Sign in with
Sign up | Sign in
Your question

New Computer Build Randomly Loses Power. Help!

Last response: in Systems
Share
January 17, 2012 7:41:17 PM

Question Disclaimer: Howdy. I've been browsing the forums for the last hour and a half looking for a similar problem, so first let me say that I'm really sorry if you have answered this question a million times. I kept finding related questions, but have tried everything I have read and still haven't fixed this problem.

My Amount of Technical Knowledge: This is the second time I have ever built a PC from scratch in the past 8 years, so I don't know all the acronyms and such, but I feel i know enough to at least answer any questions you can pose.

My Problem: My computer will randomly (with no set pattern) lose power for 3-5 seconds and then just restart itself. Upon restart, it shows me the options for how to restart windows (in Safe Mode, in Last Known Good Configuration, or Normally). Sometimes upon losing power it will not restart, and pressing the power button on my case causes the case fans, CPU fan, and Video Card fans to stutter but not start, and the computer does not power on. If I reset the power supply for 15-20 seconds and try again it will start most of the time, but on certain occasions it won't start at all.

My Computer Build:
CPU -- Intel i5 2500K
Motherboard -- ASRock P67 Extreme 4
RAM -- Corsair Vengeance (2 x 4GB) Low Profile
Graphics -- MSI GeForce GTX 560 Ti - 448 Cores
Storage -- 1 old Seagate 1TB drive. 1 old Seagate 300GB drive.
Power Supply -- XFX Core Edition Pro 550W
Case -- My old Antec 900 Changed to Antec P280.
Optical Drive -- ASUS DRW-24B1ST
Heat Sink -- COOLER MASTER hyper 212 Plus
SSD -- Intel 320 Series 80GB SATA II (MAIN HD)
NOTE: If you need links to the exact parts I bought, I have them. Parts are from NewEgg.

History of Troubleshooting:
1) Computer worked fine upon first build and lasted 3-5 days without a shutdown during heavy usage (Addiction to SWTOR).
2) Around day 5 the computer starts to randomly lose power. This happens when I have SWTOR and Steam up. I assume that is the problem and don't run the programs together.
3) Day 6 - Computer starts to power down multiple times in a row (3 to 4 times) but then works fine for the next two days. I did not restart the system during this period. The system could have rebooted itself when I was away at work and I would not have known as I leave no programs running when I leave it.
4) Day 8 - The computer finally won't power on at all after a power down, and I get the fan twitch error mentioned above.
4a) I discover that holding down the power button on my old beat up case as hard as I can will start everything up fine and assume the button is the problem.
4b) I buy a replacement case (Antec P280), disassemble the computer from the 900, and reassemble it in the new P280 case. Everything again works fine for the next (1) day.
5) Computer starts to randomly lose power again.
6) Come to Tom's Hardware forums for help.
7) I have done all 23 steps in the checklist of troubleshooting a new build that won't POST (I'm not 100% sure what that word means BTW). Everything was OK. I am currently running Memtest86+ because I saw a thread about it.
8) I am now posting here for help.


I'm sorry for the super long message, but I assume the more information I post the better. If anyone needs more information, please let me know. If this is the wrong forum, please let me know.

For anyone that made it this far, Thank you so much for reading!
January 17, 2012 7:56:59 PM

"Losing power" is not a good way to characterize this. Your PC is doing a full reboot without your input. If you were losing power all the lights would go out on the board, and I doubt that is the case. You still have standby power.

POST means Power On Self Test. It's the steps your motherboard does before the operating system boots.

If you have truly been over the checklist then you will have already tried the following:

Re-seated your CPU cooler and made sure it was correct, and had the thermal paste CORRECTLY applied.
Tried each stick of RAM individually.
Re-set your CMOS to defaults

Does this only happen in Win 7? Does it ever happen before you boot? If you sat in the BIOS menu would it happen?

m
0
l
January 17, 2012 9:08:22 PM

Unfortunately, "losing power" was the only real way I could think of describing it. Whenever it happens all of the fans on my case, on the CPU, and on the video card stop spinning. There are no glowing lights on my motherboard so I have no idea if it has standby power or not.

