Terrifying problem with high end computer Please HELP!

G

Guest

Guest
Hello, this is kind of a long story but i'll try to make it as short as possible. First i'll list my parts.
msi gtx680 sc 2gb gaming temperature 48-60c no higher
i5 2400 @ 3.4ghz with enhanced turbo gaming temperature 50-58c no higher
msi z77a gd65 motherboard
8gb corsair vengeance ram
corsair tx850w psu

Well, a few months ago I noticed that my psu was making an annoying ticking sound that is similar to two marbles hitting each other softly. I decided that I will RMA the power supply later on but I still used the computer for another few weeks. During those few weeks, I experienced a BSOD caused by caused by driver nvlddmkm.sys while gaming. After the blue screen, I decided to reinstall my driver, but for some reason the Nvidia 3d vision files refused to delete and I think was probably caused by me deleting it using this new software that I downloaded, which probably deleted it improperly. After trying to use driver sweeper several times to delete these files, I decided to give up and just let the nvidia installation do a clean install. A week later, my computer goes nuts. I start to get artifacting and driver crashing. I will be RMAing my ram and PSU tomorrow morning, but now I don't know if I'll have to RMA the GPU too since the psu screwed it up or something. I am in a really scary situation here and I hope that you guys give me some advice on how to deal with this problem and figure out what the hell is going on exactly. Since i'll be RMAing the psu and ram tomorrow I don't even have time to re install the operating system, and even if I had time, it would still be a bad idea to still use the computer with a possibly failing PSU (something that I should've thought of months ago when I heard that noise coming out of the PSU) .Anyway, what should I do at this point, besides crying myself to sleep?
 
G

Guest

Guest
It is not a guarantee that the PSU is bad. Just because it is making a sound doesn't mean that it is. Plus the readings for it in the bios are normal. Mocking me doesn't really help by the way so if you have nothing helpful to say then don't bother commenting and leave the thread open for others who might be capable of resolving the problem. Thanks.

3.3V I got 3.325
5V I got 5.003
12V I got 11.968
 
We always say the most important part of your build is the PSU. The reason we say that is because a bad or poorly manufactured PSU can go bad, and ruin other parts of your system at the same time. :pfff:

The FIRST thing you need to do is RMA the PSU. You cannot, I repeat, you cannot make any assumption at all about any other part of the machine until you have a properly working, quality PSU installed. Period. Once you have a properly working PSU in place, then you can start to troubleshoot for any other part of the machine that may have been damaged. But only then.

In the meantime no, don't even turn the machine on with the bad PSU in it.
 



This means absolutely nothing. And if it is making a sound, yes most likely its bad.
 
While the PSU noise is Not a good thing and a good canidate for replacing, That in itself does not "clearly" state PSU Caused any problem - YET.
1) Did you take a look at the PSU Voltages. Noticed you listed temps. (Added - See you looked at Idle)
.. PSU Voltages were: at Idle and max load (ie gaming).
.. Specs for +5 V are 4.75 -> 5.25 Volts, For +12 V they are 11.4 (I use 11.6) -> 12.6 V. For the +12 V Max decrease of 0.3 Volts when switching from Idle To heavey load.

2. You listed RMAing the Ram - So you ran Prime 95 (Or memtest 86) and are getting memory errors???????

3. Then your Thinking of RMAing the GPU.

4. The only problem you have Identifed is a BSOD (UNKNOWN REASON), and then a inability to fix a software/driver problem.

NOT an advicate of the "shotgun" approach to fixing a problem.

.. By almeans RMA the PSU, for Noise as it should not have this noise (Not a Normal noise - Most PSU noises are a high freq whin). But before RMAing other components, First determine if they are defective.

Added, as other post poped up:
1) Yes your BIOS voltages are Great, BUT you need to verify under load as Bios readings are slightly above Idle. Second, Being with in voltage specs does not always mean the PSU is NOT a problem. The diff between idle/load will indicate ability to regulate, that is why I also place a 0.3 V shift as part of my criteria. There are two other factors that the Average user can NOPT check - A) Ripple and B) Spikes.

