Tech Support Forum banner
Status
Not open for further replies.
1 - 12 of 12 Posts

·
Registered
Joined
·
23 Posts
Discussion Starter · #1 ·
I put together my rig just over a month ago with parts I bought from Newegg. Here's an image of the cart,
.
Except I actually ended up using an XFX 6850 instead of a Sapphire.

It worked fine on Day 1, and Day 2. By the end of Day 3, Crysis was occasionally crashing. By the end of Day 4, Crysis was always crashing. Around Day 5, every game crashed eventually. Starcraft II crashed whenever a game began (but it was fine in the menus).

I went crazy with frustration, and ended up thinking it was the memory sticks, since Memtest86 reported failures constantly. So I RMA'd my sticks and got them replaced, and they came in the mail just the other day. I thought they'd solve it, but nope. Still crashes whenever intensive gaming (or any intensive computing at all) is done.


So today I got fed up and brought it into a new TigerDirect store that had just opened up near me. Well, I kept the PC in my car, and brought in only my parts list. I didn't want to get charged anything yet.

So I showed this really smart, really nice guy there my list and he looked at it and became convinced my 500w power supply wasn't cutting it. Especially since it was the "green" version, which makes it kinda wimpy. I agreed, and decided replacing the power supply to a heftier wattage would help. It seemed to fit the symptoms: absolute PC hanging whenever I tried a game or intensive computing.

Next, just for kicks, I went into Best Buy. I wanted to see how what they said compared to the Tiger Direct guy; I didn't expect to be given good information or to fix the problem. Just out of curiosity, you see.

So I brought in my list and told they guy there my symptoms, and he said it was most likely the CPU, possibly the GPU, probably not the power supply, and also maybe the motherboard.

That didn't help much. And it conflicted with the guy from TigerDirect.

I haven't made any changes yet. I'm here to ask what you guys think. Here's my symptoms:

* When I turn my PC on, it first shows nothing. No BIOS, no OS, NOTHING. Screen displays nothing. The motherboard emits 2 short beeps, then a pause, then 5 short beeps, then it repeats. It was obvious this was a mobo code for something, and so I looked in my manual to see what it meant. THE MANUAL DIDN'T SAY. I STILL have no idea what those beeps mean; no matter where I looked or googled, I can NOT find 2 beeps, pause, 5 beeps. This is a NEW problem that began just today when my memory sticks came back; the beeping began the very first time I tried to start the PC with the brand new sticks.

* When I restart my PC from those beeps, it always boots perfectly. It's like turning it on is a 2 step process: I must turn it on once, listen to the beeps, then restart to make it work.


Here's a section on tests I've run:

* FurMark crashed 2 minutes in at 70 degrees celsius for the GPU. And when I say "crashed," I mean it hung the PC to the point that a restart was required. So I restarted, and...nothing. No beeps, no BIOS, no OS, nothing. It seems that now whenever the PC crashes, I need to wait about 10 minutes before I can turn it on again.

* Prime95 on the small FFT setting hung the PC instantly. Like, literally the instant I hit "go." Not even a second's delay.

* Memtest86 is running now, though I think it'll pass just fine. I'm not concerned with my brand new sticks, anyway. They seem fine.

* The PC works fine, generally, when I'm NOT doing intensive stuff. I can run Chrome all day long, along with iTunes, Word, etc, and be fine. It's only when I play Starcraft II or Just Cause 2 or Crysis or GTA that I inevitably hang.

* Here's an interesting fact: it's inconsistently bad. There was a point when I was crashing EVERY TIME I tried to play a game, or do anything. Then, out of the blue, it decided to start working. Battlefield Bad Company 2 would play just fine. Starcraft II would play just fine. This was about a week after having problems. I thought maybe they had gone away on their own by magic. Then, a couple of days later, it started again. Nothing would play; no games at all. That's where I am now; back in the time where I an navigate the SC2 menus, but not start a game.

Well, guys, what do you think? I have way too many options, and way too many things that can be wrong.

I think the TigerDirect guy was right about the power supply; yes, it should be stronger. In the end, I suppose I'll end up getting a better one. But can that be what's causing EVERY one of these problems? Does that fit? Or does it seem like there's more to it?

It could also be the CPU.....that would explain Furmark and Prime95 failing, yes? Or could those both be accounted for by the power supply not being strong enough?

