Death of a computer

Glenn Zetlitz

Reputable
Jun 29, 2014
19
0
4,510
Sorry l. This was the supposed text...

Death of a computer

Hey all.
I have a serious computer problem and I can't seem to find out what's causing it.

It all started when I got a new gpu and set of ram.
Occasional bsod's started occuring.
I ran memtest86 and it found no errors. The Windows memory diagnostic tool found no errors either.
The bsod message in Windows mentioned a bad amd gpu driver.

I flashed my bios and made sure all drivers were up to date, but the bsods continued.
They were not frequent or consequent so I couldn't replicate them.
The computer worked fine after rebooting so I didn't give this the attention it probably deserved.
Now just the other day, I turned on my computer and it wouldn't boot Windows.
So I pulled out the system repair disk and it did it's thing. Computer was ok again and I could continue gaming.

The next day when I turned on my pc, and again it could not start Windows.
So I get all of my disks and do a clean install. After everything was up and running again, I thought the problem was solved.
I left the computer running whilst attending to something else. It went into sleep mode and when I woke it up, it was as if I never installed the video card drivers. (resolution was 800*600 or something).
So I figured I'd do a system restore, but it failed.
The Windows repair tool would not work either. Report mentioned some damaged .sys file.

Ok , yesterday I format my system disk, and do another clean install. After everything was ok yet again I shut the computer off.
On startup today, it yet again fails to boot Windows. After a recovery I got in, but it runs extremely slow.

Now my question is, what could cause this?
Damaged hd ?
Faulty memory ?
Damaged gpu?
Underpowered system (need better (psu)?

Anyone have any idea?

Thanks!

System specs(can't get copy/paste comp is dead):
Intel I7 860
Ballistix sports 2x8gb ddr3 ram. 1600mhz
Asus p7p55d pro mo-bo
2xsata hard drives (different brands)
Msi r9 290 gpu
Mist 500w psu
 
Solution
You just upgraded to a very power hungry GPU, an R9 290 needs 275W for itself. The specs i can find on the mist 500W PSU show it having dual 18A rails and a combined capacity of 380W on its 12V rail, 275W for the GPU + 50W for the motherboard/drives + 95W for the CPU puts you at 420W so your PSU never would have been able to handle the load. You cannot upgrade to a high end video card without checking that the rest of your specs can support it, you need to buy a new PSU

http://www.silentpcreview.com/forums/viewtopic.php?p=162173
I thought I had edited the text to include my text, but obviously I had not. My bad.
Sorry guys. Yes, I do understand that people might need just a tad more info than "Windows won't boot" :)
 
You just upgraded to a very power hungry GPU, an R9 290 needs 275W for itself. The specs i can find on the mist 500W PSU show it having dual 18A rails and a combined capacity of 380W on its 12V rail, 275W for the GPU + 50W for the motherboard/drives + 95W for the CPU puts you at 420W so your PSU never would have been able to handle the load. You cannot upgrade to a high end video card without checking that the rest of your specs can support it, you need to buy a new PSU

http://www.silentpcreview.com/forums/viewtopic.php?p=162173
 
Solution
@vic 40 Yes that's the one :)
as for the ram, I checked that my mo-bo supported it, and since I couldn't find any errors when running memtest and win mem diagnostic I just thought it was all good. Could the mem actually still be an issue?

@hunter315
I always suspected I had to get a new psu.
But I'm no expert by far when it comes to hardware. The site I bought the gpu from stated it needed at least 500w psu, so I figured I'd give it a go. I assumed (until now) that a 500w psu gave me 500w . Would you mind explaning to me(or linking me to a good explanation in layman's terms) how power distribution actually works?

Thanks for your help so far guys.
 
Well a psu is most efficient when at around 60% load, so you don't want to be pushing that, and the GPU would consume more power under heavy load, so that might be it, but Hunter can probably give a more in depth answer.
 
A modern computer system draws 80-90% of its power from the 12V rail of the PSU, it powers your GPU, CPU, motherboard, and anything with a motor. Older systems used to power the CPU off of the 5V rail so they had much strong 5V and 3.3V rails and much weaker 12V rails, those units are awful for modern systems. When picking out a PSU check the label, a good unit should have the majority of its power available on the 12V rail, you usually see a 500W unit with 40A(480W) on the 12V rail, in most modern units the 3.3V and 5V rails are generated off of the 12V rail, but the loads on those two rails are quite low.

