New Pc Build Constantly Freezing

[Draebesonrm]

So I've posted on a few threads about this but I don't know what to do, i don't know what i'm doing wrong if it's a hardware issue or a driver issue, more maybe something just isn't connected right.

Honestly this weekend I'm taking mostly everything apart and trying to find out what's going on, I didn't put the cpu in the motherboard, or connect any of the major wires my friend did that for me..so it's possible he did something wrong, idk.

But basically what was happening was I was getting random lockups where I couldn't do anything except shut down the computer manually. My mouse and keyboard didn't work and everything just completely locks up. Before I was getting blue screens, but now I just get the freezing (at least so far it's been about a week since I replaced the RAM and HDD).

I thought my old storage device was the problem or the RAM so I just replaced both, the RAM is the same brand Crucial Ballistix Sport DDR3-1600, but I got a WD Blue 1tb HDD instead of the Samsung SSD I had in there.

So after I replaced this two things, everything seemed fine for the past week, and then just today it started freezing on me again. While idle, and while gaming three times so far. No BSOD's yet but I'm waiting to see.

So all I can think now is that it must either be my mobo, processor, or psu. Now I'm gonna just get a different psu to try out in there because I needed an extra one anyway. But as far as the mobo and the processor go I don't want to trade those in if I don't have to, considering I'd have to get windows registration figured out with a new mobo which might be a hassle, idk never done it before.

So I'm trying to figure out which one of those three it could be. My temps are fine I ran stress tests and my cpu never goes about 50 at 100%, my gpu goes pretty damn high though around 70, but the cooling in my case isn't the greatest right now.

Voltages are
CPU Core AVG - 0.5
+3.3v AVG - 3.332
+5v AVG - 5.066
+12v AVG - 12.051
DIMM AVG - 1.488

now my RAM is rated for 1.5v 9-9-9-24 and I set the voltage at 1.5 manually but it won't run at 1.5 idk if that's a problem or not, but otherwise the RAM is set properly as far as I know.

If theres anymore info that would be helpful in figuring this out please let me know I'm happy to post it, but this is all I can think to put right now. I'm just trying to see if I can figure out exactly what is broken (if it's hardware) or not working properly (software, drivers). So I don't have to RMA the whole machine just to find out it was a software issue.

My Build
i5 4670k
Corsair H75 water cooler
MSI Z87-GD65
WD Blue 1tb HDD
EVGA GTX 660
Corsair Carbide 200r
Corsair CX750M
Crucial Ballistix Sport 2x4gb ddr3-1600 cl9

 

[Powerbolt]

Your CPU probably doesn't like that your memory voltages are so high. I know my CPU (4770K) will throttle my memory modules if the voltage goes over ~1.35V ish. Is there not a utility in the BIOS that automatically overclocks your memory based on it's configuration (X.M.P Profiling I believe)?

Might check your SATA ports on the board as well. Sounds stupid probably, but I've had a burnt out one cause all sorts of mayhem on my system. In fact it wasn't so long ago it was freezing, loading things at a snail's pace, etc.

 

[bgunner]

before trying new hardware run these free programs and see if it is a hardware issue.

1. run this one first, Memtest86+. burn it to a disk of flash drive and boot from that media drive. let it run for 11 passes. this will take time so don't rush these tests. this tests the memory
2.Intel Burn Test, run this on maximum stress level and for 20 runs. tests the CPU
3. Prime 95, run this for 5 hours minimum. tests the CPU yes run both because one may not catch the error the other one will.
4. OCCT, run this test only for the GPU with error check enabled. this tests the GPU.
5. HDDscan, this checks the smart features of the HDD.

It also could be a bad setting in the BIOS that is causing this issue and as you mentioned before it can be a software conflict. I like to start with hardware to rule out and bad piece there as these are easier to find and tell if they are bad.

 

[jcol87]

before trying new hardware run these free programs and see if it is a hardware issue.

