[Troubleshooting] $3,000 Rig, First-time Builder, Serious Problems

BuildBuildBuild

Honorable
Mar 24, 2012
10
0
10,510
Specs:

Processor: Intel Core i7-3820
Motherboard: Asus P9X79 Deluxe
Case: Cooler Master HAF-X
Fan: Noctua NH-D14 SE2011
RAM: G.Skill Ripjaws Z DDR3 1600 16GB 4x4GB CL7 (Model #: F3-12800CL7Q-16GBZM)
HD: Samsung 830 SSD 512 GB (Model #: MZ-7PC512D/EU)
PSU: Seasonic Platinum 860W (Model #: SS-860XP Active PFC F3)
Graphics: PowerColor HD 7950 PCS+ 3 GB (Model #: AX7950 3GBD5-2DHPP)

Note: All the latest Drivers and Windows Updates have been installed.

Pic of Build: http://i.imgur.com/90l4L.jpg
Close-up: http://i.imgur.com/GIuSo.jpg

Notes:
No major issues during build.
All latest Drivers and Windows Updates have been installed

---

Turning On First Time:
First time I turned it on, got a green flashing screen. Turned it off. Waited a few minutes, turned it back on, worked OK (never had the green screen problem since). Got CPU Fan errors. Read on forum that’s normal for Noctua NH-D14 fan due to efficient RPM, so I lowered CPU Fan Alarm threshold and never got the CPU fan error again.

---

Main Issues:

When it wakes up from from a short sleep, the Q-Code stays at “30” (which refers to the sleep cycle). A couple of times I saw it stay at “30” and then it changed to “80” (DXE Initialization) while still working, but I never saw it go back to “AA” after waking up. Shouldn’t the Q-Code change to normal “AA” status once the computer wakes up and has been used for over 30 minutes? In case it helps, when I put it to sleeps, it goes from “AA”, to “03” (System Agent initialization before microcode loading), and then goes off.

At first, computer would wake up immediately from a short (few hours or less) sleep. But it did not seem to respond to an overnight sleep. After leaving it asleep overnight, pressing the keyboard or moving the mouse did nothing. I had to shut it off.

Now, after installing the last 36 Windows updates, it only sleeps for about 30 seconds and then wakes up by itself but the screen is still blank. If I move the mouse, Windows then appears normally on the screen.

Computer has frozen at least four times so far. Once when editing the Windows Hosts file in Notepad, second time while opening a Youtube Video, third time when running 3dMark11, fourth time while opening a Youtube video again. I checked the Q-Code of the MB when the second freeze and fourth freeze happened and it showed a Q-Code of “30” on both occasions, so it might have to do with the Sleep. When it freezes, the screen stays on whatever it was on, but nothing moves or reacts or changes (neither on the screen nor on the computer itself). It simply stops responding and reacting. When it froze on 3dMark11, the screen was just dark and stayed that way (I ran 3dMark 11 4 times, it only froze on one of those four times).

---

Diagnostic Screenshots:
Fans & Voltage: http://i.imgur.com/ct8cc.png
RAM: http://i.imgur.com/pQbU0.png

---

Underperforming Benchmarks:

1) SSD

Samsung 830 512GB with latest firmware.
Partitioned and properly aligned.
S.M.A.R.T. Screenshot: http://i.imgur.com/ULKf9.png
Crystal Disk Marks: http://i.imgur.com/WPI0O.png

Based on average Crystal Mark Scores from 4 different sources online (see below), my SSD is underperforming an average of -36%:
http://www.legitreviews.com/article/1795/5/
http://www.pureoverclock.com/review.php?id=1455&page=8
http://www.reghardware.com/2012/01/06/review_storage_samsung_830_ssd/page2.html
http://www.reportlabs.com/testbed/version1/hdv1/detailssd.php?recordID=128

After doing the last 36 updates on Windows, I ran it again, and it’s still underperforming: http://i.imgur.com/vngtJ.png

2) GPU

3DMark 11 Score: P7187
Full Details: http://i.imgur.com/QKewv.png

---

Based on all of the above, it is clear to me that something is very wrong with my build, which is quite disheartening after all the time and money I invested in it.

