the drive that only make matters worse. The same reason contacts on the drives PCB suffer from impurities (oxidation) in the long run. –  XXL Jan 5 '12 at 21:48
    
The magnetic head stack degradation is mostly an internal problem, and hard drives are properly sealed against humidity. High temperature conditions indeed have an effect, but at home this mostly should not happen. But well, it's indeed still possible for people that don't cool their computer well. Good point about the contacts... –  Tom Wijsman Jan 5 '12 at 22:11
    
But how does an internal problem of a hard drive stop being one at that if its influenced by the thermal conditions (environment) the drive is kept in? MHS degrades noticeably faster if the drives operating temperature is more than 45* C (50-55* C for high capacity models) or less than 20* C. That is exactly a problem that occurs at home much more often than in the server farm (where things are cooled properly). Especially in a tight low-grade chassis that only accumulates dust over time. I would personally rate that much higher than "just a possibility", more of a pretty common occurrence –  XXL Jan 5 '12 at 22:47
    
An important word in that sentence is operating. While the room temperature can be lower than 20°C it will still operate at a higher temperature than 20°C, and those high temperatures of course result in wear. Which I agreed upon... –  Tom Wijsman Jan 5 '12 at 22:52
    
Oddly, I do not see much of a difference in pinpointing that word. I was just saying that why would you classify it as an "internal" problem, even though it's largely influenced by the thermal conditions? I mean, what did adding the word "internal" add to the point? That the operating temperature could not be altered or.. ? Also, "but at home this mostly should not happen" - while this happens at home the most - especially taking into the account that the majority of users simply purchase prebuilt computers (where internal cooling is clearly not one of their strong sides). –  XXL Jan 5 '12 at 23:10
show 1 more comment

One problem I've seen: Accumulating enough software that there is no longer enough memory to run it at a reasonable speed. Also, the updates to various software often makes them take more memory to run properly.

If your motherboard allows it, you can add more memory - either in empty memory slots or by replacing the old memory boards with new ones with more memory per board.

Make sure that any new memory you buy is matched to the type your motherboard expects, or the new memory may refuse to let your computer boot.

Also, most 32-bit operating systems (most of the older ones) have a built-in limit close to 4 gigabytes on how much memory they can actually use.

I found the web site I bought more memory from:

http://www.crucial.com/

Have the information about your computer's maker and model number ready to help find which of their many memory boards should be suitable.

Another problem some people have: Their computer has many viruses using up the computer time. In this case, you'll need an antivirus program to find and remove them. Such programs should normally be run at least once a week. You might want to consider if this one is suitable:

http://windows.microsoft.com/en-US/windows/products/security-essentials

However, avoid having two or more antivirus programs installed at once - they tend to interfere with each other.

This program is free if you have a genuine, registered version of Windows.

share|improve this answer
    
The OP specifically mentions even reinstalling the OS, so malware is not really a factor here. And having a fragmented disk or never having upgraded the RAM aren't malfunctions. –  Daniel Beck Jan 10 '12 at 20:40
add comment

Worst impact on the performance of my last machine were devices that only failed from time to time.

Two examples I remember: 1) a DVD drive with "stuck" optic - the first reads were quick but when the computer started searching around then the speed degraded.

2) too many USB devices and not enough power. All looks fine - but if by coincidence several require lots of power at once reads and writes fail and all gets slow. I hear similar problems can occur when a too energy hungry graphics card is used.

Unfortunately these problems are hard to diagnose - cause they rarely show up when you run a test :D

share|improve this answer
    
Does the USB problem only occur with older models? Or how is it related specifically to older computers? –  Daniel Beck Jan 10 '12 at 20:41
    
I would not say specifically with older models. It's just that USB also is a power supply. I have the feeling many manufacturers underestimate what people will plug in. I guess usually people don't have much devices to plug in on new machines - so this is unnoticed for some time. –  bdecaf Jan 10 '12 at 22:36
    
The point why I think this is related to the question is that the number of devices amasses over time. Maybe the opener just plugged in that single one device too much... –  bdecaf Jan 10 '12 at 22:41
add comment

I know that the question already has an answer marked as correct but, I would still like to throw in my own thoughts.

