PC freezes with a buzzing/looping sound, any tips?

[Fenix]

Seems I can't go long without feeling stupid. I assumed this was a HDD issue and naively put it off since it appeared not to be good in CrystalDiskInfo. Well after cloning it to a 'good' drive the freezing with sound happened while I was off watching TV. Not sure if it was at screensaver or shut the monitor off. As far as I recall I had two browsers open, and one had a flash game up.

So far I've done the following;
-Did a memtest86+ for 1 pass - no errors. Plan to do another, even longer test, while I sleep just to be sure.
-I did a prime95 for 20 minutes or so and had no errors. Should I do a longer test?
-According to CPUID HWMonitor my CPU is in the 32-60 range, depending on what I'm doing. GPU I've never seen above 50, I think. Think that's fine. Right now with just steam, firefox, and a notepad/HWMonitor. The CPU is at 32, GPU is at 29.

So far tonight I've put my CPU back to stock, and updated video drivers. At a loss at this point, hoping it was just a driver issue.. but I don't know what to test. Though could it be my motherboard? Not sure if the sensors are screwed or if HWInfo64 and CPUID HWMOnitor are not reliable with my GA-Z170-HD. But it shows the following under each for it's temps:

Under CPUID HWMOnitor:
TZ00: 28c
TZ01: 30c
TMPIN0: 31c (Based on HWINfo65 I assume this is "System")
TMPIN1: -54c (?)
TMPIN2: 32c (Based on HWInfo64 I assume this is "CPU")

Under HWInfo64:
PCH Temp: 37.5c
System: 31c
CPU: 32c
Temp 4: 87c (Is this HWMnitors TMPIN1? Surely if it was that hot I'd have some serious issues)


Such a bad time for potentially needing to sink more money into this thing. lol
Gonna go to bed and do a longer memtest, I'll check on this in the morning-ish.

 

[keefy]

You got any molex connectors in your system? They can often be wired differently and cause issues with hardware receiving power on the wrong pins.

 

[PBM]

Download HD Tune and run an error scan with the "quick" option enabled: http://www.hdtune.com/download.html

It's not free, but it's a fully functioning trial. The use of the error scan is permanently free though.

This looks for bad sectors on the drive and whatnot and does it very quickly. You can run the full test (quick option unchecked) if you're going to be doing something else for a few hours because it takes a while.

If you get any blocks that aren't green at all (yellow is not good, red is very ver bad), you should change your drive. You should probably be looking into changing your drive anyway if you're already starting to see bad joojoo with crystaldisk.

 

[MjW]

Does your motherboard have integrated graphics? If yes remove the graphics card and go with the integrated to make sure it's not the gfx card.

 

[Fenix]

[QUOTE="keefy, post: 6523968]You got any molex connectors in your system? They can often be wired differently and cause issues with hardware receiving power on the wrong pins.[/QUOTE]

Just my case fans are via molex.

[QUOTE="Brandon, post: 6523973]Download HD Tune and run an error scan with the "quick" option enabled: http://www.hdtune.com/download.html

It's not free, but it's a fully functioning trial. The use of the error scan is permanently free though.

This looks for bad sectors on the drive and whatnot and does it very quickly. You can run the full test (quick option unchecked) if you're going to be doing something else for a few hours because it takes a while.

If you get any blocks that aren't green at all (yellow is not good, red is very ver bad), you should change your drive. You should probably be looking into changing your drive anyway if you're already starting to see bad joojoo with crystaldisk.[/QUOTE]

I've already taken the bad seeming drive out because I hoped that was the problem. Today the computer froze on me while playing a silly flash game and browsing in another browser. Didn't do the "Freeze with sound" thing yet, though. So I guess it's progress?

Quick scanned my current primary and all is, sadly, good. If it's my Motherboard or PSU I'm a little fucked - Motherboard = PC-less, got no money for a new PSU.


Could something be corrupted from my apparent failing drive? Or is this just wishful thinking.