I know what it's NOT; it's NOT software or drivers. This is all hardware; I'm almost positive. (I say "almost" because it's impossible to be positive about anything anymore with this issue)


So, amazing people in this forum, I ask of your incredibly generosity only two things:

1. What do you think about the above information I provided you with? I know it's hard to diagnose without being there in person, but from what I provided you, what do you think? How does it add up?

2. Since you probably can't help me a ton with just the information above, who do you suggest I go to? If it comes down to me saying "I give up. I don't know. I can no longer do this myself," and I have to take it in to some brick and mortar shop, where do you suggest I go?


Thanks so much, you guys. You're the reason I'm able to at least have the hardware put together, even if something's busted or I did something wrong.


EDIT 1:

Just dropped $65 on a 650watt PSU. Fixed absolutely nothing. I still get the 2 beeps pause 5 beeps, and Prime95 still hung the computer almost instantly.


I guess the next step is to try a new CPU....

EDIT 2:

I just remembered something incredibly stupid I did. When I first assembled the PC, the CPU fan wasn't attached 100%. It was still on pretty tight and making direct contact with the CPU, but it was able to slide around a little across the CPU (even though it was still making direct contact with the CPU). After it started crashing, I realized I had probably put it on wrong, so I reattached it but this time I put it on all the way. That didn't fix the crashing. But maybe the CPU blew out from when I had the fan on wrong? Does that seem very probable? Or do we still think the mobo is the most likely cause? I'm trying to figure out what part to RMA next, and it's between those two (even though people say RMA the GPU too, I still don't think that's what is causing the problems)
 

·
Emeritus
Joined
·
51,675 Posts
Ther PSU is lower quality and isn't powerful enough for the 6850. A 650W good quality would be what I would want.
Do a bench test

Remove EVERYTHING from the case.
Set the motherboard on a non conductive surface. The motherboard box is perfect for this. DO NOT PLACE THE MOTHERBOARD ON THE STATIC BAG! It can actually conduct electricity!
Install the CPU and heat sink.
Install 1 stick of RAM.
Install the video card and attach the power supply connection(s) to the card if your card needs it.
Connect the monitor to the video card.
Connect the power supply to the motherboard with both the 24pin main ATX Power connection and the separate 4 or 8 pin power connection.
Connect power to the power supply.
Do NOT connect ANYTHING else. Make sure you have the power connector on the CPU fan connected.
Use a small screwdriver to momentarily short the power switch connector on the motherboard. Consult your motherboard manual to find which two pins connect to your case's power switch. Then touch both pins with a screwdriver to complete the circuit and boot the system.

If all is well, it should power up and you should get a display. Then assemble the parts into the case and try again. If the system now fails to boot, you have a short in the case and need to recheck your motherboard standoffs.

If the system does not boot after this process, then you most likely have a faulty component. You'll need to swap parts, start with the power supply, until you determine what is defective
 

·
Registered
Joined
·
23 Posts
Discussion Starter · #3 ·
Thanks. I have a few questions about that process:

First off, can I remove and reattach the CPU to the motherboard again without applying new thermal paste to the heatsink, or should I reapply? Of course, the CPU and heatsink will still have thermal on them, but should I wipe that off and apply new stuff or can I use the stuff that's already on them?


Also, if you read the whole thing you'll know what happened with the PSU (I replaced it with an Ultra brand 650watt. didn't fix anything at all). The memory sticks are also fine: they're brand new, and pass Memtest86 with flying colors.


I'm really hoping it's the CPU; I'm pretty sure that's it, unless it's the motherboard (or the GPU, but I'm fairly confident it's not the GPU...because why would it hang when I run Prime95? Prime95 doesn't use the GPU, right? Just the CPU).

So basically, it's either the CPU or the motherboard. And I have evidence to suggest it's the CPU: the fact that I had the heatsink on wrong when I began experiencing problems. So I'll start there.
 

·
Registered
Joined
·
23 Posts
Discussion Starter · #4 ·
I've been told by some to upgrade the BIOS as well. However, I'm really afraid to do this. First off, I honestly don't think the BIOS is causing all my problems. It doesn't seem to make sense.


Secondly, what if it crashes while updating? In fact, it probably WILL crash while updating. Not a good idea.
 

·
Registered
Joined
·
23 Posts
Discussion Starter · #6 ·
Duanomo, I will check temps in the BIOS right now. Except the crashing has started to become exceptionally bad. Usually it'd only hang during games or tough processing, but now it's even started freezing and hanging in the BIOS. That's not a good sign at all. Sometimes it won't even boot up.