1. run this one first, Memtest86+. burn it to a disk of flash drive and boot from that media drive. let it run for 11 passes. this will take time so don't rush these tests. this tests the memory
2.Intel Burn Test, run this on maximum stress level and for 20 runs. tests the CPU
3. Prime 95, run this for 5 hours minimum. tests the CPU yes run both because one may not catch the error the other one will.
4. OCCT, run this test only for the GPU with error check enabled. this tests the GPU.
5. HDDscan, this checks the smart features of the HDD.

It also could be a bad setting in the BIOS that is causing this issue and as you mentioned before it can be a software conflict. I like to start with hardware to rule out and bad piece there as these are easier to find and tell if they are bad.

Since you already replaced ram and hdd it would be strange if both new ones were faulty.

Do you have spare psu to try, do you have onboard graphics, if so you can remove gpu and uae lower wattage ps, if nothing fails thwn there's the problem

Those tests may not tell you much since it will probably crash before findinf anything, it doesn't spund like a driver isue, but ir couldnt hurt to roll back ur graphics drivers to check

Disconnect ghe case fans when u test the other psu, if it still fails then the problem isbprobably cpu or mobo, but then again once you reinstall everything and rwconnect it might work fine

 

[Draebesonrm]

Yeah I've already got memtest and a bunch of other diagnostic stuff burned on a disc that I use, I just got done running it for a couple hours and it passed 4 times with no errors, so I'm gonna leave it on tonight anyway just to be sure that should give it enough time to do a bunch of passes.

I haven't tried prime95 or intel burn test yet but I have been using AIDA64, as long as I'm not pushing my graphics card too, because that was when it was running 70 celsius, I had it at 100% for 10 minutes the last time and the temps never went above 50 and it was stable, but I do need to run it for longer, and I'll try it with those other programs too.

I've never used OCCT before though so I'll have to check that out. I am gonna have an extra psu this weekend but it's gonna be brand new out of the box so I can't say for sure if it's gonna work or not either lol. I'm buying it for another computer but I can always try it in this one first. But I'd hope that if it is the psu that Corsair wouldn't sell me two bad psu's in a row, especially since I've heard such good things about some of their psu's, not mine in particular but some.

and also I didn't think about possibly lowering the voltage to the RAM, I was assuming it was possibly not getting enough power but I never thought about too much, I'll look into that too. Thanks for all the suggestions I'll let you guys know tomorrow how memtest went overnight as well as all the other tests.

 

[bgunner]

Occt is only good for the GPU test. This is because the other tests are very unreliable and do not seem to catch the errors like the other programs I mentioned. It is very accurate for GPU errors and is great tool when OCing the GPU to tel if you have gone to far or not.

 

[bgunner]

before trying new hardware run these free programs and see if it is a hardware issue.

1. run this one first, Memtest86+. burn it to a disk of flash drive and boot from that media drive. let it run for 11 passes. this will take time so don't rush these tests. this tests the memory
2.Intel Burn Test, run this on maximum stress level and for 20 runs. tests the CPU
3. Prime 95, run this for 5 hours minimum. tests the CPU yes run both because one may not catch the error the other one will.
4. OCCT, run this test only for the GPU with error check enabled. this tests the GPU.
5. HDDscan, this checks the smart features of the HDD.

It also could be a bad setting in the BIOS that is causing this issue and as you mentioned before it can be a software conflict. I like to start with hardware to rule out and bad piece there as these are easier to find and tell if they are bad.

Those tests may not tell you much since it will probably crash before findinf anything, it doesn't spund like a driver isue, but ir couldnt hurt to roll back ur graphics drivers to check

Disconnect ghe case fans when u test the other psu, if it still fails then the problem isbprobably cpu or mobo, but then again once you reinstall everything and rwconnect it might work fine

Actually if the test fails or the system locks, this helps point to a bad piece of hardware. They will tell you a lot if you know how to trouble shoot. If the system locks on all tests except memtest86+ then it would be either software or motherboard. If it only locks when the CPU test is run this points to Bad Bios setting or bad CPU or Ram, this is why the memory needs to be tested first to rule out memory. If the system only locks on the GPU test this tells you that the GPU is the issue and is either OCed to far, the GPU is bad or Driver issue. Now if the program just errors it tells you the issue is in the same place but not severe enough to lock or crash the system.