Please help me figure out all these issues as I’m way in over my head and have no idea how to diagnose or resolve them.
Building a PC was far more complicated than I anticipated (especially due to the poor instructions in the manuals regarding so many cables), but I thought it was still worth it when I finally finished the build and saw what I had created with my own two hands. Now, with all these problems that my old Dell never gave me (it pains me to say that), I’m starting to have regrets and second-thoughts, especially since I saved up a lot of money to buy top-of-the-line components to avoid precisely these kinds of issues. To be honest, after this experience, I probably won’t be building a PC ever again, but I want to at least salvage this one so that it lasts me at least 3 years.

I’m willing to test, diagnose, and try anything, so I’m all ears.

Please help. Thank you.
 

pcfxr4u

Distinguished
Mar 7, 2011
7
0
18,510
First, I know this is a lot of work, but take it aprt and build it on a tabletop first.
Start with the mainboard, memory, video card, and turn it on check BIOS...if it's okay,
continue to add dvd, and hard drive, check BIOS once more...now load OS. Which OS have you selected?
PCFXR4U Stafford, VA
 

Darkendnox

Distinguished
May 17, 2009
22
0
18,520
Sounds like your hard drive, make sure everything is plugged in correctly. Go into windows power options in control panel, go to advanced tab and turn hdd sleep to never. Might aswell turn all sleep settings off... Also turn off Yang suspend setting.
 

BuildBuildBuild

Honorable
Mar 24, 2012
10
0
10,510
Thanks for your help. Why do you recommend taking everything apart and putting it back together and checking the BIOS?

By the way, is there any diagnostics I can run beforehand to maybe isolate this issue so that instead of taking everything apart, I can focus on something more specific?

Thanks.
 

BuildBuildBuild

Honorable
Mar 24, 2012
10
0
10,510
Thanks for your assistance.

What makes you think it's the HD? At $850, I really hope it's not it.

Also, I don't wish to lose the Sleep function as I normally use it overnight to continue working where I left off the next day. Why do you recommend turning it all off?
 

devils_advocate

Distinguished
Nov 13, 2011
53
0
18,640
In general when you building a PC from scratch first time, its best you build the system with basic component out side (Not in side tower). Personally i just use the box that mother board comes in, and put it together on it with basic component and see if i can boot in Bios. Save alot of time compare to building system in the case and then having to debug because there are problems. And then having to take it apart...

As for the advice to take it apart and put it back together with basic parts out side of the case. Its usually how you go about debugging the hardware. Once you certain hardware is not the problem. Then you move on to software Like bios, O.S and drivers.

Well that's how i pretty much go about fixing my new builds. Never had any problems, i couldn't narrow down and fix it.
 

BuildBuildBuild

Honorable
Mar 24, 2012
10
0
10,510
Thanks for the tips, but are PC components that unreliable that we are forced to essentially build our PC's twice just to make sure everything is OK? I'm also not sure that I would have caught any problems during an "outside" build because at first everything seemed to be running fine...then the freezes started, strange sleep/wake behavior, and underperforming SSD & GPU.

My questions is, even with an "outside" build at this point, how can I test each component to figure out what might be the cause?

Right now I'm running Memtest which I will leave overnight. Tomorrow during the day I'll run Prime95. These systems are so complicated that it seems really difficult to pin-point any kind of issue, which makes resolving these problems extra difficult.

Any additional insight and suggestions would be appreciated. Thanks.
 

Ohmybad

Distinguished
Nov 16, 2011
327
0
18,810
seems like a SSD issue, idk if you have the most current firmware for that 830, I've read some places (other forums) with problems with SSDs and sleep mode. the difference from sleep and a fresh boot is what? 15-25seconds?
 

Darkendnox

Distinguished
May 17, 2009
22
0
18,520




Sometimes SSD's don't like the sleep function, Why I dunno... but it might be something you have to live without.


Just leave your computer on and never turn it off, Just turn monitors off and call it good.
 
G

Guest

Guest
Did you install the ssd in AHCI mode?

I would run memtest as well. http://www.memtest.org/
 

BVKnight

Honorable
Mar 18, 2012
79
0
10,660
Did you update the BIOS on your motherboard? Just reading the reviews on newegg, it seems like a lot of people have trouble with that one.
 

fil1p

Distinguished
Nov 29, 2010
944
0
19,360
I think it could be your motherboard, as you have had so many issues in so many components, that all lead back to the motherboard. Also in the past some mobos have had sleep issues. Try updating the bios ad BVKnight said, also make run some more benchmarks, on individual components to see if everything else is running fine now. So how did the benchmarks go?

Hope this helps!
 