1) I have not re-seated the CPU cooler yet as I don't have any thermal paste handy. I need to make a run to Fry's tomorrow to pick some up.
2) I am currently trying out each stick of RAM individually and am on the first stick, but because the power cycling is random I have no idea when it's going to happen again. I'll post an update once I have tested both sticks.
3) I re-set my CMOS to it's defaults with the jumper on the motherboard following the instructions in my manual. I then set the jumper back to "default".

4) As far as I know it has only happened in Windows 7. Each time I've seen it reboot itself has either been on the "input password" screen of Windows 7, or when I've logged in. After I have checked the RAM I'm going to boot into the BIOS menu and leave it on for a day to see if it reboots.

Thanks for the quick help. Are there any other steps I should take?

EDIT: I absolutely do have stand by power when the fans stop spinning. I just remembered the first time it happened when I had the side cover off. My motherboard has two buttons (Power and Reset) that glow red on start up. The fans did stop spinning once during boot and both of those buttons continued to glow red well after the fans stopped spinning. I was able to restart the computer by hitting the power button on the motherboard itself. Sorry for characterizing it incorrectly :( 
m
0
l
Related resources
January 17, 2012 9:29:28 PM

Right click computer, properties, advanced system settings, startup and recovery, settings, uncheck "Automatically Restart".

I think you are getting Blue Screens of Death, and Windows is not showing them because (I suspect) this button is checked by default.

Also, download a program called MemTest86+ and copy it to a CD and boot with it, let it run for many hours when you aren't trying to use the computer.

This would help finish ruling out a RAM problem.

Also, take the sides off of your case and see if you continue to lose power.

Write down as much of the error message as you can from a BSOD if you get one.
m
0
l
January 17, 2012 9:44:01 PM

I've had "Automatically Restart" unchecked since the first time my power cycled because I too thought I was getting BSODs and they weren't displaying. In the 8 or so restarts that I've been at the computer I have never once seen a BSOD, even with that box unchecked.

I have already copied MemTest86+ to a CD and am waiting to run it, but figured I would test each RAM stick one at a time. Instead, I'll hook up the other stick and run it as it seems like my computer has decided to not restart itself in the last 2 hours. It never happens when you want it to, does it? :(  .

I currently have one side off of my case and am running it that way. Just for reference, my case has 5 fans running in it at the moment. There is one in the rear blowing out, two on top blowing out, and two on the front sucking air in thru the Hard Drive Bays. Also, the Power Supply fan is sucking air in from the case instead of from the outside VIA the panel on the bottom. My computer is on carpet, so I figured it's better this way. Do you need me to remove the other panel that allows access to the back of the CPU mount? Should any of these fans have their directions changed? Should I flip the PSU around so it's sucking air in from outside the case?
m
0
l
January 17, 2012 11:41:31 PM

It's still early in the process and I won't venture a guess, but I do want to point out that it could be many things. It would be as simple as a bad connection somewhere, or something more complicated.

Booting from the memtest CD and running overnight will reveal much. If it survives that long.
m
0
l
January 18, 2012 11:00:02 AM

Update: My computer stayed on all night and went through 10 cycles of the MemTest86+ software with 10 successful runs and 0 errors (both RAM sticks inserted). I left it running while I'm at work because I don't know how many times it needs to run the test. I figured it would stop itself when everything is ok.

Wow, thank you for those links! For cooling: I'm going to switch my PSU around so it is sucking air from outside the case. My carpet isn't super "fluffy", and I can always set the entire case on a spare piece of Wood I have sitting around so that carpet fibers getting sucked into the fan don't become a problem. Also, because I have two exhaust fans on the top and back of the case I'm going to keep my CPU cooler in it's current "Upright Tower Cooler Installation" orientation.

Question though: Positive vs. Negative pressure - I currently have 2 fans in and 3 fans out (Negative Pressure). The writeup says that Graphics cards without the ability to exhaust heat don't benefit from this setup... my card exhausts heat correct? It has two cooling fans on it, and a slotted vent on the back of the computer (after connecting it to the motherboard). I assume it does, I just want to make sure that air is going into the card via the side fans, and out of the card via the back of the case.

Also related, I notice that the image with this configuration (http://media.bestofmicro.com/0/C/312348/original/Verhal...) has air flow measurements listed. My Fans have speed switches on them (High and Low). Should I purposely set the rear exhaust fan at it's highest setting in order to have a higher air flow and set the top 2 and front 2 fans to their low settings? I'm worried the m^3/h equation (in vs out) that I'm creating is wrong.

m
0
l
January 18, 2012 1:29:18 PM

If it has two center mounted fans like this model

http://www.newegg.com/Product/Product.aspx?Item=N82E168...

