Weird pc issue

centra28

I've been having some ongoing computer issues that I have thus far been unable to nail down. I've built and maintained my own computers for years, but this one I'm not 100% sure about the exact issue. Unfortunately I didn't write down any of the specific error messages I've received, so I'll just be generallizing the issue...

I started building a new gaming pc, but used a couple of components I already had on hand (this is where I thought the original issue stemmed from). I bought a new AMD processor (Phenom X2 560 BE), coupled w/ an ASUS M5A88-V EVO (880G chipset). The processor was brand new, but the mobo I got on sale as an open box item. Aside from a couple cosmetic issues, the mobo seems to work fine (from what I can tell at least). I checked the supported memmory list from ASUS's website for the board, and purchased the appropiate DDR3 memory (Corsair TW3X4G1333C9 XMS3 4GB 2 X 2 GB PC3-10666 1333Mhz Dual Channel 240-pin DDR3). I pulled over my sata dvd burner drive, 700 watt psu, Radeon HD 3870 gpu, & my Nirvana CPU heat sink fan. OS is Windows 7 prof 64 bit. For the hard drive, I reformatted an older WD drive I was just using as a backup driver on my previous pc.

Once I got the OS installed, and all the appopriate drivers and bios settings figured out, I started having weird drive issues. I would get blue screen, and weird error messages that led me to believe the problems stemmed form the hard drive. I ran a chkdsk f/ and that seemed to resolve the issues for a while. Then I started having worse problems. The computer would crash and not boot to windows, and I would be forced to do a system restore. After the chkdsk, the new issues didn't start occuring until after I had installed and playing Neverwinter Nights. The pc would crash during play, then I would have to restore. After the restore, some system utilities would be gone or unavailable. Like the control panel for example. I could click on it, but it would say that is was unavailable or did not exist. Weird stuff I've never seen before. So at the point I wasn't sure if the issue was just due to the HD or maybe from the older GPU. I assumed the HD was no good, and wasn't storing data properly. To solve this I purcahsed a brand new WD black edition. I also went ahead and purchased a new gpu (xfx 6870 BE) to eliminate any issues I might have been having w/ my older 3870. I was in the market for a new gpu anyways, so I didn't mind spending the money on that.

After installing the new HD and GPU, teamed w/ a fresh OS install, I am still having issues. The computer will just shut down randomly, like it's losing power. I will have to jiggle the psu connections, and it will have power again. The first game I downloaded was Black Prophecy. Everytime I run the game it freezes, gives me blue screen, or the pc just shuts down. Now I'm at a point where i'm not sure if the issue is coming from the open box mobo, a bad psu, the game, or just all or some of the above. My thoughts are that the psu definitely needs to be replaced. How many of these problems could stem from a bad PSU though? Maybe the components aren't getting all the power they need? I'm also worried that maybe some of the mobo connections are no good (like the PC-E port for example). I would just like some additional addvice before I throw anymore money at it. Sorry for the tl;dr, but I do appreciate any advice or ideas on this. Thanks.

genoski

I would put my money on the motherboard. That was your first mistake. Never buy an open box anything... more often than not, you're buying what someone else once bought and sent back because they had problems with it.

If you had no issue with the power supply in the previous build, it's unlikely that is an issue now. Bite the bullet for a new motherboard with a warranty.

iggy8n

I second the opinion that it's the motherboard. It seems like the other parts have been working in the past. The only unproven to be working part is the motherboard.

ManofGod

The very first thing I would verify is that the ram is good. Run a ramtest and see if you get any errors. Also, verify that the ram is running at 1.6V and the timings are correct. (9-9-9-24) That, and run one stick of ram at a time to see if that helps. (Perhaps also boost the ram voltage to 1.65v to see if that changes anything.)