BuildBuildBuild

Honorable
Mar 24, 2012
10
0
10,510
Thanks for the additional help, folks.

Here's an update plus answers to the questions above:

I've been running Memtest for over 11 hours. It's done 5 passes with no errors. Will wait for the 6th pass to complete and then end it. If there are no errors after the 6th pass, does that essentially mean the RAM is NOT the problem?

Yesterday I ran the latest Prime95 torture test (in all the default settings) for about 45 minutes. Although the CPU was running at 100%, I was using the PC during that time, and it did not crash, and I did not see the CPU temp go above 45 C. I will do a longer test today after the Memtest. How many hours of the Prime95 is recommended to rule out the CPU?

I have NOT updated the BIOS to the latest firmware. I'll do that today to see if it at least resolves the sleep issue as I've read elsewhere that it might actually fix that.

The SSD came with the latest firmware already and is installed in the Intel X79 Controller 6 Gb/s port, and AHCI was set by default when I installed Win 7 Pro. Any other ideas why it might be running slow? I've read elsewhere that doing a FULL format could resolve the speed issue, is that recommended? Or should I just have it replaced since it's brand new and cost so much?

What about the low 3dMark11 score on the GPU? Few have mentioned it.

Lastly, what would cause the computer to just freeze in place (with no BSOD) and not respond at all during the crashes I've experienced? Maybe that behavior helps narrow down one of the problems, so I just wanted to point it out. Also, is it possible to get a dump from one of those crashes and analyze it? If so, where and how would I do that?

Thanks again to everyone for your help, and please keep your thoughts, suggestions, and insights coming. I'll continue to update this thread until my PC is working 100% correctly.
 
G

Guest

Guest
Ok well first memtest is not end all memory test, but a good place to start from. If ran prime95 with all cores active that is good. I am not exact on temps spec on that cpu but you seem to be in line with what to expect. One thing that would help is to try note what that computer is doing when it carshes. Then report back on if their is any thing that makes crash more often then not. Other not prime 95 with test the cpu and ram but it is not end all ram test due to will not use all the ram. One the other test is to try intel burn test. I have old q6600 overclocked to 3.6 ghz it will pass prime 95 but will not always pass intel burn test(some times it will).Other then that system is very stable and has been so for many years. Getting back to your system if is seems to happen very randomly. You will have to start tiring to remove components and see which one helps. In general I will remove everything but very basic. Have one stick ram and cpu if onboard video just that. Then use flash drive or cd image of known good bootable live image of some sort. Then add components back until the machine has a issue again. It very improtant to add one at time. Note that ram can be huge pain toubleshoot because differnet configurtion might cure or hurt the issue. Because with ram issue and can be that cpu, ram, or motherboard(slot order). Or all them. Good luck
 

pcfxr4u

Distinguished
Mar 7, 2011
7
0
18,510
I don't see anyone adresssing this for you...if you go into the sleep mode, you won't be acomplishing anything. With an SSD, your boot time from a cold start will be a minute to a minute and a half...disable sleep, hibernate etc...download an auto shutdown gadget or use the Windows 7 built-in shutdown command when you want to keep running until a job is complete. And, your apps can be set to auto run at boot if you really, really, really need to have them up at boot time.

Everyone here is giving you good advice, and yes, PC componentsare beta tested usually buy consumers, just like Windows is. That is a life lesson. Before you buy anything, wait for revision 1.1 or 2.0 then google what you want to buy this way...as an example, EVGA SR-2 issues...or whatever your interest is...
 

BuildBuildBuild

Honorable
Mar 24, 2012
10
0
10,510
UPDATE (Sunday, March 25 @ 3pm GMT):

Ran Memtest for over 12 hours and after 6 full passes, there were no errors. Is it safe to say the RAM is not the problem and is OK?

Updated BIOS to latest firmware.

Since updating BIOS firmware:

* SSD: benchmarks are all normal EXCEPT 4K Writes which have improved but are still underperforming by about 33%. This is definitely progress, but any ideas why that specific benchmark could be underperforming while everything else is now normal?

* Sleep issue: Put computer to Sleep, and it stayed asleep (remember, before it would wake up automatically after about 30 seconds). Woke it up after 15 minutes, and several minutes later, noticed Q-Code was at “AA” (remember, it used to remain at "30", which was the Sleep cycle). This is the first time I’ve witnessed the Motherboard report a normal Q-Code of “AA” after waking up from a Sleep. Again, this is definitely progress.