PSUs are heavily impacted by age and temperature, a good unit will be built to function at full capacity at 50C, a cheaper unit will be built to function at full capacity at 25C and will lose 1-4W/C above that, so by the time you get to 50C you could be short 100W of the rated capacity. As components get older they also start to lose capacity, capacitors are a good example of this, they smooth out the power and lose capacity over time as a function of temperature, for every 10C different they age 2x faster, so a good unit built with 105C capacitors will last 4x longer than a cheap one made with 85C caps.

As for efficiency, yes they are more efficient in the middle but the difference is rather small so that is a moot point, the extra 2% efficiency at 60% load will never pay off the extra spent on an oversized PSU. As long as you get 80+ bronze rated PSU and up the efficiency curve is pretty flat.


Where are you able to buy from? There have been some suprisingly good deals on XFX PSUs lately, the EVGA G2 Supernova Series is also good and well priced generally but it really depends on your selection. For a system with an R9 290 i'd really suggest looking at a 650W PSU.

 
Thank you for your thorough answer. Great stuff. I had no idea a psu lost such significant power efficiency.
And just to clarify: an underpowered psu could cause the problems I wrote about in my initial post?

I have been hinting to my underpowered psu at several forums but answers in general have focused on the ram.

I'm planning to purchase either the cooler master v1000 or the xfx black edition 1050. Both are in the 80's and have gotten some great reviews.
They are however somewhat costly c so for now I'll bite the bullet and uninstall my shiny new r9 290 and revert to my old gpu.
I really hope that'll help and that I haven't inflicted irreversible damage on my computer.
And if that doesn't work I guess I have to take a better look at the memory chips as well.


I'll report back when I know if this helped.
 
Those over a 1000watts are way to heavy or you must want to go for crossfire.A 650watt psu should be fine for one card and overclocking,this will be less pricey too.

It's also possible that the new ram just isn't compatibel with your motherboard so probably why people point that way.Have you tested with the old ram?
 
@vic 40
What do you mean by "way too heavy"? You mean it's overkill ? I'm just trying to learn from my mistakes. When I built the computer I thought the 500 would be more than enough for years to come..
i checked the mo-bo/ram compatibility. They should be ok. (Mb:Asus p7p55d pro ram: Ballistix sport 2x8gb ddr3 1600 (pc3 - 12800) cl9 @ 1,5v.)
I've just put back my old gpu and my old ram, but the computer runs so slowly it's hard to get anything done.
I have memtest86 on a usb, so I'm going to check the new sticks one by one.
But seeing the computer running so extremely slowly now, should I promote my hard drive to the prime suspect ?
Any good tools besides chkdsk to determine if my hd is corrupted?
I have to hard drives btw. Maybe I should reinstall win yet again, make the other one the system drive , remove the one I use as system drive now to see if the problems persist?
 
You could try to run a test on the hdd's with a tool from the manufacturer or the next,
http://www.hdtune.com/download.html
do the error test and look at the smart values.

You could also swap both drives to see if the current systenm drive is bad,but that's more work.
Have you tried the bios reset/clear cmos with the old ram too?
Do you have ahci enabled in the bios?
Did you check the sata cables or with different ones? Maybe try the one from the second drive with the system drive.

Look at this and see if the hdd is in pio mode,
https://www.google.nl/url?sa=t&rct=j&q=&esrc=s&source=web&cd=4&cad=rja&uact=8&ved=0CDsQFjAD&url=https%3A%2F%2Fwww.ifixit.com%2FWiki%2FEnabling_DMA_Mode_Transfers&ei=SZCxU_r4EYbEOcOfgaAL&usg=AFQjCNEZNUsLVGdwS8kgIAMxOA2uLvdXcw
Look under "Enabling DMA Mode Transfers" how to enable this and/or look if it is enabled.
 
@vic 40
Thanks for your suggestions, I shall look into it once I've memtested my ram.

