Posted: 7/28/2006 10:10:10 AM EDT
|
How does a computer "wear out?" As in the hard drive and all that jazz. Does the silicon actually break down in the chips until it finally dies or what? It is a 2001 Gateway, replaced the modem [lightning] as dead, and it would not boot up [no power up] Its up and running ut the repair facility said the memory and drive is starting to go and its just cheaper to get a new computer [not really but better off in the long run actually]. I don't doubt that but just curious as to why one begins to break down and how. Is it the heat? I would suppose so but why does it fail in that manner? Thanks. |
Same could be said for video cards, and processor fans etc... I don't even know if in a newer say dell, if the processor fan quit, would it get hot enough to do damage? Not oc'ing of course. |
Yeah it could. |
|
depending on how old the system is,it can exceed the value of the system to be repaired. all i see, all day long is bad harddrives, memory, powersupplies, and motherboards. can be from lightning, power outage. or just something went bad. with how cheap newer computers are its just not worth getting some older ones fixed |
|
Yeah things fail over time. Also remember where your parts come from. I forget when it happened but there was a tsunami that whacked Singapore about 3-4 years ago. Memory prices, and other pieces that used circuit boards, went through the roof. Most manufacturers that make our cheap home pc parts get their circuit boards overseas. Heat is the #1 killer of circuits over time. That's why computer rooms are freezing cold. Also you have a 2001 pc. Is it a 800 Mhrz PIII or around there? I found that unless I scavenge old PC's that our work throws away, buying the right technology is more expensive than current technology because most of the companies producing parts are making more of the current technology. Thus the cheaper to buy new. That being said I have a Pentium 90 running my VM system in the basement. I have about 10 spare old PC's sitting next to it for spare parts. I can't imagine how much a new AT power supply would cost. I am sure they are out there, but it's easier to go to my scrap pile. |
| Aside from capacitors blowing, static electricity, power surges and moving part failures, there is also electromigration. This puts a finite lifespan on ICs. Lifespan actually decreases as transistor size decreases, so CPUs end up dying within a decade or so. They are only rated for so many thousand hours of operation. |
|
most logic problem are due to overheating and sloppy voltage <power supplies> Microprocessor and their connecting circuits are VERY fine circuits. Over voltages cause excess heat which causes the circuit to expand and contract beyond it's designed tolerances. eventually this causes a short in the board. As PC board repair is not economical in the field most PC circuits are now considered disposable. 99% of hdisk problems are failed motors that spin the platters. Newer drives tend to be less prone to platter damage issues and generally the motors die before the platters. Failure to achieve proper speed is the most common. the best way to insure long life on a pc is a high quality power supply, routinely clean all fans and heat sinks every 3-4 months and keep ambient temperture as cool as possible. <cool air outside means cooler air inside> when you have added what you beleive to be adequate fans--- add one more. |
not only would it die it would likely get hot enough to bubble the laminate off the cpu. most causes od memmory death are genrelly associated with impoper ESD precuations on installation. minute electrical shorts caused by general handling can cause damage. as this type of damage tends to compound over time with voltage irregularities it may not manifest for a long time. but the root failure cause can generally be traced to poor ESD precations. |
| for you guys with older machines, if you look at the system planars you will notice they are considrably thicker than newer boards. the older boards used a bit different technology for manufacture that required thicker/heavier circuits. that equals less heat buildup and stronger circuitry. |
Yup. Lack of OS support is what will have them retired since Solaris 10 dropped support for the 32 bit cpu. |
most newer computers have settings in bios and sensors that would cause it to power off before frying the cpu but if those safeguards failed or were disabled and the cpu fan failed it would very like overheat to the point of damage very fast |
actually a 2ghz processor in a dell will begin to smoke after about 26 seconds and the laminate sheets on the cpu will seprate and flame after about 45 seconds. don't ask me how i know this. mike |
|
Capacitors also have a lifetime that is effected by ripple currents. Charging and discharging them causes them to wear out. Heat breaks them down. On older stuff the parts were bigger and stronger and put out less heat so thermal forces were not as great. RAM, flash, silicone, all get electrical spikes. Burn them out. Flash wears out after 100,000 rewrites (this number varies alot). Alot of devices will last practically forever if they aren't mechanical or subjected beyond their tolerances. -Foxxz |
|
The hard drive, being a moving part can and will wear out. Chips can go and capacitors can fail but is less common. I have seen memory fail over time too, but it's really rare. Because everyone here is giving old hardware stories. I have a Cobalt Raq 2 (now owned by Sun and defuncted(sic)) that I got in '99, its been running 24/7 save a few power outages. The one fan inside the computer failed, so I removed it and the Raq has been running fine for the last four years that way. The Raq has a 6gb Hard Drive and 128mb memory. It's a perfectly good web-server that runs off a 40 watt power supply. All hail the mips processor. -JTP |
|
The moving parts of hard drives wear out. And Tin whiskers kill circuit boards by shorting out the chips. Tons of links explaining them on Google. NASA's tin whiskers pages |
|
After two years of running the platters of a hard drive might have gone around over seven billion times. The bearings just wear out. Then there’s something called “head slap,” where the read/write heads actually touch the disk while it’s spinning. (They are supposed to float over the disk at a distance far, far less than the thickness of a human hair.) This can damage the disk surface… and the microscopic debris from this collision starts gradually chewing up the surface of the disk. Mechanical parts like fans and CD drives just wear out, dust and cigarette smoke speed this process along. Thermal expansion and contraction of circuit boards can cause marginal solder joints/parts to go bad. I believe Quantum had a problem with some of their 4.0 GB drives doing this about 8 years ago. It was just a batch of faulty resistors that would fail after heat cycling a few dozen times. Worst part was the drives just gave CRC errors at random when they failed, so the problem was very hard to diagnose. Electrical surges and static electricity can cause parts to fail. Worst part is they can weaken a part and cause it to fail a month or two after the damage is done. Corrosion/tarnish can cause memory errors and failures of PCI cards too. Cleaning the contacts usually fixes these problems. Extreme heat can damage chips too. Dust acts like an insulator and can make computers overheat. And nothing lasts forever. Parts just fail sometimes. However, new parts are more likely to fail than older parts because new parts might have hidden defects while old parts probably don’t. Once a part has been burned in for a year or so it will probably keep chugging along indefinitely unless one of the above things damages the chips. (I have a couple of working calculators from the 1970s.) |