[QUOTE="MjW, post: 6523974]Does your motherboard have integrated graphics? If yes remove the graphics card and go with the integrated to make sure it's not the gfx card.[/QUOTE]

If it was a graphics card wouldn't it be a near constant? Legit asking, I wouldnt assume so but I've never come across this issue before either.

I've heard the GTX 1070/1080s don't have thermal pads, and have sent away for the "thermal mod" as they call it. Could the VRM part of my card get REALLY hot to cause issues despite the thing only being 30c in areas with sensors?

Guess I'll try taking out the GPU after some more tests, the thought of having to RMA something makes me sad.

 

[Fijiandoce]

[QUOTE="Fenix, post: 6523984]I've heard the GTX 1070/1080s don't have thermal pads, and have sent away for the "thermal mod" as they call it. Could the VRM part of my card get REALLY hot to cause issues despite the thing only being 30c in areas with sensors?[/QUOTE]
The EVGA thermal pad issue affects their ACX coolers (which is essentially all their cards tbh), EVGA were apparently footing the bill for those affected cards, so at the very least, it may be worth following up on it regardless?

As for the lock ups. It sounds strange. I would have said the Hard drive was on its way out, but having switched it, i'd guess it was either a software issue, or something hardware related (which would be strange since your temps are quite low).

Have you run the "sfc /scannow" command to verify windows files? If the old drive was bad, and you merely cloned it over, corrupted files would have carried over also.

 

[robvandam111]

A buzzing sound? My computer used to freeze with a buzzing sound whenever I was watching videos from youtube at first. Just a couple of weeks ago it happened twice while I was watching a movie from Windows Media Player.

 

[MacP]

If worst comes to worst get a Windows disc and just format it and re-install Windows. 9 times out of 10 it solves most problem's just backup your data

 

[keefy]

[QUOTE="MacP, post: 6523994]If worst comes to worst get a Windows disc and just format it and re-install Windows. 9 times out of 10 it solves most problem's just backup your data[/QUOTE]

If not got a disc You can download it from Microsoft direct to a USB and even use the image to repair rather than reinstall completely.

https://www.microsoft.com/en-gb/software-download/windows10

 

[PBM]

I'd recommend a clean install. There could corrupt drivers or who knows what.

 

[Fenix]

Welp, happened again while I was out; been so long I was hoping it was over. Went out and left two flash games open and came back to my computer frozen and buzzing. Shame I can't see temps in the task bar.

Never paid attention before but is it normal for motherboards in temp monitoring programs to have a negative number or a wonky number? TMPIN1 is -54 under CPUID HWMonitor. Under HWiNFO64 Temp 4 is 89C, but considering everything else is <35 that must be HWMonitors TMPIN1.

[QUOTE="Fijiandoce, post: 6523990]The EVGA thermal pad issue affects their ACX coolers (which is essentially all their cards tbh), EVGA were apparently footing the bill for those affected cards, so at the very least, it may be worth following up on it regardless?

As for the lock ups. It sounds strange. I would have said the Hard drive was on its way out, but having switched it, i'd guess it was either a software issue, or something hardware related (which would be strange since your temps are quite low).

Have you run the "sfc /scannow" command to verify windows files? If the old drive was bad, and you merely cloned it over, corrupted files would have carried over also.[/QUOTE]

Yeah, I just got the "thermal pad" "mod". I just need to get up the courage to take my GPU apart, but I can't see the VRM being that hot while browsing/running flash games. Just ran a 3Dmark and got 5,334 in Time Spy. Which should be about what to expect with a stock clocked CPU/GPU. is this an adequate test for a GPU? I heard furmark.. but I'd honestly rather not. lol

Just did a Prime95 "Small FFTs" Test:
Worker #1: Torture Test completed 52 Tests in 1 hour, 11 minutes - 0 errors, 0 warnings.
Worker #2: Torture Test completed 53 Tests in 1 hour, 11 minutes - 0 errors, 0 warnings.
Worker #3: Torture Test completed 58 Tests in 1 hour, 11 minutes - 0 errors, 0 warnings.
Worker #4: Torture Test completed 59 Tests in 1 hour, 11 minutes - 0 errors, 0 warnings.

