Computer turns itself on and off...

johnsta13

Honorable
Sep 13, 2013
17
0
10,510
So for a month or so now, I have to hold down the power button to turn on my computer. If I simply press it, it turns on for a second, I hear the water cooler turn on, and then it shuts off. Today when I turned it on, it turned on, I heard everything boot up, and then it turned off. Then it turned on again and off again at random intervals. I flipped the switch on my PSU, pressed the button again to let out the residual power, counted to thirty, and then flipped the switch back. When I powered it up, it turned on. Later I shut down and the same thing happened except after toggling the PSU, it reset itself until I pressed the power button while it was on. Then it jumped out of the funny off-and-on mode and booted up. I'm really worried here. Is my PSU bad or do I have a bad switch?

For the record, I also have this thread open which is still unsolved: http://www.tomshardware.com/answers/id-2991759/psu-gpu-allocate-bytes-crash-games.html I don't know if theyre potentially related. My Arma 3 also keeps giving me a error that says "the instruction at 0x0(random different hexadecimal ending) referenced memory at 0x0(other random ending). The memory could not be read error. Do you think my PSU is not powering the card enough to handle the games and is also causing this weird startup issue?

I have MSI Afterburner logs I can post if y'all tell me how and want them. Please comment and tell me you're baffled and don't know whats up just so I know you read it:)

Specs:

PNY GTX 780ti @3Gb
i7 5820k
AS Rock X99 Extreme4 LGA2011-3 ATX Intel Motherboard
Silencer MK III 750W
16GB RAM
 
Solution
Do you have the most recent bios installed?

Have you tried booting with only one stick of memory or the other, to see if it's a bad module?

Are you overclocking, and if so, have you tried restoring to default settings? In some cases I've seen a few DDR4 systems that required a voltage increase even when XMP settings were employed.

Most of these on off cycles that don't make it past POST are memory related, but unstable overclocks and PSU issues are certainly suspect.

The fact that you're getting errors even with in windows makes me want to take a closer look at PSU voltages and also makes the GPU card a possible culprit as well. I know that doesn't narrow it down much, but doing a few of these things as well as posting HWinfo...
Do you have the most recent bios installed?

Have you tried booting with only one stick of memory or the other, to see if it's a bad module?

Are you overclocking, and if so, have you tried restoring to default settings? In some cases I've seen a few DDR4 systems that required a voltage increase even when XMP settings were employed.

Most of these on off cycles that don't make it past POST are memory related, but unstable overclocks and PSU issues are certainly suspect.

The fact that you're getting errors even with in windows makes me want to take a closer look at PSU voltages and also makes the GPU card a possible culprit as well. I know that doesn't narrow it down much, but doing a few of these things as well as posting HWinfo sensors window screenshots while running Furmark might help to eliminate some of the potential suspects.
 
Solution

johnsta13

Honorable
Sep 13, 2013
17
0
10,510




Is there a particular program I can use to write this into a txt? I'm not quite sure how too share the MSI Afterburner logs. Just tell me what to do and I'll do it :)



I will try booting with one stick of RAM. I have not overclocked, but I have tried underclocking the power by 5% like some other people did to fix the issue. It did not work though and I reset it to the defaults. I also had PNY RMA my card and it passed with flying colors and was sent back to me. I'll get those logs for you and make another post.
 
There is no memtest96. You mean Memtest86 and Memtest86+.

Run for a minimum of 3 passes. 7 would be better, but each pass takes a very long time. It will take hours and hours to run 7 passes, or even 3, and most memory problems, if they are bad enough to cause problems instantly when trying to POST or boot, will show up within the first pass. Even two passes should be enough to tell if there is a major problem with one of the modules.


http://www.memtest.org/

http://www.memtest86.com/download.htm


I do not think however, that if you cannot boot or enter the POST process by normal means, that running Memtest is going to tell us anything useful as a first look procedure.

Better would be trying different modules to see if one of them is causing the problem.

First though, what operating system are you running? If it's 10, try turning off hybrid sleep and hibernation in the control panel power applet advanced menu options.

 

johnsta13

Honorable
Sep 13, 2013
17
0
10,510

For the record, this fixed my startup problem. One of my two 8Gb sticks boots, the other does the weird thing. Thanks for that! Do you still think theres something up with my PSU? Also, this means I need to get another stick of RAM/deal without, right? I only just bought this set at Christmas from MicroCenter...



I had a GPU temp of 42 degrees Celcius, a CPU of 47, and a SSD temp of 27. Here are the screenshots of the HW logs. Perhaps it will help with the error?
nJ64Mcb.png
XmBloh7.png
5lMDz1v.png
iht5sZ4.png

 
If you're not still having other issues, then I doubt anything other than the memory was the problem.

What is the brand and model number of your memory? Most major memory manufacturers warranty their modules for life. G.Skill, Geil, Corsair, Mushkin, Crucial, Kingston, Patriot and I THINK even PNY, all have lifetime warranties so long as you have or have access to the purchase receipt or invoice. I'd contact Microcenter since it's hasn't been very long, first, and contact the manufacturer for an RMA, second.
 

bailojustin

Distinguished
Possibly your solid state drive is failing, to me just the fact that your SSD read/write speeds are lower then your HDD.

You should replace the SSD, please someone correct me if they might think otherwise.
Those speeds write/read on the SSD and the totals done are not good. most people don't expect a ssd or hdd bottleneck a computer, but it will do that, and slow it to a crawl.

I just read more into your problem as well, As a computer repair technician specifically hardware this is an often occur the boot loop eventually into booting up, this is a tell tale sign that the drive is immediately going bad. I would remove the SSD, try booting the computer and see if it boots up in 1 try. This is often the case.

Hope it helps, best regards.
 
Yeah, something looks fishy with those SSD speeds, but it's hard to say without actually running a benchmark for comparison that's going to really utilize the drive. Could be more than one reason for those readings. Could also be a problem with the SSD, especially if it's an older unit.

Kind of wondering about the use of an Apple SSD in a windows machine too. Partition/formatting differences? Is it even fully compatible?