Then it isn't blowing air out the back of the case in huge amounts.

All the MSI 560 TIs that I can see are like that.

These are called dispersal fans, they push air straight down onto the heat sink and they don't care where the air goes after it goes through the heat sink. Usually that means that most of it stays inside the case.

Exhaust fans would have a single fan on the end opposite of the exhaust port and they direct a single stream of air long ways across the whole card and out the back.

With the plastic that is around your heat sinks, it is likely that the side fan would not be blowing air directly on any fins. The job it would be doing would be increasing the heat dispersal beyond what the video card fans are doing. The video card fans would just try to get it directly off the heat sink on the video card which would keep it close'ish to the video card and the side fans would take that one step farther and push it farther away from the video card.

With this sort of video card cooling method, I would want to get the maximum air in and the maximum out to try to cycle the heat out more quickly.

The m^3/h has a lot to do with the size of the fans. Basically, the size of the fan dictates how much air it can move. In 1 RPM a 120mm fan will move more air than 1 RPM of an 80mm fan. A 200 mm fan will do a lot more than both the 120 and the 80.

It isn't a linear rate either. A circle with a radius of 100 mm is more than twice the area of a circle with a radius of 50 mm. Pi * r^2 and all that from geometry class.

The smaller fans have higher RPMs usually so they can kinda make up some of the difference, but that causes a lot of added noise as well.

This is all kinda complicated stuff.

The easiest thing to do is to use fans of the same size all over the case, then you can just count in vs out.

If you can't use fans of the same size, tell me which size fans you have where and I will try to advise more fully in that regard.

As a rule of thumb, you probably want the top fans to be running at the higher speeds if possible. Heat wants to go straight up, not upwards then sideways, then upwards. Top exhaust fans are better at getting heat out than upper rear exhaust fans are if they are the same size. One works against the natural tendencies of heat and the other not.

BTW, nitpicky detail, you probably want your CD/DVD drive to be in the lowest possible bay. Heat pools around it more easily if it is in the highest possible bay.
m
0
l
January 18, 2012 7:11:42 PM

Update 2: I'm using this case -> http://www.newegg.com/Product/Product.aspx?Item=N82E168... -> Antec P280. It came with (2) 120mm top Exhaust Fans and (1) 120mm rear exhaust fan. I added (2) 120mm intake fans to the front at the Hard Drive Bays. All fans are currently at full speed/power. Also, I have turned the PSU around so that the fan is facing downward and pulling air from outside the case.

MemTest86+ results: It successfully Passed 16 cycles of tests with zero errors. I think it ran for 16 hours (unfortunately, the picture I snapped befor turning it off in order to rearrange parts didn't catch the hour digits). Picture at http://imageshack.us/photo/my-images/6/img0084ahf.jpg/

I have rearranged various parts of my machine now.
1) The CD drive, at the request of Raiddinn, has been moved to the lowest possible bay.
2) I have separated the Hard Drives more. There is now 1 bay of space between each HD.
3) The PSU has been flipped to suck air in from outside the bottom of the case.
4) Because of the new PSU arrangement, I was able to feed a few more cables through the Cable Routing holes in the back of the case to clear up space. Hopefully this helps with Air Flow.
5) I have set all fans to their maximum speed and placed the side cover back on.

My computer currently looks like this: http://imageshack.us/photo/my-images/513/img0088xd.jpg/

Now, since I have put the cover back on, I have booted the computer up twice from a completely off state. Each time, it has booted to Windows 7 and lasted for around 5 minutes, then cycled the power off and on. After cycling power, it has stayed on until I have shut it down manually....so....progress? Hooray no second random restart!

I have also checked my CPU temperature (VIA Core Temp) and no Core is getting higher then 33 degrees Celsius.

What's next? Are we at the point where I need to re install Windows 7?
m
0
l
January 18, 2012 7:26:10 PM

Owensboro said:



Now, since I have put the cover back on, I have booted the computer up twice from a completely off state. Each time, it has booted to Windows 7 and lasted for around 5 minutes, then cycled the power off and on. After cycling power, it has stayed on until I have shut it down manually....so....progress? Hooray no second random restart!

