Problem so strange I want to cry!

[peterb]

Glad you got it sorted - but don't leave it too long before changing that PSU!

[Blitzen]

Hi thanks for the comments, as it turned out it was the harddrive with the faulty SATA connection. I installed windows onto another HD on a small partiton, ran the PC on various games and benchmarks and it didnt crash after 6hrs of stress testing (And is yet to crash after running for 30hrs)

I then did the ultimtate test and managed to complete Fire & Flames on easy! (Its stupidly hard even on medium) Bunged in the faulty HDD and powered off at the same point.

In the distant future I'll be getting a new PSU - thanks for the advice & help.

If you wait until the 'distant future' i suspect you will be having to change out alot more than the PSU.

[soptom]

If you wait until the 'distant future' i suspect you will be having to change out alot more than the PSU.

To be fair, that is pure speculation, 3 pages of insisting it was a psu problem (1000000%) and it wasn't. That, unless the guy is a liar, is a fact.
As long as the guy isn't reaching the limit of the psu's power, and he could be doing this with a top name psu, it will work as reliably as any generic psu does, you have to remember, bad psu's are STILL the exception, hardly any low/mid range pc's are sold with branded psu's but they all run fine.

[Sema4]

To be fair, that is pure speculation...

Cheap PSU's weigh about as much as a box of tissues, and because of this build quality and the components used in their manufacture, they fail more often than expensive ones, and more significantly, when they do fail, they often pass 240v mains down the 12v and 5v rails, in turn destroying every single component in the computer, without exception.

Expensive PSU's on the other hand, weigh as much as a house brick, and if you are unlucky enough to have one fail, they are designed to 'fail safe' and in doing so, they commit suicide, blowing them selves to bits in order to protect the 12v, 5v and 3v rails, and equipment connected to them.

Consider it like this, if you will...

I ride motorcycles, I learned on a bike with a 125cc engine, and if you drove it flat out, it would go 70mph, poor little engine was screeming, but it did it....

Now I ride a 600cc, and it does 70mph in second gear, stick it in 6th, and it will bob along at 70 while the engine is hardly breaking a sweat, in fact it's idling.

Now consider what would happen if you drove 5000 miles at 70mph on each bike....

Which engine do you think is going to fail first ? The 125 that's running close to it's limit ? or the 600 thats practically idling ? Both engines will produce 70mph just as two 600w PSU's will produce 600w, but it's just not as simple as you think is it... far from it...

I don't know what makes you think cheap PSU's are just as reliable as expensive ones, but please believe me, you are mistaken in that belief, facts are facts, and whether you like the facts or not, you can't argue with them.

When you build and repair systems all day long, you can say with certanty, that statisically, cheap PSU's fail more often than expensive ones, and when the cheap ones fail, they kill the computer, sorry if you don't agree with that point of view, but agree or not, it's still a fact, as most people on this forum will confirm, it's anything but speculation...

[peterb]

Cheap PSU's weigh about as much as a box of tissues, and because of this build quality and the components used in their manufacture, they fail more often than expensive ones, and more significantly, when they do fail, they often pass 240v mains down the 12v and 5v rails, in turn destroying every single component in the computer, without exception.

Expensive PSU's on the other hand, weigh as much as a house brick, and if you are unlucky enough to have one fail, they are designed to 'fail safe' and in doing so, they commit suicide, blowing them selves to bits in order to protect the 12v, 5v and 3v rails, and equipment connected to them.

=====snip=====

Although I agree with 99% of the above post, in fact it would be an extremely rare failure mode of any PSU to pass direct mains down to the system power rails. All SM PSUs (even really cheap and nasty ones) have an iso;ating transformer in them, so there is a vanishingly small risk of direct crossing over of the mains suppluy.

However in the lower quality build units, as Sema4 says, there is a greater risk of destructive failure where a higher valtage may be passed to the Mobo causing damage to the associated components.

