My computer, oh my lovely gaming machine, has failed! It won't work at all - All help appreciated

Recommended Videos

SnipErlite

New member
Aug 16, 2009
3,147
0
0
-AC80- said:
1. attempt to enter your BIOS (please google for more info) look around temps, and you boot options.
2.if 1 fails open case and clean including the CPU and GPU coolers
3. if 2 is clean check RAM if it is securly in
4. if 3 is secure attempt a boot up off another HDD from a diffrent computer
5. if 4 fails you have isolated it to be your defective machine and NOT you HDD, now comes the expensive bit:
5.1, your cpu is fine and you GPU is fine and your RAM.
5.2, you need to cheak you PSU
5.3, your mobo MAY have a problem but dont rule it out (transisors and capasitors)
5.4, check your cpu cooler is it on right?
5.5, check all cable-ling
5.6, check BIOS for faults or any possible upgrades available online
6. if everything is above in 5.2 to 5.6 is ok, test your HDD then the back up if that still fails please see 6.
6. GIVE UP and take it to a pro/scrap it.
edit:
SnipErlite said:
Picture the scene: I turn on my computer today hoping for another good evening of gaming, and around a minute after I do so, just as I'm logging on, it just shuts off randomly. Check the power light and it's flashing amber, instead of a nice solid green. Uh oh.

This happens every time I switch it on (along with a message saying files on my C: drive have to be checked, checking right now blah blah blah....this happens now and then and ordinarily it checks all the files and we go on as normal so I'm not worried about that).

Have a quick google and various answers pop up. Also the Dell website (it's a Dell Dimension 8400) suggests having a look at the diagnostic lights on the back. These 4 lights apparently can help you work out what the fault is depending on what combination of green and yellow they display. However while booting up they flicker between green and yellow really fast. Then decide on steady green on all 4 (which indicates there's no fault) about 20 seconds before the whole thing fails.

Now my main question is, does anyone have any idea what's wrong with my beloved PC? Has anyone else had this problem and found a fix?

Has..........has my favorite gaming system finally broken?

=(
after re-reading this and hearing that you managed to enter your OS means that it is most likely a temp issue/precaution you bios has shut you down because a component is overheating, due to the shout amount of time you spent since pressing the on button. Have you moved you computer around recently or heard a wired noise and is it cooled by the intel stock cooler ps that is THE DEVIL to overclokers. this is the most likly error but it could be others.

ps every one remember this check-list above it may help YOU!
I'll go through those steps in the next few days. Thank you.

Also nah I haven't moved it about or anything, this was completely unexpected. One day it works perfectly fine, the next this. Also as vf501 said no I'm not overclocking.



tahrey said:
1. It's a Dell...
2. That mention of the disk check thing happening "now and again" is NOT a good sign. I think that's happened a grand total of maybe once... twice? in the ~4 years I've had the laptop I'm typing on, both times after what amounted to collosal handling or driver upgrade fuckups (one other or both). NT flavours of windows just don't suffer the problems that cause that reaction anywhere near as much as 9x did unless there's something inherently broken with them. A healthy computer in normal use I'd expect to NEVER see it.

Things that are typical of causing this sort of thing:

* Dodgy memory, particularly if it's in a part of the map where windows likes to load into. It goes there, gets corrupted, and poof - lockup.

* Bad hard disk, whether from manufacture or mistreatment. Some of those custom cases put the disk in ludicrous positions that are ripe for it badly overheating. And brand PCs like that often use cheap and nasty disks, like the lower-end Western Digitals or Seagates (as opposed to, say, a Barracuda) that appear to be made out of wadded-up tissue paper.

* CPU fan failing or CPU-heatsink assembly becoming detached. The latter shouldn't be common any more given how well-secured they tend to be, and the former shouldn't give all so much trouble, though. OLD processors would have just smoked and died. Certain previous generations progressed to having safety cut outs that killed the power at a certain temperature (before it could enter "thermal runaway", ie Halt and Catch Fire). More modern and sophisticated versions of this step the processing speed way down as the temperature approaches critical. You may end up only running at like 200mhz instead of 3200, but it will still sort of work (instead of crashing/resetting) and won't cook itself to death.

* Iffy power supply. If that starts malfunctioning, particularly under load, the rest of the machine also gets sick. (including suffering errors that could be suspected as memory or disk at first, as those parts could be directly affected)

* Pernicious malware buggering your files up bad enough to require the disk check then make the machine hang.

Anywayup the diagnosis is going to require time, confidence with computer innards, spares and experimentation. I'd start gently by downloading and making use of memtest86+. Watch it for the first few minutes, then le it run at least an hour. If the machine runs that long without crashing, that's a good start; without crashing or any errors being flagged, then it's probably not the RAM.
If it does give errors, try removing a stick (or bank) at a time and seeing where the problem does and doesnt occur. And as MT86+ runs off external storage, you can probably discount hard disk problems if it DOES hang/reset during the tests. Unplug the disk completely to be sure. You can make sure the CPU fan's spinning at the same time.

