computer locking up

[PopFly]

I am running a AMD Sempron 3000+ processor and one of those Gigabytpe Triton Mother boards. The Video card is a GeForce 6600 LE. When I load the computer everything is fine and it plays correctly. However, after a few minutes the computer will give me a black screen for a split second before freezing up on me completely. I have looked and I thought it was perhaps an AGP driver that needs to be updated, I really have no clue though. The computer was just built and I salvaged the hard drive, I used to run a GeForce 4. However the hard drive was reformatted. Thanks in advance!

[preacherman481]

The computer was just built

When you say the system was just built, do you mean you took an older case (with the original fan) and put these components in it?

Is it possible that you don't have sufficient cooling?

[ZT3000]

This problem could be due to anything from insufficient cooling (which I doubt in this case, unless you have the back of the computer less than 6 inches from a wall, I'm assuming the CPU fan is correct and spinning at a proper speed) to a defective motherboard, defective video card and bad harddrive.
Check your event viewer log files for any clues.

[PopFly]

thanks for your suggestions.

I bought the card and an additional stick of ram a couple of weeks ago only to discover my power supply was insufficient. I had a small box and the bigger power supply was expensive so, I replaced the case, the board and updated the processor too.

I have a feeling the video card may be the problem ... just a feeling. Before realizing what was wrong, I ran it with insufficient power for the new card for a little while right after buying it. I wonder if that could ruin the card? Device manager showed no problems. Anyway, I took it to a shop today. I'll post as to what the problem was if anybody is interested.

Thanks,
John

[.:front2back:.]

Howdy.

I'm guessing that heat is causing you the crash's. And running a lower power supply would also not help the problem..

Maybe invest in a few fans, get some better thermal paste for the processor, Also make sure that there is enough cool air running through the system, maybe leave the side of the case of and see how that goes.

And maybe give the drive a good defrag etc.

cheers
f2B

[The Duck]

Sounds to me like it's either a heat problem (take fronties advice and invest in some fans!), or a faulty video card. How old is the motherboard? It could be a faulty AGP slot or something else on the mobo. If you have an older PCI video card, you might want to try installing that and see if you still get the problem...

You should also update all your drivers...

I doubt a PSU lacking the amps for the video card would cause this problem... Why would a lack of current destroy it?? ...


EDIT***

Anyway, I took it to a shop today. I'll post as to what the problem was if anybody is interested.

That pisses me off... don't ask for help if your not going to at least attempt to listen to us... Maybe I'm over reacting...

[ZT3000]

Originally posted here by The Duck
I doubt a PSU lacking the amps for the video card would cause this problem... Why would a lack of current destroy it??

To answer your question: Because the peak power draw is too high for the power supply which tries to compensate by drawing amperage from the other rails causing them to dip below specification, resulting in a locked up (black screen) computer.

In my last post, I said I doubted that insufficient cooling is the problem. That does NOT mean it's still not an overheating condition. HUH?

(long educational read follows)
Here's why:

There's a value range called "Tolerance" and each individual electrical component (capacitor, resistor, SCR, transistor, inductor, etc) has a range of tolerances that defines the range of operating parameters it must meet for acceptability as a component. Those components then combine together to create a circuit which also has a range of tolerances it must meet to be considered a working circuit in a particular situation. Many of those circuits then combine into a computer part which has it's own particular range of tolerances it has to meet in order to be a successful part in a complete working computer.

Sometimes (too many times, these days) one, two or more individual components are at the threshold of their individual tolerance and singly they barely squeak by as being "good" individually, but when combined in a circuit or in a component and then a part (hence a motherboard, video card, hard drive, power supply, etc) their operating range is "under tolerance or bad" which causes other circuit, components, or parts not to operate at all or at least sporadically.

Don't think that all manufacturers of computer components, circuits and parts test each component, circuit or part. They simply don't have the resources or time to do so.

In a nutshell, the computer design engineers develop a standard operating design, function and tolerance for a particular circuit based on known dependable (well working) parts and test/modify the design to ensure it works as planned. Manufacturing then mass produces the component and gives it to Production, who then randomly test a number of components to qualify it for acceptance. The final step prior to shipment is QC (Quality Control), they ensure a random number of parts are again tested.

Throughout this process, not every part, circuit or component is tested and not all individual electrical components in a computer are at their peak operating efficiency even when coming fresh out from the factory.
There are plenty of complaints about bad components to attest to this.

(Please avoid overanalyzing the scenario I gave here. Each manufacturer can call each step/process by another label and/or have a different procedure they follow. I was merely giving a nutshell view based on my early 8 years as a board level (Depot) electronic circuit/system technician).

So to bring it back around, the computer can have sufficient cooling but still one component can be overheating or defective, causing the whole thing to crash. (Even as was mentioned, a slot can be defective or an internal motherboard circuit trace could be split which when heated expands causing an open circuit).

Let me add that in the OP scenario, bad memory and a funky power supply can also be the cause. In cases like this, where many things can give the exact same indications, it's always best to "divide and conquer", meaning strip the computer down to basic components (remove modem, unplug CD drive, take out sound card, whatever) then test, and add more components one by one. Only add replacement components when you think you've narrowed down the problem by using the "divide and conquer" technique.
Go from a known good state to an unknown state one step at a time.



Edit: Also not having the white 4 pin 12V plugin to the motherboard missing can cause this.

[The Duck]

Because the peak power draw is too high for the power supply which tries to compensate by drawing amperage from the other rails causing them to dip below specification, resulting in a locked up (black screen) computer.

Yes, I know this, and this is not what I was talking about. He stated that he previously had a weaker PSU powering the computer. From what I read, he now has a more powerful PSU and was wondering if a lack of current going to the video card (from the previous PSU) would make the card faulty, I said not likely... My statement wasn't really a question...

[Ghost_25inf]

a copy of pc check would help you here. get your hands on a hardware diagnostic tool and see if there is any faulty hardware. I am placing my bet on hard drive. I would say it was the memory but that would result in a BSOD. Same thing would happen if you had issues with the Video card. did you look at the event veiwer?

[morganlefay]

Freeze ups are usually....of course in my humble experience

overheating CPUs...or failing hardrives.

MLF