Screensaver came on so I'm gonna assume that's why there is a difference and also why Temps got up to 66C. But that can't really be all that high considering most threads I see about high temps are in the 80s. So that rules out CPU problems, right?

Did a "sfc /scannow" now. No issues.


[QUOTE="Brandon, post: 6524005]I'd recommend a clean install. There could corrupt drivers or who knows what.[/QUOTE]

Yeah, it looks like that's gonna have to be what I do. Looks like it's time to backup my documents, ect. Was only planning to do a clean install when I got an SSD. lol

 

[-Dj-]

have you try buying a better pc?

 

[keefy]

Perhaps the program doesn't fully support your motherboard.

Battlefield 4 used to do crash and lock my system with repeating funky noises,I had a rather large over clock at the time so reduced it. After doing that it stopped crashing.

Is flash fully up to date?

 

[keefy]

If your power supply is modular check you are using the correct cables for that make.

Some info in this video
https://youtu.be/SDsC_PNo84I

 

[Fijiandoce]

[QUOTE="keefy, post: 6524558]If your power supply is modular check you are using the correct cables for that make.

Some info in this video
https://youtu.be/SDsC_PNo84I[/QUOTE]
That video was the result of this guy:
https://www.youtube.com/channel/UCmbkRUS_4Efdt5UIhwNqtcw

His PC went AWOL, and while trying to diagnose, killed the entire system (two systems actually). He concluded it was the result of the Hue+, so sent it off to NZXT, who sent it off to GN who concluded that Greg (the guy with the PC) used the wrong cable.

Which prompted this video:
[video=youtube;b3mmfNErFWU]https://www.youtube.com/watch?v=b3mmfNErFWU[/video]

NZXT were boss though, iirc, they handled the issue very well. I think they replaced his parts, as well as paid the shipping fees for the PC.

But since this thread has been bumped; any updates Fenix?

 

[PBM]

I've had nothing but good experiences with NZXT. Their customer service is awesome.

 

[Fenix]

Didn't realize it's been so long. But I do have an update, and it doesn't really make much sense to me due to the 20-60 min memtest bringing no errors up, but after doing a complete reinstall of Windows 10 it ended up freezing the next day. I had a flash game up and I ran to the store and came back to buzzing.

Not sure why but I got the brainstorm to put my RAM to stock (So everything in my computer is now "stock") and well.. it's been 4 or 5 days now with no issues. I'm feeling confident, so far, that was the issue but I can't say 100%. (EDIT: Last "unexpected shutdown" that I can see was ‎2016-‎12-‎13 at 10:59:20 PM)

Though, if it is the issue.. is that a RAM issue or a Motherboard issue?
Could I have stumbled upon some kind of motherboard/ram incompatibility thing? Which I didn't even know was possible but I guess that's what I get for going with an unknown company cause it was cheap and had blue LEDs. (which I thought was pretty awesome and didn't know was a "thing".)

For reference, my RAM is 2x 8GB Avexir Core Series DDR4 (AVD4UZ124001608G-2COB) and the Motherboard is a Gigabyte GA-Z170-HD3

 

[PBM]

You could try different RAM ... you could try a different GPU ... or remove the GPU entirely. You could try swapping the PSU. Hopefully it's not the motherboard.

How's the thermal paste on the CPU?

 

[keefy]

Over clocking varies from system to system.
Also if you have instability after an over clock its best to reduce it until it goes away. If we knew you had over clocked your RAM, pretty sure someone would have told you to remove it long ago.

 

[Fijiandoce]

[QUOTE="Fenix, post: 6524654]Didn't realize it's been so long. But I do have an update, and it doesn't really make much sense to me due to the 20-60 min memtest bringing no errors up, but after doing a complete reinstall of Windows 10 it ended up freezing the next day. I had a flash game up and I ran to the store and came back to buzzing.