As far as the computers running slower, could it be possible you have gotten used to newer technology? I always feel like an older product has something seriously wrong with it when it goes slower compared to its newer counterpart. Our expectations have grown since their, era. Could it be that your expectations are slightly unrealisticly high?

Just a thought, depending on your circumstances (and most of them, such as the constant freezing) probably have nothing to do it.

share|improve this answer
add comment

There are basically five categories of problems:

1) Electrical problems due to hardware age -- failed capacitors, "tin whiskers", cracked PC traces, et al. These have been covered pretty well above.

2) Electro-mechanical problems, primarily in disk drives -- loss of lubrication, accumulated wear in bearings, etc.

2b) Failed fans -- power supply fans, CPU fans, video chip fans, etc

3) "Transient" problems -- oxidized printed circuit contacts, dust accumulating and blocking cooling.

4) Software problems -- viruses, corrupted data, the simple problem that new web pages are too complex for old browsers to handle, etc.

The first two are somewhat inevitable and generally can't be fixed without replacing major components. However, #2b is often relatively cheap to fix, if overheating hasn't caused permanent damage.

#3 can usually be corrected by disassembling and cleaning things (once again, if overheating hasn't caused permanent damage).

#4 obviously can be fixed, at least to a degree, by fixing the software. (Though if the system is old enough there likely is no current browser that will work with it.) (One particularly irritating problem is that M$ always arranges it such that new OSes won't network with older ones. Sometimes you can circumvent this, but beyond a point it's not worth the effort.)

share|improve this answer
add comment

Tin Whiskers

From Wikipedia Tin whiskers don't have to be airborne to damage equipment, as they are typically already growing in an environment where they can produce short circuits. Tin whiskers (accelerated by autocatalytic "tin pest" expansion) caused the failure of the Galaxy IV satellite in 1998.[4] At frequencies above 6 GHz or in fast digital circuits, tin whiskers can act like miniature antennas, affecting the circuit impedance and causing reflections. In computer disk drives they can break off and cause head crashes or bearing failures. Tin whiskers often cause failures in relays, and have been found upon examination of failed relays in nuclear power facilities.[5] Pacemakers have been recalled due to tin whiskers.[6] Research has also identified a particular failure mode for tin whiskers, where in high power components a short circuiting tin whisker is ionized into a plasma that is capable of conducting hundreds of amperes of current, massively increasing the damaging effect of the short circuit.[7] The increase in use of pure tin alloys in electronics due to the RoHS directive drove JEDEC and IPC to release a tin whisker acceptance testing standard and mitigation practices guideline intended to help manufacturers reduce the risk of tin whiskers in lead-free products.[8]

http://en.wikipedia.org/wiki/Whisker_%28metallurgy%29

share|improve this answer
    
This would be somewhat enhanced by pointing out the importance of environment regulations against lead content: leaded solder doesn't do this, but to protect the water supply we either need to eliminate it (hard) or insure the proper disposal of all electronic kit (even harder). –  dmckee Jan 6 '12 at 22:17
add comment

No-one has yet mentioned software degradation that is unconnected with hardware or environmental issues:

Most applications and operating systems have faulty code and over time corruptions can occur at driver, app, OS or other level purely because of coding mistakes.