I have also checked my CPU temperature (VIA Core Temp) and no Core is getting higher then 33 degrees Celsius.



I have never come across anything exactly like this before. Strange behavior is often linked to a bad PSU though. It would be theoretically possible for the PSU to hit a specific temp which causes out of spec voltages and so creates this event. Then, as the PSU warmed more the problem would not re-appear.

Obviously you would test that by swapping out the PSU.... and it's just a wild theory at this point.

The other thing I see here is the successful overnight memtest. This makes me think it might be the GPU that is at fault, since there is no stress at all on the GPU during that process.
m
0
l
January 18, 2012 8:24:23 PM

He has an XFX Core 550w Pro which is a Seasonic OEM PSU.

It doesn't sound like a PSU problem based on that.

It is possible that it was somehow damaged in the shipping process or something, but the XFX PSUs generally have very low RMA rates.

If the computer made it through 16 passes with Memtest86+ it probably isn't the RAM either.

On a separate note, is there not any way you can run that CPU power cable behind the motherboard? It is sticking out like a sore thumb.

Have you tried to boot into safe mode and tried to crash it?

If not, turn it all the way off and then turn it back on and immediately boot it into safe mode. You can hit a key like f8 or f12 to bring up an options menu if necessary.

Then see if it crashes after 5 minutes or any amount of time.
m
0
l
January 21, 2012 9:22:35 PM

So after getting the computer to work for a bit, I was scared to restart it and try safe mode. This morning I finally powered it down and let it rest for a few hours. I just now tried the power button on my case and it didn't turn on. The case fans jerked, but never spun. It still won't turn on, even after resetting the power supply.

so, is this what a grounding issue looks like? Or is the PSU bad? I have no idea what to do.
m
0
l
January 21, 2012 9:29:56 PM

Do you have another equal PSU laying around that you could try?

How about an old video card?

Also, you should try hitting the power button every time the computer turns itself back off. Sometimes if a PSU is failing it will work if you do this. If it does work when you do this you probably need a new PSU.

Do you have a graphics port on the motherboard that you could plug the monitor into? If so, take out the video card and try it with onboard video from the 2500k instead.
m
0
l
February 2, 2012 11:44:33 AM

Alrighty. I think I have solved the problem. I didn't want to post until I had given it a few days of testing (power off, run for two days, power off multiple times in one day, etc..) so i'm sorry for the late response.

I replaced the Power Supply with a Corsair TX650M and now everything works 100% fine. I find this funny because the last time I built a computer, the Power Supply was the weak link in the entire system. Of course, the last time it was because I had no idea that Graphics Cards needed so much power to run.

I went from a 550W to a 650W PSU because of everyone's previous comments about XFX being a great brand and it probably not being broken. I figured the only thing I hadn't tested was more power. Thank god it worked. I'm thinking that perhaps because I had so many extras connected to the thing I was drawing more power then it could handle.

Thanks so much for the help everyone. There is absolutely nothing worse that building a new computer and it working just long enough for you to think you dodged any problems, only to start randomly screwing up and you have 0 idea what went wrong.
m
0
l
February 2, 2012 6:49:51 PM

Yep, any PSU can be bad, especially after being tossed around by UPS. It's been a rule here for a while... the more bizarre the symptom the more likely it's the PSU.
m
0
l
May 30, 2012 3:09:13 AM

Owensboro said:
Alrighty. I think I have solved the problem. I didn't want to post until I had given it a few days of testing (power off, run for two days, power off multiple times in one day, etc..) so i'm sorry for the late response.

I replaced the Power Supply with a Corsair TX650M and now everything works 100% fine. I find this funny because the last time I built a computer, the Power Supply was the weak link in the entire system. Of course, the last time it was because I had no idea that Graphics Cards needed so much power to run.

I went from a 550W to a 650W PSU because of everyone's previous comments about XFX being a great brand and it probably not being broken. I figured the only thing I hadn't tested was more power. Thank god it worked. I'm thinking that perhaps because I had so many extras connected to the thing I was drawing more power then it could handle.

Thanks so much for the help everyone. There is absolutely nothing worse that building a new computer and it working just long enough for you to think you dodged any problems, only to start randomly screwing up and you have 0 idea what went wrong.



This might be my problem as well, but I'm running my computer through a backup/power surge utility rated at 450w although I'm using a 750w PSU. Could this be my problem?
m
0
l
!