As you can see there is a lot you can tell from these stress tests.

Out of pure curiosity why disconnect the case fans? They have nothing to do with the system locking, crashing, stuttering or anything else. They are a 12 volt and a 5 watt max draw a piece so this will not effect a powerful PSU like a 750W.

 

[jcol87]

Actually if the test fails or the system locks, this helps point to a bad piece of hardware. They will tell you a lot if you know how to trouble shoot. If the system locks on all tests except memtest86+ then it would be either software or motherboard. If it only locks when the CPU test is run this points to Bad Bios setting or bad CPU or Ram, this is why the memory needs to be tested first to rule out memory. If the system only locks on the GPU test this tells you that the GPU is the issue and is either OCed to far, the GPU is bad or Driver issue. Now if the program just errors it tells you the issue is in the same place but not severe enough to lock or crash the system.

As you can see there is a lot you can tell from these stress tests.

Out of pure curiosity why disconnect the case fans? They have nothing to do with the system locking, crashing, stuttering or anything else. They are a 12 volt and a 5 watt max draw a piece so this will not effect a powerful PSU like a 750W.

I was saying that about the fans because if the PSU is faulty then any little fluctuation in power could affect the system. Obviously, at that wattage it's such a small difference that it probably won't make a difference, but I was just being cautious. I also said to do that if you test another, different power supply and I was saying to disconnect the fans if the backup was lower power. You know, if you wanted to just check mobo, cpu, ram you would only need like an 250 watt PSU.

As for the tests, I know it sounds counterintuitive because obviously the whole point is for it to FAIL and reveal the problem, I was just saying that it might not reveal it because one test will fail because of a different component that is not being checked. Though clearly, he is testing everything, and memtest obviously just tests the RAM because the computer hasn't fully booted up. But stuff like IntelBurnIn and Prime95, which are for the CPU, still use other components. I am not too familiar with them though.

I just noticed the OP is using a K-series CPU, are you overclocking? If so, check your voltages, you may need more or to disable certain power states. Also, try standard settings. I don't why anyone would get a K-series CPU and not overclock though.

 

[Draebesonrm]

Actually if the test fails or the system locks, this helps point to a bad piece of hardware. They will tell you a lot if you know how to trouble shoot. If the system locks on all tests except memtest86+ then it would be either software or motherboard. If it only locks when the CPU test is run this points to Bad Bios setting or bad CPU or Ram, this is why the memory needs to be tested first to rule out memory. If the system only locks on the GPU test this tells you that the GPU is the issue and is either OCed to far, the GPU is bad or Driver issue. Now if the program just errors it tells you the issue is in the same place but not severe enough to lock or crash the system.

As you can see there is a lot you can tell from these stress tests.

Out of pure curiosity why disconnect the case fans? They have nothing to do with the system locking, crashing, stuttering or anything else. They are a 12 volt and a 5 watt max draw a piece so this will not effect a powerful PSU like a 750W.

I was saying that about the fans because if the PSU is faulty then any little fluctuation in power could affect the system. Obviously, at that wattage it's such a small difference that it probably won't make a difference, but I was just being cautious. I also said to do that if you test another, different power supply and I was saying to disconnect the fans if the backup was lower power. You know, if you wanted to just check mobo, cpu, ram you would only need like an 250 watt PSU.

As for the tests, I know it sounds counterintuitive because obviously the whole point is for it to FAIL and reveal the problem, I was just saying that it might not reveal it because one test will fail because of a different component that is not being checked. Though clearly, he is testing everything, and memtest obviously just tests the RAM because the computer hasn't fully booted up. But stuff like IntelBurnIn and Prime95, which are for the CPU, still use other components. I am not too familiar with them though.

I just noticed the OP is using a K-series CPU, are you overclocking? If so, check your voltages, you may need more or to disable certain power states. Also, try standard settings. I don't why anyone would get a K-series CPU and not overclock though.