If none of that helps, then there is a good chance that the board will need to be replaced. (I honestly believe the symptoms point to a ram issue however.)

centra28

well that is disheartening to hear. i was hoping to get a unanimous "it's the psu for sure". i'm currently kicking myself for purchasing the open box item. and of course it took me more than 30 days to go through all the processes, so it's too late to return it.

on this psu, i was using it for my previous build, which would also get me bsod every so often. but that i nailed down to a memory issue. this i don't believe is a memory issue. i did run the memory diagnostics test, and it came up clean. that's not a 100% guarantee, but my thoughts were it was the psu or the mobo, or even both. i wouldn't allow myself to believe it was the mobo, so i've been leaning heavily towards a new psu. damn i would hate to have to buy a new mobo. i don't really have a good way to test out either... i don't have another psu that i can use for testing, and my other pc has the memory issue still.

centra28

I think I know what I'm going to do before I buy anything else. I'm going to have to take the time to nail down the memory issue on my other pc, and pull out the bad stick. I'll then switch the psu over to that pc and see what happens. I have a friend who has a psu I gave to him that he no longer needs. I'll have to drive my butt over to his house to grab that to use on my new rig to test the mobo out. /sigh

centra28

i've cross posted this over on the eggxpert forums, and i'm curious to here your thoughts on this response:

"Well, you're right that the first part sounds like a bad hard drive.. And in that situation you should have pulled out a multi-meter and checked your 12V rails for voltage stability.. I've seen many times where a bad PSU will undervolt the 12V and cause the hard drive to fail to start up after being spun down (due to power management).. The drive just doesn't get enough juice to get the platter spinning and it just times out and gives up which would cause a kernel panic / crash no matter what OS you were running..

The second part about it randomly shutting off, is pretty clearly a PSU issue or a home electrical / grounding problem.. All my Antec PSUs died in this manner after 3.5 - 4 years where they started randomly shutting off... The 12V would start to dip in voltage as the PSU warmed up and right when it dropped below 10V, the PSU would just shut the whole system off.. That's the undervoltage protection of the PSU kicking in.. Ran the hardware for years before it started doing that, so the PSU certainly wasn't overloaded.. The Antec PSUs that I had were pretty good quality too, they certainly weren't cheap junk..

The CPUs on motherboards these days have got so much voltage regulation and capacitors behind them that they'll actually run on significantly less than 12V.. But it seems like the hard drives and the fans are the ones that really suffer the most when undervolted since they don't have the luxury of so much voltage regulation circuitry that's on the motherboard.. So they just don't spin up, simple as that.. ..

Though I've always been curious.. If you ran an SSD instead of a hard drive, I wonder how far below 10V your 12V rails can slip while the system is still usable.."

TechGuy

iggy8n

If you could get a meter on your 12 volt line when the hard drive is doing the crashing, you should see if the voltage has dropped or the hard drive has failed for some other reason. That's if the hard drive is really failing ever. If the motherboard doesn't get 12 volts from the PSU, it doesn't matter how many voltage regulators or capacitors there are on the motherboard, you just can't build up a voltage if you don't supply it at the input. There are voltage doubler circuits that could do it, but, I don't think it would be worthwhile because it would require more current from the PSU to double the voltage then regulate it back down to 12 volts. Inefficient, to say the least.

genoski

genoski

centra28

centra28

TechGuy

iggy8n

centra28

centra28

iggy8n

TechGuy

ManofGod

Sorry Techdude, but I cannot not agree with you statement at all. The mainboard itself is rarely the problem of system failures by my experience. (Hardware wise, that would be harddrives or ram in that order followed by power supplies.) Also, people do not usually fix computers themselves for the same reason they do not fix cars, they may not have the knowledge/experience or have the time.

That and just guessing does not usually work out to well. Centra, once you are done, you are going to have a pretty fast computer. I am just glad we are able to help as much as possible.

genoski

TechGuy

ManofGod

iggy8n

Since these are all electronic parts made of silicon, they can all fail one way or another. A transistor can get a voltage spike that burns it out. This transistor could have been on a board with banks of memory chips, a PSU voltage regulator, a GPU with all kinds of silicon chips making video frames and shipping them out to a screen. There are plastic connectors, wire, soldered connections that can heat to the point of melting the solder, etc. So, diagnosing a problem is not always like what happened last time. It's great to have a logical way of diagnosing. This will find the bad part sooner. But luck is also a part of solving the mystery. Sometimes, you just happen to have a part that will replace the bad part in an old computer. Other times you don't.

TechGuy

centra28

TechGuy

centra28

TechGuy

genoski

iggy8n