Parts designed to die?

D

Dario D.

Gawd
Joined
Dec 8, 2004
Messages
582
Hi all.

This may sound a little vague, but...

I was at the computer store a few weeks ago, and saw this mobo for sale (I can't for the life of me remember which one it was. Maybe Asus or something) that said something on front of the box about having "long life parts" or something to that effect. Basically, it was boasting having some certain type of part in it that wasn't prone to dying, I think it was...

This made me think. I'm very used to chips dying on me after about 3-5 years (not all, but more than I'm comfortable with), and was wondering if some chips are perhaps designed with with certain parts that are expected to fail after X amount of time.

I know hard drives and CPU's have a certain lifetime expectancy (is it like 5 years or so?), but most of the things I'm thinking of don't have moving parts, don't burn like lava, and don't have an obvious excuse for dying that makes a whole lot of sense to me.

Example: what do they design the space shuttle, space stations, and satellites with? I would assume that their chips aren't prone to failing after 5 years... In fact, I would assume that they aren't prone to failing at all ("Uhh, Houston, our hard drive just died.") and yet half the crap in my computer has either already died on me (over the years, here's what I've lost: 1 sound card, 1 video card, more fans than I can count, 2 motherboards, 1 hard drive, 2 CD roms), or can probably be expected to die a sudden and mysterious death one day in the near future.
-

Anyway, I don't want to sound like a conspiracy theorist. If they just design consumer hardware with sucky standards, that's nothing new. I'm just wondering if anyone knows whether or not some of these chips we're using are specifically designed to fail or not, or if they're purposely using weaker, cheaper materials that are known to deteriorate. That would NOT surprise me, so that we have to buy new stock every couple years.

...would help if I knew what that board was.
 
Was probably advertising solid capacitors.

Gigabyte's been calling their boards "Ultra Durable" since they started using all solid caps.
 
Hmm, could be. Google-searching still won't reveal the box I saw though. :(

Here's something of note from this article on solid capacitors [here].
"without fear of excessive capacitor wear or exploding capacitors."
Click to expand...

It's quite obvious that overclocking can potentially overheat stuff, and I guess wear on capacitors... but what I'm wondering is if moderate, average heat can also have a slow decaying effect on capacitors (or other things).

I'm normally not one to speculate this kind of thing, but I'd like to be able to isolate what aspect of chip design is causing most of the random 3-5-year failures we see out there.
 
MTBF = Mean time between failures...all PC parts have this number...It's usually advertised moreso on HDs or fans but they are usually in the neighborhood of 50,000 - 100,000 hours...that translates to about 5 - 10 years but every part has their own MTBF and this is just an average, some parts last longer, other less.

It all depends on how it was treated during its' life as to when it will fail. I've got a 10 year old PC running at home with no problems.....*crosses fingers*
 
whether or not some of these chips we're using are specifically designed to fail or not, or if they're purposely using weaker, cheaper materials that are known to deteriorate.
Click to expand...

In general no, specifically concerning CPUs, absolutely not ! Concerning other ICs used in electronics, absolutely not !

Now discrete components, like capacitors, resistors, diodes etc typically have a very long life if the circuit they are designed into is done properly. Transformers and Inductors will last forever, ditto typical resistors. Capacitors are the biggest issue by nature of their construction. They do have a limited life (typically +10 years or so for a good quality electrolytic) depending of the quality of construction and the stress the circuit they are installed in places on them. (Note there are capacitors that can last much much longer but I am restricting my comments to a typical PC motherboard. )

Heat kills electronics, but an IC in a properly designed circuit and kept within the manufacturers operating parameters should work for years and years and years.

As with anything, cars for example, you get what you pay for and how you maintain and care for it probably has a greater affect of life than the initial build quality.

The majority of failures we incur are due to
1) excess heat
2) excess voltage, includes spikes/surges on the commercial AC.
3) Poor quality parts. (cheap)
4) engineering errors or cost constraints (poor design) causing items 1 2 3 and leading to failure.
5) Actual failure of the component for various reasons.


Military and Aerospace ICs are typically the same design as commercial parts with better packaging (gold leads, better bonding, ceramic case, wider operating temp specifications)
and testing. Basically they are the "cream" of the run skimmed off during basic device testing and given better packaging and additional stress testing at the factory. Of course there are specialty parts way over engineered made specifically for those applications but in general the same base design is used and put in a stronger "box" and tested more rigorously.