Whip the disk out and put it in a USB enclosure or use one of those fancy disk-dock thingies to run a full virus and malware scan on it unsing a different machine.

In fact just try booting it at all off any suitable external media... crash = its not the disk that's at fault... and now i'm literally falling asleep at keyboard so this needs to hand off to someone else.
Well that disk check thing never seems to coincide with any change I've made on the computer or any new software or hardware. It just appears occasionally and has never caused a problem.

Also I would download memtest86+ only I can't even logon, let alone download something. If I exist the disk test (not recommended apparently but oh well) then I can logon for about 5 seconds. Then *poof* it dies.
 

SnipErlite

New member
Aug 16, 2009
3,147
0
0
Oh, also if i enter setup (F2 during boot) and leave it, it still fails after about a minute. So it isn't being caused by something processor-intensive. I'm thinking it's unlikely the processor or RAM.
 

Nimbus

Token Irish Guy
Oct 22, 2008
2,162
0
0
stinkychops said:
Nimbus said:
Again, sorry to hijack, but the poeple in this thread seem to know what they are talking about.

I'm running a system with an i7 920, a GTX295, 4GB RAM, and a couple of HDDs. Could I get an estimate as to what kind of PSU I should get? I've currently got an antec Truepower 850, but apparently I need a new one.

Nimbus's Wallet: Oh God! Not again! Nooooooooo!
850 would do it. Don't go over 1000 thats for sure, a waste of money. The best thing to do is go for ones with good rails. You have a list of options or open to all recommendations?
I've only ever bough the one PSU, so I don't really have much experience. I'm open to all suggestions.
 

SnipErlite

New member
Aug 16, 2009
3,147
0
0
stinkychops said:
4 years? Your PSU's life might have run out. If you're going to troubleshoot, swap that out first, then mobo. Make sure to buy one with 500+ watts and a decent name antec I recommend.
Indeed - I swapped it out for a 650W PSU and everything is now fine.

So yeah, turns out it was just the PSU failing. Thank you everyone for your help, I've got my lovely lovely PC back ^_^
 

Sebenko

New member
Dec 23, 2008
2,531
0
0
Nimbus said:
Again, sorry to hijack, but the poeple in this thread seem to know what they are talking about.

I'm running a system with an i7 920, a GTX295, 4GB RAM, and a couple of HDDs. Could I get an estimate as to what kind of PSU I should get? I've currently got an antec Truepower 850, but apparently I need a new one.

Nimbus's Wallet: Oh God! Not again! Nooooooooo!
None, just send it to me. Or, y'know, a quality 1Kw. that'll power anything. Probably overkill though. if you have 850w already, that should work. If you need a new one, just by another one of those.
 

tahrey

New member
Sep 18, 2009
1,124
0
0
Good to see you've got it working again, and it was the dreaded PSU thing rather than anything difficult to solve. Hopefully is was just marginal rather than completely ruined, so it won't have had lasting knock-on effects on your other hardware!

I wouldn't go overboard on the PSU front though. Just buying the biggest one you can get is like shoving the largest engine you can find into your car. It doesn't solve all problems, can be expensive, and if it doesn't cause any other actual issues it will at least waste power. Computer power supplies are most efficient at something like 50-80% load, and still quite good when nudging their limit, but get rapidly less efficient towards 0% load. So if you originally had a 200w that either just randomly died, or was overloaded because you occasionally demanded a peak 250w of it, and you replace it with a 1000w (or even that 650)... well it'll certainly have a pampered life, but you'll be using a fair bit more wall power than you previously did just to get the same output to the motherboard. Particularly when it's idling, as that'll be even further down the curve.

This is ameliorated somewhat if you get an "80 plus" supply (particularly one with a "medal" rating) and are planning on adding lots of extra stuff to the system, but you'll still get more benefit if you match it up. What was the rating of the one you took out, and had you done any serious, potentially power-sucking upgrades? If it's a company-built one which you haven't altered, one of the same rating should be fine... just use a better brand than they did.

Slimline, eco/budget systems these days can be 150w or less, particularly those partly based on laptop components. You only need much above 3-400w if you're doing heavy gaming stuff, e.g. a graphics card with its own power plug. Or running something along the line of Nimbus :D

Also when I was saying "download this 'n' that", I DID sort of mean use a DIFFERENT computer for it! Do it at work, borrow a friend's, etc (MT86 is less than a meg - it can even be installed on a floppy - so it's not going to abuse their connection... and it's useful enough in a general fashion, even just for benchmarking, that I suggest you download it and make a bootable disk RIGHT NOW whilst your system is working ok).

Same with the putting the disk in an external enclosure for testing - the idea there is that you get the target system booted and the best possible antivirus setup already running, disable autorun, THEN plug in your disembodied hard drive for checking. You isolate everything as much as possible, and certainly don't try to go online with your PC if it's acting suspiciously.
 

tahrey

New member
Sep 18, 2009
1,124
0
0
If you have the money for it, Compar... many of us don't. Components are a lot cheaper and a lot less bother to replace than whole systems.