[this_is_gav]

To be fair, that is pure speculation, 3 pages of insisting it was a psu problem (1000000%) and it wasn't. That, unless the guy is a liar, is a fact.

Dude, it's going to die at some point (they all do, but I'm going to say it'll do it sooner). What separates cheapies to not so cheapies is that when cheapies die, they tend not to just just stop working. They tend to do it spectacularly. Whether that's taking out some other components or setting fire to your wall-paper (as a mate's Q-tec did) you won't know until that happens.

As long as the guy isn't reaching the limit of the psu's power, and he could be doing this with a top name psu, it will work as reliably as any generic psu does, you have to remember, bad psu's are STILL the exception, hardly any low/mid range pc's are sold with branded psu's but they all run fine.

The largest OEM of them all tends to use FSP PSU's, which while not exactly overclockers heaven, are generally very solid. Any decent PC will have something bordering on decent in it. A CWT or something. The really cheap PCs you tend to get at PC World will just have something put into them that has the right connectors, but then again they only have to power about 4watts of onboard graphics 95% of the time, so will be under far less stress than if you used such a cheapy in a gaming PC.

[soptom]

Cheap PSU's weigh about as much as a box of tissues, and because of this build quality and the components used in their manufacture, they fail more often than expensive ones, and more significantly, when they do fail, they often pass 240v mains down the 12v and 5v rails, in turn destroying every single component in the computer, without exception.

Expensive PSU's on the other hand, weigh as much as a house brick, and if you are unlucky enough to have one fail, they are designed to 'fail safe' and in doing so, they commit suicide, blowing them selves to bits in order to protect the 12v, 5v and 3v rails, and equipment connected to them.

Consider it like this, if you will...

I ride motorcycles, I learned on a bike with a 125cc engine, and if you drove it flat out, it would go 70mph, poor little engine was screeming, but it did it....

Now I ride a 600cc, and it does 70mph in second gear, stick it in 6th, and it will bob along at 70 while the engine is hardly breaking a sweat, in fact it's idling.

Now consider what would happen if you drove 5000 miles at 70mph on each bike....

Which engine do you think is going to fail first ? The 125 that's running close to it's limit ? or the 600 thats practically idling ? Both engines will produce 70mph just as two 600w PSU's will produce 600w, but it's just not as simple as you think is it... far from it...

I don't know what makes you think cheap PSU's are just as reliable as expensive ones, but please believe me, you are mistaken in that belief, facts are facts, and whether you like the facts or not, you can't argue with them.

When you build and repair systems all day long, you can say with certanty, that statisically, cheap PSU's fail more often than expensive ones, and when the cheap ones fail, they kill the computer, sorry if you don't agree with that point of view, but agree or not, it's still a fact, as most people on this forum will confirm, it's anything but speculation...

The analogy above is useless when you read the point where I said, "As long as the guy isn't reaching the limit of the psu's power" If you ride the 600 and the 125 at 30 to 40 miles an hour, they'll both be as reliable as each other. Indeed in that analogy you can argue that the 125 is made for those kind of speeds.
As for "most people on the forum", I just read 3 pages where everyone jumped on the bandwagon and said it's a psu problem, it wasn't. Those people were wrong. To now argue that the cheap psu will soon pop is irrelevant. I agree that if the guy is going to push his pc by gaming and upgrading then his choice of psu is poor, that's obvious and that is the point that an expensive psu will prove more reliable than one struggling to put out the watts, but again, if he's not pushing it, he won't have the problems that everyone suggests and reliabilty won't be an issue.

[this_is_gav]

It is pushing it. He's got an 8800GT. I'm not going to state that it will fail, because you might start crying, but it's of my personal belief that it won't last a year with such a card. To put it another way, it's going to dramatically decrease the life of the PSU, and that couldn't really be said of a good, solid PSU, though of course they do die too, and the death will be brought forward the heavier it is loaded, but not to such an extent.