Just recently, though, when it was in a "good mood" and decided to work for a little while, the temps were really good. Core Temp was reporting around 20 C for the idle CPU. When it used to play games for a little while, the highest I saw it get was around 50 C. Just yesterday I ran Furmark, which got my GPU up to 70C. That's really hot. but that's not so hot for a GPU, right? Because on Furmark's scale, 70 C was about right in the middle, in the "light red" for sort-of-hot, and 120C was the very max of the scale. I'm not sure if 70C is a good temp for a super stressed GPU. I hope it is, and I found a site that said 70C was a pretty normal max temp for a GPU. However, it crashed the moment it got to 70C, about two minutes into the test.

That would, of course, lead you to think the GPU is the problem. But if the GPU is the problem, then explain why Prime95 fails? If you ask me, that points to the CPU being the problem. I'm sure that, even though it's a GPU test, Furmark probably uses a ton of CPU usage as well, correct? So Furmark tests the GPU and the CPU, and it fails. Prime95 tests just the CPU, and it fails. That would lead me to think it's the CPU.


My theory is that the CPU "burnt out" back when I first installed it. Like I said, I had the heatsink on incorrectly. It was on, but not fully latched down. The PC worked fine for about two days. I'm guessing that the CPU was running really really hot because the fan was on wrong, and then it started to fail and get worse and worse over time, because I kept overheating it. Then, I finally realized the heatsink was on wrong, and I corrected it. But by that point, it was too late, and the damage to the CPU had been done. Even though the heatsink is now properly on and I'm not getting hotter than 50C (which is when it usually crashes, anyway. then or long before then), it's still failing because it's damaged from when it used to overheat.


Does that make sense? Does that sound likely? Mostly that's just me really hoping it's the CPU, because that's more easily replaceable than the motherboard. What do you guys think?
 

·
TSF - Enthusiast
Joined
·
7,670 Posts
Update BIOS only if the update directly addresses an issue which you are experiencing. NEVER attempt to update BIOS on a computer which is not stable.

edit: The CPU may well have been damaged if it has been overheated.
 

·
Registered
Joined
·
23 Posts
Discussion Starter · #8 ·
Ok, thanks gcavan, that seems to be the case. At least, I'm *assuming* and *hoping* it's the case. I'm gonna try to RMA it as soon as possible.


Just wondering, does this sound like likely symptoms of an overheated CPU? Like, does a busted, overheated CPU work as long as it's not stressed? Or is it one of those things where if it's not working 100%, it's not working at all? You know what I mean? I can't really think of an example....

I'll just rephrase it: even if a CPU is damaged, is it possible it still works partially and under little strain? Because that's how mine is right now: it works until you stress it. If busted CPU's can work a little, that's good, because it's more evidence that the CPU is what's wrong. However, if a overheated CPU shouldn't work AT ALL, that's not good.


Sorry if that's really confusing; it's not hard to understand, I'm just horrible at explaining things.
 

·
TSF - Enthusiast
Joined
·
7,670 Posts
In my exp it has always been an on/off thing. Either it works or it doesn't. But then I haven't had many bad CPU's. One of the guy's doing this as a full time business may be able to tell you more.

Now that being said, let me go slightly off topic for a moment. I'm sure you have heard of people unlocking the "hidden" cores on dual or triple core AMD processors. Those hidden cores were often locked at the manufacturer because the proc didn't pass QC as a quad core but did as a dual, say. When unlocked, the procs often exhibit symptoms such as yours.

I'm thinking maybe only one of your cores has been heat damaged. But I'm just guessing here.

In any case, at this point, you are down to swapping out parts. Most likely suspect is probably the CPU.

By the way, did you complete the benchtest described by Tyree in post #2? If not, I would try that first. There is a sticky on that subject here.

Also, if you have access to another PCI-e graphics card, it wouldn't hurt to swap it out for testing. (doesn't have to be anything great, only has to work)
 

·
Registered
Joined
·
23 Posts
Discussion Starter · #11 ·
Just in case I somehow didn't mention this, the PC WILL boot. Most of the time. If it doesn't boot, I can keep restarting it until it does. I CAN get it to boot, and only maybe 2 times out of 10 does it not boot. Well, actually, it DOES do the "two beep pause five beep" thing EVERY time, and I have to restart from that, but it still boots in that scenario.

My problem does not occur during booting. It occurs in intensive use of the system. Prime95 will hang the PC instantly. Starcraft II will hang the PC when a game begins (but it's fine in the menus).
 
1 - 12 of 12 Posts
Status
Not open for further replies.
Top