For checking voltages (primarily the +12 V) I normally run prime 95, In blend Mode, as this will also indicate CPU/RAM problems or CPU Temperature problems. Then I run Furmark to also check stability of +12 V and verify the GPU.
 

Z1NONLY

Distinguished
RMA the PSU.

Run memtest to check your RAM..Then RMA if it's bad.

Try system restore to a time before you think you "improperly deleted" the driver.

Let us know the results, and we can proceed from there, if it's still having issues.
 
G

Guest

Guest
Thank you guys for your input. I've already started the RMA process and i'll be sending the psu and the ram in as soon as I could. I am RMAing the ram because I ran memtest and got "Unexpected Interrupt - Halting" message from memtest so I thought I'm better off RMAing the ram too. I forgot to mention that when I heard the noise, I didn't know that it was coming out of the PSU right away. In fact, it took me a long time to eliminate all the possibilities of it being caused by a fan or GPU coil noise etc. I didn't decide to take apart the computer piece by piece to figure out where the noise is coming from until recently and that is why I kept using the computer until that point. The noise is by no means loud enough to know it's source which certainly made my life a little more difficult. @RetiredChief I've never really been a big fan of torture tests and that is why I've been reluctant to do that. Is there a way to monitor your 12v rail readings during a torture test tho? if so then that would be great. One more question though, can I still be covered under warranty if my gpu or motherboard were damaged due to the faulting psu or does that still count as damage that companies will refuse to repair like physical damage?
 
G

Guest

Guest
I just took out the PSU and looked at the PSU fan and I noticed white liquid like matter right under it. I know that this is obviously a bad sign, but does anyone know what that could possibly be? it looks like it was dripping then solidified.
 

Kari

Splendid

it is quite common for the manufacturers to use some sort of hot glue in the assembly process.. (to prevent coils from vibrating etc)
take a look at the pics on this psu review
http://www.techpowerup.com/reviews/Seasonic/X-750/4.html
that is a high quality unit and there is that white stuff all over the place :p
 
G

Guest

Guest
Oh, that's interesting. Please excuse my ignorance, i'm not really well versed when it comes to hardware at all which you've probably already noticed. I would really like if someone could answer my previous question though so I can at least know whether I have a backup plan during the worst case scenario (if the psu did damage some of my parts).
 

The manufacturer will most likely say that its damage caused by the user, in this case you, and will not refund or replace. But still RMA them to see what happened and learn from the experience. Not trying to be condescending in anyway just saying what i would do if i were in your position. Best case scenario you will get back a new PSU and RAM and everything will function as they should. Worst case scenario your PSU fried you GPU and other various parts of the computer and manufacturer will not replace. Either way building is a learning process whether or not you have a good or bad experience.

-Best of luck
Helper800

 
Can not answer your above question, But in regaurds to Voltage and load.
1) To check Voltage under load, I use HWMonitor which displays A) current value, B) Min Value and C) max value. (min and Max can be reset).
2) Generally only need to run "Stress test" (prime 95 or Furmark) for about 5 Min. 5 Minutes does not contsitute much STRESS.

Some times these software programs do not work. Example if your +12 V reads 16 V or 8V, but the system boots then these values are bogus. In these cases I use a digital Multimeter (DVM). They are cheap and very easy to use, and come in very handy around the home and the car - Not just for computers.
Cost - about $18 at Walmart or auto store.
.. Just open side of computer.
. . Plug black Meter lead into on of the two middle pins of a molex (power) connector Also Has Black wire.
.. Plug Red meter lead into End of molex connector with the orange wire - This is your +12 V (Molex end with red wire is your +5 V).
.. Power on, red voltage when OS is done loading = Idle. Then Run prime 95 or Furmark for 5 Min to get loaded value.

How is is it to use, Take your pick from the link below.
http://www.google.com/#hl=en&sugexp=les%3B&gs_rn=1&gs_ri=hp&cp=11&gs_id=17&xhr=t&q=Using+a+digital+multimeter&es_nrs=true&pf=p&tbo=d&output=search&sclient=psy-ab&oq=Using+a+dig&gs_l=&pbx=1&bav=on.2,or.r_gc.r_pw.r_qf.&bvm=bv.41248874,d.dmQ&fp=35aa3d940bc48dd8&biw=1229&bih=669