Read the HEXUS PSU Roundup from 2005 if you want to understand why - the Qtec one on page 19 is particularly enlightening.

[soptom]

It isI'm not going to state that it will fail, because you might start crying.

How old are you, you join in a thread where the majority of people are later proven wrong in their diagnosis, so you then state that he'll go on to have a problem anyway I'm guessing to deflect away from said fact and then you post "you might start crying".
You need to grow up and stop acting like a child.

[this_is_gav]

I apologise for your not being able detect a little tongue-in-cheek comment (wasn't blatantly obvious to be fair, though I did mean it as a dig too I guess).

I'm 28 and work as a technician. I deal with hundreds of different systems continually and do support for community members too. Don't question my credentials or validity because you don't agree with what or how I post.

I've read the thread from when it was first posted, and I still maintain it will die if not now, then in the not-so-distant future, and he will regret it. I don't have any motives for posting other than in the best interests of the original poster and others that read this, as 99% of the forum do. I'm sure anyone who's worked in the technical aspect of IT for any length time will go on to say the same. Because it wasn't the PSU this time doesn't mean won't be the PSU the next time.

[Sprite]

I think he get's the PSU bit now...

[PinkPig]

Since this has become a debate on whether an expensive PSU is worth the money, I'm going to take the middle ground. Some people seem to suggest that a cheap power supply will "explode" or cause a fire, and certainly make any system unusable. That's very certainly not the case - I have not heard a single documented case of a fire or explosion or anything of the sort resulting from a cheap PSU. Anything for sale will meet relevant electrical safety standards.

Similarly, lots of computers work perfectly well for a long time on "cheap" power supplies. Few people on an enthusiast forum would be satisfied with the PSU included in most OEM machines yet few are completely unreliable and none start fires! Lots of people buy and use cheap units and are perfectly happy with them.

That said, if you're on the hunt for a very reliable system, if you want to overclock or if you're buying very expensive and power-hungry components elsewhere then of course a well built (and relatively expensive) PSU is a very good idea.

[soptom]

I apologise for your not being able detect a little tongue-in-cheek comment (wasn't blatantly obvious to be fair, though I did mean it as a dig too I guess).

I'm 28 and work as a technician. I deal with hundreds of different systems continually and do support for community members too. Don't question my credentials or validity because you don't agree with what or how I post.

I've read the thread from when it was first posted, and I still maintain it will die if not now, then in the not-so-distant future, and he will regret it. I don't have any motives for posting other than in the best interests of the original poster and others that read this, as 99% of the forum do. I'm sure anyone who's worked in the technical aspect of IT for any length time will go on to say the same. Because it wasn't the PSU this time doesn't mean won't be the PSU the next time.

That's fine, I wasn't questioning anyones credentials and their motives are irrelevant, what I take issue with is people being 100000% certain that this guys problem was caused by the psu, if this guy had taken everyones recommendation, he would have spent more money and still had his original problem. Now I'm not saying that his cheap psu will last as long as an expensive one, but I will say that it is UNLIKELY that it will catch fire or will destroy his pc as has been stated previously. As long as he is aware that his psu won't like it if it is running at it's nominal top whack all day long, it should last him fine. PinkPig has managed to put it succinctly above.

[this_is_gav]

Since this has become a debate on whether an expensive PSU is worth the money, I'm going to take the middle ground. Some people seem to suggest that a cheap power supply will "explode" or cause a fire, and certainly make any system unusable. That's very certainly not the case - I have not heard a single documented case of a fire or explosion or anything of the sort resulting from a cheap PSU. Anything for sale will meet relevant electrical safety standards.

Of course it seldom happens, but it does. I've seen the aftermath as it happened to a mates QTec, as I stated above. I expect he'll still have a photo he took too. It may meet standards, but it didn't stop a line of MSI motherboards from randomly catching fire around 5 years ago (//edit: actually it might have been Gigabyte) - again far from a given, but not isolated to individual cases either.