Not sure why but I got the brainstorm to put my RAM to stock (So everything in my computer is now "stock") and well.. it's been 4 or 5 days now with no issues. I'm feeling confident, so far, that was the issue but I can't say 100%. (EDIT: Last "unexpected shutdown" that I can see was ‎2016-‎12-‎13 at 10:59:20 PM)

Though, if it is the issue.. is that a RAM issue or a Motherboard issue?
Could I have stumbled upon some kind of motherboard/ram incompatibility thing? Which I didn't even know was possible but I guess that's what I get for going with an unknown company cause it was cheap and had blue LEDs. (which I thought was pretty awesome and didn't know was a "thing".)

For reference, my RAM is 2x 8GB Avexir Core Series DDR4 (AVD4UZ124001608G-2COB) and the Motherboard is a Gigabyte GA-Z170-HD3[/QUOTE]
How hard were you pushing the RAM overclock?

It could be the RAM timings are a bit too tight, or maybe there is a BIOS update for the Mobo that addresses some issues with power delivery?

 

[Fenix]

It wasn't really an overclock, or so I thought, considering all DDR4 is 2133 unless "overclocked". I merely turned on XMP to make them 2400 which is what newegg/the manufacturer says they are. Didn't really think of it either since I ran a few memtests are varying amounts of time that showed no problems.

Would this be a RAM issue, or a motherboard issue? Is it even worth bitching about?

EDIT: BIOS is "F6", only one passed that is "F20" but that's for "Support next generation CPU
* This BIOS prohibits updating to earlier version BIOS"


[QUOTE="Brandon, post: 6524661]You could try different RAM ... you could try a different GPU ... or remove the GPU entirely. You could try swapping the PSU. Hopefully it's not the motherboard.

How's the thermal paste on the CPU?[/QUOTE]

What's the GPU/CPU have to do with putting the ram to "stock" potentially fixing it? Legit curious, this entire thing has been whack as hell.

Don't have the money to just start buying parts to swap. Pretty much put all my spare cash into buying the 1070 SC.

 

[Fijiandoce]

[QUOTE="Fenix, post: 6524683]It wasn't really an overclock, or so I thought, considering all DDR4 is 2133 unless "overclocked". I merely turned on XMP to make them 2400 which is what newegg/the manufacturer says they are. Didn't really think of it either since I ran a few memtests are varying amounts of time that showed no problems.

Would this be a RAM issue, or a motherboard issue? Is it even worth $#@!ing about?

EDIT: BIOS is "F6", only one passed that is "F20" but that's for "Support next generation CPU
* This BIOS prohibits updating to earlier version BIOS"[/QUOTE]
Do you have CPU-Z installed? It should provide a list of XMP timings that the RAM supports as well as the Voltage for said ram.
Like this:

If the voltage from the 2133 clock is different to the 2400 clock, it may just be a power delivery bug that is exposing something with the hardware. You could try messing with the Voltages manually to see if you can remedy this if that is the case, or ensure that you are getting the correct voltage set. Mobo may mess up and keep the voltage for 2133, or may fluctuate voltage is set to 'auto'. Which may be the issue when the PC tries to enter a low power state.

Outside that, reviews from people on Newegg suggests that seating the RAM modules in their slots is a bit of a chore, could require a bit more force to ensure they are seated correctly... though i personally don't understand how this could result in booting one moment, crashing another, but worth looking into.

 

[Fenix]

Yeah, I got CPU-Z installed. This is what I get when xmp is off:

Went to turn on XMP and I saved/restarted and as it rebooted I got a beeps and a boot failure screen. Got options like optimized settings or go to bios, so I went to bios and turned XMP off and then it booted back up fine.
According to http://www.gigabyte.com/support-downloads/faq-page.aspx?fid=816 , assuming it's right I'm going to guess it was a long beep and a short beep which indicate a memory error. (I don't know the difference between AMI and AWARD bios, and I can't recall if I heard more than two beeps.)

Wanted a picture of the boot failure screen so I tried it all again and this time it didn't happen again. Bios says my ram is 2400 Mhz, but CPU-Z is exactly the same as the picture above.