I'm not overclocking yet but I was planning on it. I've just been putting it off until I could get everything else working properly, I just didn't want to add to the problems. Once I get this figured out I'm gonna at least run it at 4.3 or 4.4, but that's only if I can't stay stable at 4.6. But that's for another time.

Anyway ran memtest came up with 11 passes 0 errors, so my new RAM is at least ok. Ran Intel Burn Test as specified above and that went fine, and I ran AIDA64 again. The only thing I haven't done yet is Prime95 but once I have time I'll do that next, but I think the CPU is ok.

I also still have to do the OCCT GPU test as well, but I'm starting to think it may be the graphics card. My SSD that I replaced was definitely bad, so maybe it was just a combo of a bad SSD and a bad VIdeo card. Because before I replaced the SSD I was getting freezes and blue screens(stop code 7a which points to memory) even when I wasn't gaming. But now I get no BSOD at all just the random freezing and only while gaming. I was able to play dishonored today at normal settings for 3 hours but I put a desk fan next to the tower and opened up the side and there was no freezing.

The EVGA GTX 660 only has one fan and idk maybe the heatsink isn't getting rid of heat like it should because on the normal graphics i was still running 70-75 degrees.

Regardless this weekend I will have an extra power supply, and next weekend I'll have my new graphics card. So I'll let you know the results of OCCT and Prime95 once I do them.

 

[Draebesonrm]

Well we're going on another day without any freezing, I even tried some newer games today(Arma 3, Bioshock Infinite) on high/very high settings but left the case open with the fan next to it and I played for a couple hours each with no problems.

So the only other thing I did was push on my 24 pin and cpu power connections on the mobo to make sure they were connected correctly. So either my gpu is overheating during gameplay, or my power connections weren't correctly attached and now I have them on there right.

But idk i'll wait and see and let you know if I have anymore freezing because it's only been a couple days.

 

[Powerbolt]

Well we're going on another day without any freezing, I even tried some newer games today(Arma 3, Bioshock Infinite) on high/very high settings but left the case open with the fan next to it and I played for a couple hours each with no problems.

So the only other thing I did was push on my 24 pin and cpu power connections on the mobo to make sure they were connected correctly. So either my gpu is overheating during gameplay, or my power connections weren't correctly attached and now I have them on there right.

But idk i'll wait and see and let you know if I have anymore freezing because it's only been a couple days.

It's entirely possible that your connections just weren't on there well enough. I've done that a time or two with my CPU power connectors which raised a noticeably lower temperature reading until I fixed the issue.

 

[bgunner]

There are a few free temp monitoring programs that are pretty good. All monitoring programs from time to time can give a bad reading like CPU is 132°c but most times they are obvious errors. CPUID's HWmonitor, Open hardware monitor are both good programs and are free to use. The will read out current, max and low temps. They will read CPU, motherbord, GPU, PSU temps and more.

for kicks and giggles try gaming again with the side panel on and see if it locks. IF it does open it up and put the fan back and see if it locks. Both times run one of these programs to keep an eye on temps.

 

[Draebesonrm]

Alright I'll put the side back on and do that today and let you know what happens.

When I switch video cards though (because I was planning on doing that anyway from the beginning because this gtx 660 was meant for another computer) I'm gonna be getting a MSI GTX 760 to go with my board.

My main question was when I switch the cards since I already have the latest drivers and they're both nvidia cards, do I need to delete the drivers and all that and do a clean install? Or do I just swap out the cards and they should work like that?

 

[bgunner]

Yes any time you swap a card, unless it is the same model, you need to reinstall the drivers.

You can go from a Sapphire 660 to a Power Color 660 and not reinstall drivers but you can not go from a 660 to a 760 and not reinstall the drivers. Driver installation is necessary when changing models.

 

[jcol87]

run a burn test on the gpu, like unigine and also crank up the fan speed in the control center, and watch the temperature changes, if things don't level off even at full speed then that is definitely your problem, but not something easily fixed by increasing ciruclaton in the case

the card cooling solution may not be functioning properly, other programs such as msi afterburner can adjust fan profiles and clock settings, try lowering the speed to reduce heat