Generally coding mistakes do not self repair, so over time these increase making the likelihood of a malfunction more and more likely. Replacing faulty code with new versions can and should help, but often upgrades and updates are not perfect, so legacy code may be left (it may sometimes conflict with the newer version).

tl;dr Entropy Wins! Computer Fails!

share|improve this answer
    
Huh? Software degrades? Could you elaborate on that? I mean it's not like the bits have a best before date... –  Daniel Beck Jan 6 '12 at 22:18
    
I elaborate on it in my answer - applications have mistakes. Mistakes can write incorrect code. Even if infrequent, this all adds up. –  Rory Alsop Jan 6 '12 at 22:33
    
You mean corruption of the data the software handles? I.e. the more processing between data input and output the more likely they won't match? –  Daniel Beck Jan 6 '12 at 22:36
    
And the application code updated by patches, updates etc. as well as accidental overwriting of app space with data. These all happen. –  Rory Alsop Jan 6 '12 at 22:41
add comment

The vast majority of problems you see are related to either heat or mechanical failure.

Both can because by the build of of dust, which blocks fans airflow and so on. This causes components become warmer then they should.

Another less common possibility is bad power. What you get from your AC outlet is not always a nice clean 120V @ 60HZ (or whatever is appropriate for your location). A computer power supply can tolerate some variability, but not always.

There can be lots of problems with software, but they can almost always be solved by a re-install. If a computer is 'locking up' with no other apparent symptoms it would be a good idea to boot of a livecd and run memtest86 and similar diagnostics.

Keeping your your computer free of dust, running it on a line-interactive UPS, and making sure it doesn't get get overly warm will lower the chance of hardware failure.

Laptops/portable systems are a special case. In is almost certain that they will be dropped or have something spilled on them or so on. Even the most cautious person will almost certainly drop it, or bang it against something. The damage of this kind always seems to be cumulative. A small drop, may not seem to hurt the system today, but that combined with heat, and be moved around and so on will almost certainly cause something to fail eventually.

share|improve this answer
add comment

Unmentioned in the existing good answers are electrolytic capacitors, which vary widely in quality and life-time and whose failure can bollox the performance of your power supply or the time constants of connections between components.

These little guys have been blamed for a number of "lemon" laptop models and high failure rates in other equipment over the years.

share|improve this answer
4  
+1 most hardware failures I've encountered have been dust/heat related, moving parts (hard-drives), or capacitor plague. Everyone mentioned the first two; you are the first to mention the last. –  BlueRaja Jan 6 '12 at 0:57
3  
+1. See en.wikipedia.org/wiki/Capacitor_plague for more informations about this damn issue. –  petrus Jan 6 '12 at 9:32
1  
@petrus: That's my second link... –  dmckee Jan 6 '12 at 16:18
    
My mainboard has "All Solid Capacitor design". Does this prevent these problems, or at least prolong the lifetime of the mainboard, or is it just marketing drivel? –  Daniel Beck Jan 6 '12 at 19:35
1  
@Daniel: Electrolytic capacitors should last for a very long time - capacitor plague is just due to large quantities of poorly-made capacitors. Unfortunately, it's still happening today, and there's really no way to tell whether a capacitor is well-made or not until it breaks. This is what happens when companies buy from the lowest bidder... –  BlueRaja Jan 6 '12 at 20:43
show 5 more comments

There are some possibilities.

Due to thermic stress on the parts, like turning the computer on and off, putting heavy usage on certain parts for short intervals etc. you might slowly damage components. Tiny cracks will build which can later damage electric connections.

They probably will make it easier for oxydation or other reactions (plastic will lose its softeners over time and crack) to get to critical parts. Pure Silicium is very prone to oxidation as far as I heard.

Another possibility is mechanical stress inside harddrives or optical drives and maybe even non stable electric input through the power supply unit.

All those are more or less speculative, but googles paper on harddrives is a good read on the thermal and mechanical topic.

share|improve this answer
add comment

protected by Daniel Beck Jan 10 '12 at 20:35

Thank you for your interest in this question. Because it has attracted low-quality answers, posting an answer now requires 10 reputation on this site.

Would you like to answer one of these unanswered questions instead?

Not the answer you're looking for? Browse other questions tagged or ask your own question.