Similarly, lots of computers work perfectly well for a long time on "cheap" power supplies. Few people on an enthusiast forum would be satisfied with the PSU included in most OEM machines yet few are completely unreliable and none start fires! Lots of people buy and use cheap units and are perfectly happy with them.

I said exactly that a couple of posts back. It doesn't automatically mean BANG! but it's going to happen sooner rather than later, and for the reasons those Hexus tests showed.

That's fine, I wasn't questioning anyones credentials and their motives are irrelevant, what I take issue with is people being 100000% certain that this guys problem was caused by the psu, if this guy had taken everyones recommendation, he would have spent more money and still had his original problem. Now I'm not saying that his cheap psu will last as long as an expensive one, but I will say that it is UNLIKELY that it will catch fire or will destroy his pc as has been stated previously. As long as he is aware that his psu won't like it if it is running at it's nominal top whack all day long, it should last him fine. PinkPig has managed to put it succinctly above.

I never once blamed the PSU (post #38 was my first in the thread) but I will admit that when I was reading through the thread earlier, I too thought it was caused by the PSU until he tested his 2nd PSU - everything pointed to it. I don't think anyone could have guessed the real culprit, and to be fair I'm still trying to get my head around that one, as it makes absolutely no sense to me.

I certainly doubt it will catch fire, and it's far from guaranteed to destroy any components if it does let go, but I don't think it's going to last while it's powering a C2D, 2 hard drives, presumably 2x1gb RAM and a 8800GT, but there is a significant risk that it will take a hard drive/CPU/RAM/mobo with it if it does. It's certainly not worth running the risk.

For the record I have a good percentage of machines at work using an OEM PSU, and my HTPC too, but they're under no load whatsoever. I'd swap them out whatsoever if I was to load up some higher demand 3D apps.

[killie99]

blimey, hasn't this thread turned into a

Well just to stoke the fire a bit more, in 22 years of repairing computers I've never seen this problem caused by a hard drive or flakey connector.

[Sema4]

In response to those people here (you know who you are) that make statements like 'I have never seen a PSU go bang' and 'I have never seen a cheap PSU damage a motherboard', I would just like to say in response, 'I have seen them go bang', and 'I have also seen them take out every single component connected to them'.

Just because you have never seen it, don't proclaim that it's not possible to happen because you say so, it can and does happen!

Thats like saying Australia doesn't exists because you've never seen it, you've only ever heard other people talk about it and you don't believe, or agree with them !

And soptom, you can twist my analogy if you like, and say that at 40mph, a 125cc engine will be as reliable as a 600, but I'm afraid thats not the case either, you just don't seem to get it... a 125cc engine will have to work a lot harder to do the same work a 600cc engine does, regardless whether it's well within it's operating tolerances or not, the fact is a 125 will still have to work harder to produce the same output, and as a result of that, the 600cc engine will still be running when the 125 fails. (You show me a 125 with 100'000 miles on the clock, there are plenty of 600's that do)

The same goes for PSU's, if you take two units rated at 600w, yes they can both produce 600w, but the cheap one will have to work much harder to do it because it's comparatively 'inefficient' at what it does, and so has to work harder to produce the same output as a better quality unit, which can produce the same output without trying very hard at all.

Anyways, I'm done trying to help you understand your misconception, you go around telling everyone cheap PSU's are fine if you like, it's all the more money for me when they blow up

@killie99, ditto, although I have 1 year less in the business than you I have never seen a problem like this caused by a flaky hard drive either, I did suggest he check the event log for errors relating to the hard drive/ disk controller, in an earlier post I made, but it seems to have been ignored, and if it was the hard drive, Im sure the event log would be full of red X's... :shrug: Oh well, horse to water and all that... I'm done with this thread, as much as you can offer people the wisdom of experience, it seems you simply can't make them understand...