Going to try turning XMP off again and double checking CPUZ, then I'll try reseating the RAM. Maybe I got 'em in backwards or something - I don't know. lol

 

[Fijiandoce]

[QUOTE="Fenix, post: 6524689]Yeah, I got CPU-Z installed. This is what I get when xmp is off:

Went to turn on XMP and I saved/restarted and as it rebooted I got a beeps and a boot failure screen. Got options like optimized settings or go to bios, so I went to bios and turned XMP off and then it booted back up fine.
According to http://www.gigabyte.com/support-downloads/faq-page.aspx?fid=816 , assuming it's right I'm going to guess it was a long beep and a short beep which indicate a memory error. (I don't know the difference between AMI and AWARD bios, and I can't recall if I heard more than two beeps.)

Wanted a picture of the boot failure screen so I tried it all again and this time it didn't happen again. Bios says my ram is 2400 Mhz, but CPU-Z is exactly the same as the picture above.

Going to try turning XMP off again and double checking CPUZ, then I'll try reseating the RAM. Maybe I got 'em in backwards or something - I don't know. lol[/QUOTE]
That is strange, you have 2400MHz RAM. Skylake can run up to 4000MHz RAM.

From the looks of it, you aren't even getting the advertised 2400Mhz that the RAM should be running at, at least as far as the part listed goes.

On the Intel Ark page, Intel lists that i5 as having up to 2133MHz data range, at stock. This upper limit is coincidentally what your RAM is running at. Have you tried looking a little deeper in the bios? Unlike a CPU where you only really need to watch heat(voltage), RAM generally has a few factors that determine how it runs - Like the memory controller.

This is generally why people shy away from RAM overclocking :snicker

I know Asus boards have this feature, but not sure on Gigabyte. But if there is an 'Auto' overclock utility in the Bios, that also sets DRAM clocks, perhaps give that a try just to see if it sticks. If it does, then perhaps it is a case of something not being fed what it needs.

 

[Fijiandoce]

Forums are being a pain in the arse, so need to double post:

Just checked the Mobo spec sheet, it lists 2400Mhz as *(OC). Indicating you need to do some finagling to get it up to speed. XMP should set the RAM up, but wont adjust anything else, as said above, Memory has a few bits and bobs that it would like to help it run faster.

 

[Fenix]

Any idea what I need to do besides turning XMP on? Cause when I do BIOS reads it at 2400 but CPUID apparently lists it at stock.


Shouldn't have to overclock my RAM to get advertised speeds. -_-

 

[Fijiandoce]

[QUOTE="Fenix, post: 6524696]Any idea what I need to do besides turning XMP on? Cause when I do BIOS reads it at 2400 but CPUID apparently lists it at stock.


Shouldn't have to overclock my RAM to get advertised speeds. -_-[/QUOTE]
Not up to speed on newer Intel's, last one i owned was the Q6600 iirrc

However, i found this article:
http://www.tweaktown.com/guides/748...-intel-skylake-overclocking-guide/index5.html
That does a little rundown on a general list of things to overclock including memory controller voltages and such.

A table from the page as a visual of the ranges to set:

There are two (VCCIO, and VCCSA) that pertain to ram overclocking, so maybe start with these.

You shouldn't need to go crazy. You are able to boot, and use the PC at times, so whatever it needs, may be running marginal at 2400MHz, and only in need of a small increase.

 

[Fenix]

That sounds more complicated than I care to do. Guess I'll live with 2133, at least I'm happy I found the issue. Was worried it was a GPU issue and sending in a $400+ card for a refurb was gonna piss me off.

 

[Fijiandoce]

The silver lining is that there isn't much difference between 2133, and 2400. So it'll be unnoticeable. In games it'll cost maybe a single frame? And for real world stuff, it might as well be DDR3 :snicker

If you ever cycle around to it however, you'll know where to start looking

 

[Fenix]

Mostly went DDR4 cause I wanted to go intel and skylake was the latest and these sticks were like $60. The blue LEDs seemed neat, though I'm starting to dig the white from my GPU more. lol

What's more likely for my XMP profile not working properly, RAM or motherboard? Probably not an issue worth RMAing over, right?


Thanks for the time, folks. Didn't expect DDR4/skylake to be such a pain. I miss the DDR2/3 days where ram was plug and play. lol