All that said, crap is crap and there is plenty of it out there as it is a sad fact that most people have no clue and the bottom line price is all that matters. For example if you go buy a TV you look at the picture and the price. Who looks to see how well the guts are designed ? And even if you tried you probally could not find the details you need, even so called "reviews" are horrible about providing technical details because you need an engineer or at least someone with some technical training to be able to give an opinion IF the tech data from the manufacturer was made available, its not. /start rant Look at the specifications of any motherboard, it is a list of features, some marking BS like you noticed, sometimes based on some mangled engineering information, and nothing else. Only recently (and a few notable sites in the past) does a motherboard review bother to talk about the VRM, power circuits for the memory or support chips, and God forbid anyone should mention the master clock chip and who makes it. One reason we have boards that allow you to set a FSB of 500+ MHz when there is no way in Heaven or Hell the board could ever do more than 450 due to basic design limitations. And the suckers continue to shell out. /end rant

I'm very used to chips dying on me after about 3-5 years
Click to expand...

If this is true, you are doing something very wrong or need to spend a bit more.

I can build you a motherboard that will last 1000 years, but it will cost you a million dollars and be the size of a full sheet of plywood.
(and be technically obsolete in 2 years or less )
 
Manufacturers are always working on new stuff to make their current products obsolete, they don't usually try to make them die.

Cheap hardware usually is built with cheap components which don't last as long. Computer hardware with moving parts usually have shorter life spans than solid-state hardware. This is why backing up hard drives is so important and why ultra cheap optical drives always seem to fail before anything else.
 
Of course they make the products die after a lifespan, just like what manufacturers do with cars. Products lasting too long was the cause of the first depression.
 
Stuff on airplanes and shuttles also fails, but they RAIETM ( redundant array of everything that matters).
 
OP: It really matters how you take care of your PC, like good cooling, power, etc. I've got several old PC's that are still chugging away, and that's when they didn't build motherboards with solid caps. Although admittedly, I don't think I have any 10+ year old computers with their original hard drive.

Robert
 
Well-made motherboards should last you for quite a while.

I bought a Gigabyte GA-7DXR+, from Newegg back in 2002, from their "Refurbished" pile, and it was still working just fine as of last month (when I finally upgraded my work PC). I'm going to use this board (in addition to the Athlon XP 2400+ attached to it) to make a FTP server for the lab.

Well-made hard drives should also last you for many years. I still have this Fujitsu 6.4 gigabyte drive that's been humming along since 2000.

Well-made optical drives WILL last you many years. On the computer next to me is my old Plextor Plexwriter 4X SCSI-based CDRW drives, from 2000, as well as my old Plextor UltraPlex 40X CD-ROM drive. Both still work flawlessly, and I am not hesitant to make mission critical backups using the old Plextor drives.
 
My M3A has two non functional ram slots, an onboard LAN that makes it hang on boot, and I think the raid controller is broken. I thought my drive was dying but as soon as I took it out of raid it seems to be working fine (would randomly power off, and then not be detected 50% of the time on startup).
Awesome piece of shit Asus.
Why havent I RMAd it? 2 weeks unable to game + 7 or 8 dollars to trade my deffective bought as new product for a refurb.
 
Dario D. said:
This made me think. I'm very used to chips dying on me after about 3-5 years (not all, but more than I'm comfortable with), and was wondering if some chips are perhaps designed with with certain parts that are expected to fail after X amount of time.
Click to expand...

If your stuff dies that fast, then you have some other issues as others have said. Many of us that run Linux firewalls are running them on really old hardware. I don't think I have ever used a machine for a firewall that was less than 5 years old when I got it. I know a few people that are using 486 machines for firewalls from the early 90s without any issues. These are machines that are running 24x7 and need to be 100% reliable. With proper cooling and cleaning the vast majority of computers should easily last 10+ years.*

*This statement excludes moving parts. Since moving parts = friction and friction = wear these things tend to fail.

Most machines that I see that have failed components usually have never been cleaned so they have no air circulation or had one of important fans fail causing the system to chronically overheat and cook itself to death.
 
everything wears out, it's just a fact. They don't need to design parts to die because the upgrade cycle of a lot of users is around or before 5 years. Enthusiasts upgrade much sooner than that. If it was ever found out that a company was doing this there would be a for sure class action law suit. If a part dies within a reasonably short time it's usually under warrenty. I've never had to worry about it because aside from the odd part every once in a while I do a major upgrade before the good parts wear out.
Then I do random things with the older systems. I have a PII 255mhz gateway, 2.5GB HD,64MB ram and old ass voodoo running next to me right now.
 
Most computer component failure is due to a chain of failures usually stemming from electrical spikes in the line voltage. You may think you are getting 110-120 VAC when in-fact it can range from 80 to 150 VAC which can cause quite a bit of stress on parts. The under voltage side might result in lack of power to make the HD armature movements which can cause the read/write head to miss target and write to the wrong spot resulting in data corruption. On the high voltage side it can cause a chip to run out of its spec'd voltage range and burn up.

If you really want a computer to last for damn near ever do/get these things;

High quality PSU, one that can hold a good voltage tolerance and has proper cooling.