As for the cmos, this is all new to me.
I did update my bios after the bsods started occurring. (used ez flash), but I don't have that anymore because I formatted the usb to make room for memtest86.

I shall definitely try switching power cables between the drives. Maybe it'll help.
As for anything that has to be done in Windows, that's out of the question. The system quite simply runs too slowly to do anything. I'm talking maybe 10 mins just to open the Windows explorer. Seeing the content of a folder is even slower.
Basically this leaves me only with the option of working with programs I can use on boot up. (like memtest)
"Ahci"?
Also, wouldn't resetting my cmos mean I have to update the bios again?
I can however use my gf's laptop to download stuff on a usb.

 
Clearing the cmos will reset the bios,that will be the latest you put on it and it will not be necessary to update the bios again.You should after that set the optimal default settings You might also need to set the time and date right again.
Clearing the cmos can be done by moving the right jumper for about 5seconds,which that is will you find in the manual for your motherboard.
The other way to do that is by taking the pc completely of the power=cable out of the wall and removing the battery for about 15minutes,after that put it back again,again set the bios to it's optimal default settings.

Ahci will you find under the sata settings in the bios,probably under the "advanced" tab.or something like that.Again read the manual for that,it can be downloaded from the asus site.Probably a pdf file.

It could be that you could run hd-tune from the hiren's boot cd,
http://www.hiren.info/pages/bootcd
http://www.hirensbootcd.org/files/Hirens.BootCD.15.2.zip (download zip is in the green marked section)
make a bootable device with this and try hd-tune.
 
ok then.
what i've done now, is that i let memtest run a thorough control of both ram chips. they are ok.
so next i went into bios and changed hard drives to AHCI and reinstalled windows on a new partition.
Using my old gpu and my new memory chips, the system now seems ok.
I'm running a check on one of the partitions (i created a smaller one for windows, an a larger for storing files) to see if the hard drive is ok.


I haven't had enough time to state that problem is solved, but it sure looks better that it did yesterday.
 
Hopefully my final update on this subject.
The computer ran fine with the new ram/old gpu setup. So I've now ordered and installed my my xfx 850w psu and my r9 290 gpu.
Going to see how it works out, if my problems persist there must be some damage to either hard drive or the gpu itself.
Thanks for your help guys!
 
Well, I guess I'm back...
The computer has been running for a couple of days, with only some minor issues.
It freezes for about a minute or so from time to time. (not in games) and then continues as normal.
Also I have had a couple of error messages on via hd audio, and the raptr app. (amd gaming evolved).
Now when I turned on my computer this morning it booted up normally, but when I tried launching steam, it froze. After a little while the screen went black for a few seconds and then I was presented with a Windows background and a set of error messages concerning 6-engine (Asus mb software). I think there were three or four errors. (dumb as I am I didn't take a picture of them, so I can't remember what they said...)When I had clicked through the errors, I got a bsod. I grabbed my mobile phone to take a pic of the bsod, but it disappeared too fast. The error code ended in f4
On rebooting computer ran chkdsk wich deleted and restored a bunch of files.
I am now back in Windows and the details from the unexpected termination is as follows:

Blue Screen
OS-version: 6.1.7601.2.1.0.768.3
Nat.id.: 1044

Additional info:
BCCode : f4
Bcp1: 0000000000000003
Bcp2: FFFFFA800FE97060
BCP3: FFFFFA800FE7340
BCP4: FFFFF80003188270
OS version: 6_1_7601
Service pack: 1_0
Product: 768_1

I have no idea what this means. Any and all help is appreciated!
 
The old mem worked fine. But new RAM has been cleared by both memtest86 and win.mem.diag.tool.

Also New RAM/old gpu setup worked. Although I didn't test it for more than two days. I still got those freezes though.

Could this be a hard drive failure i wonder?
I was running hd tune just now. It found 2 bad sectors pretty early on. I left the computer and it showed a . I don't know what hd tune found a dos screen promting me to enter a valid boot drive and reboot.
I rebooted, hit f8 and set boot to the system hd and Windows booted as it should.

I really can't seem to figure out what's up. It's all just a little information overload atm.
I may have to bite the bullet and send it to a repair shop.