High quality UPS, that short flicker on/off of your power can cause a ton of damage if it happens at the wrong time like a BIOS flash.

High quality voltage conditioner unit, if you know for sure that your incoming line voltage fluctuates more then 10VAC off of 115VAC then you really should get one of these for expensive electronics, people who live in the sticks like I use to etc. You can tell if you have this problem by simple observation of regular incandescent lights in your house, if they go dim often you have a problem.

Computers generate heat, make sure that the heat can be dissipated, good fans and case are not all of it, you can have the best flowing case and fans in the world but if the rig is pushed up against the wall to where the rear vents are blocked well your not doing yourself any favors.

Everything is designed with an estimated lifespan, it is one of the primary fundamental parts of engineering, when you set out to design and build a product you target a lifespan and spec the materials on that. Sure some things never wear out but those are few and far between, most products are set for a specific life span; cars ~ 8-12 years, washer/dryer ~ 15-20 years, houses ~ 80-120 years, oil refinery's ~ 20 years(5 year seal and valve replacement, my field).

I have a P2 Xeon server that is pushing 10 years old that still runs great as a hardware/software firewall and file server, it has sat behind a power conditioner and UPS with shutdown command control for long duration power outages. In the 10 years of this setup the only thing I have had to replace was the primary HD, I have added other HD into it as time has gone by and replaced the 30gb drives with larger 200gb ones.
 
I'll give my opinion based on my own experience. **DUST**, is the number 1 culprit. Which causes heat build up and more stress on all the components inside the pc case.

My first PC was a cheap $300 compaq, I bought back in 2000. I knew nothing about pc's and how to maintain them back then. After about 3 years I started having problems with it, random shut downs, instability, running slow.

So I decided to take it apart, figured I had nothing top loose, the dam thing was practically unusable anyway. It looked a bit dusty at first glance so I blew it out, Put the side cover back on, still ran like shit.

So I took the cover off again and shined a flash light in it to take a better look and noticed dust caked under the cpu fan, in the cpu fins. I took the fan off and was able to pry the caked dust out of the cpu fins.

Then I took a better look at the power supply with the flash light and noticed that the inside of the power supply was completely full of dust.

I took the power supply apart and was shocked at how much dust was caked in side of it. I carefully cleaned all the dust out of it and put everything back to together. This old Compaq pc with the all the same hardware including the original 95 watt power supply is still running today.

I learned a valuable lesson and since then I give all my pc's a thorough cleaning once a year. And that includes taking the power supply apart to make sure it is free of dust.

I have never had a mother board failure or a power supply failure on any of my own pc's yet. The only failures I have had is memory and cd/dvd drives. The low failure rate of major components in large part is due to keeping the innards of the pc's free of any major dust build up.

That and a good surge protector should keep your pc running for years and years.
 
I wouldn't put too much stock into anything that Asus puts on their boxes anyway... hell for the last couple of years they have been putting "Heart Touching" on their boxes. :D
 
Valnar said:
OP: It really matters how you take care of your PC, like good cooling, power, etc. I've got several old PC's that are still chugging away, and that's when they didn't build motherboards with solid caps. Although admittedly, I don't think I have any 10+ year old computers with their original hard drive.

Robert
Click to expand...

I've got an IBM AT down at work (original 5170 model from 1984) with a 20MB full height hard drive........ last time I booted it up (beginning of the year) the 24-year old workhorse booted DOS 3.3 with no problems............ Last production use of this machine was as a print server circa 2001............
 
Companies do these things with cost in mind. Sure they can put in ultra reliable components, increase stress testing and such, but then they only see their profits down the drain. They want to maximize their profits by producing you a cheap component with a reasonable expectancy.

As for your NASA point, NASA pays millions per computer, billions per satellite, so you can see where this concept comes into play. Also, NASA transports all their electronics and handles them in clean rooms with utmost precision, the environment in which it operates is also perfected... not like home where you have dust, dirty power, temperature changes, humidity changes, etc. Also, you do more on your machine than they do on theirs (theirs are niche machines for dedicated tasks).
 
Ockie said:
Companies do these things with cost in mind. Sure they can put in ultra reliable components, increase stress testing and such, but then they only see their profits down the drain. They want to maximize their profits by producing you a cheap component with a reasonable expectancy.
Click to expand...

'Tis true.


However, I don't mind that companies do this. I have the utmost respect for manufacturers that build both a cheap product and high-end product, thus giving the buyer the choice. It's when they *only* build a cheap product that it irks me. I'd rather pay a bit more upfront, thus giving the manufacturer the money they want, in order to get something of quality that will last longer. It seems they are doing that more these days with higher end capacitors and such. Of course, only time will tell.

Robert
 
You must log in or register to reply here.
Back
Top