* Have been running Prime95 Torture Test for over 2 hours now after waking the computer up, while looping TWO Youtube clips on EndlessYoutube.com (remember, my PC had frozen twice while on Youtube previously), with no errors so far. During this time, I decided to push my computer to the limit, and also ran 3dMarks11 benchmark. System did not crash, and got a 3DMarks score of P4385. How long should I run Prime95 for?

* CPU & Fan Info after 2+ hours of Prime95: http://i.imgur.com/UqSFG.png (I have no idea how to read this, is it all looking normal?)
Graphics Card: Can someone please confirm if a 3DMarks score of around 7200 is indeed low for my specs? I’m not yet clear on that, but have not benchmarked the card since after the BIOS update (other than during Prime95 as above), so I'll have to wait until I complete the Prime95 testing before benchmarking it again.

So it seems the BIOS firmware update has helped in at least a few areas, so I'm glad to see some progress. Thanks for everyone's help so far, I know that last thing you guys want to be doing on the weekend is helping someone with their PC troubles. Hopefully I can get this all resolved and running smoothly by tomorrow, or at least pin-point the cause and get it exchanged. Thanks again and I'll continue to answer all the posts and keep everyone updated until this is 100% resolved.
 

BuildBuildBuild

Honorable
Mar 24, 2012
10
0
10,510
UPDATE (Monday, March 26 @ 12:40pm GMT):

* Ran the Prime95 Torture Test for over 10 hours yesterday after waking the computer up, while looping TWO Youtube clips on EndlessYoutube.com the entire time, and while using the computer normally throughout the day, and had NO errors. These are the CPU stats towards the end of the Torture Test: http://i.imgur.com/VQm4i.png

I think it's safe to say the CPU is fine.

* The above, coupled with the earlier Memtest Results (http://i.imgur.com/M5r9y.jpg) seem to show that both the CPU and RAM are fine.

* I've enabled XMP in the BIOS and my RAM is now running at its intended clock speed of 1600 Mhz.
Since enabling XMP, my SSD and Graphics benchmarks have improved (no idea if related, or just coincidence, but happy either way). Furthermore, I also did a “Performance Optimization” using the Samsung 830 Magician Software and now my SSD speeds are exactly where they should be across-the-board

* New SSD Scores: http://i.imgur.com/j05cu.png

* New Graphics Scores: http://i.imgur.com/FfwUD.png (P7419, nice improvement, have been advised elsewhere that the 3dMark scores are compared against other cards, including Overclocked Cards and GPU’s, which is why 3dMark is reporting my results as "low" since everything of mine is at stock speed.
I left the computer asleep overnight (remember, before the BIOS update, it wouldn’t wakeup or respond after an overnight sleep) and it woke up without a hitch.

Computer has not frozen at all since I updated the BIOS early yesterday (not even during the simultaneous Prime95 Torture Test, 2 Endless Youtube clips, and the 3dMark benchmarking).

All in all, it looks like everything is normal now and working as it should, which is a HUGE relief for me. What my inexperienced mind first thought were multiple hardware failures was simply an outdated BIOS firmware and lack of optimization. I didn’t even have to open my case to get everything working like it should.

Moral of the Story: ALWAYS update your BIOS to the latest firmware when you build a new PC. I learned the hard way, but you don’t have to! :)

I can finally begin to enjoy my creation, and that is a wonderful feeling. Thanks again for everyone's help, you’ve all been great and made this a lot less painful than it otherwise would have been. I'm very glad this type of support for newbies exists otherwise I would have been drowning alone in a sea of silicon.
If anything else happens, I’ll report back, but I’m happy to say this will not be the last PC I build.

END UPDATE.
 

fil1p

Distinguished
Nov 29, 2010
944
0
19,360
Im glad that it finally works for you, also as for the results that you couldn't read:

From the results judging by your core temps everything is well within the range, the CPUTIN is reporting crazy max temperature, but I wouldn't worry about that, what counts is the core temps, and they are fine. Your voltages seem fine, so your good, as long as you don't have any additional problems. If you want help reading the specifics of the picture you posted (http://i.imgur.com/UqSFG.png) tell me, but above i gave you the summary.
 

BuildBuildBuild

Honorable
Mar 24, 2012
10
0
10,510
Yes, please, I would really appreciate help reading the specifics of the screenshot. That would be great, thank you.

Also, are those crazy Max figures indicative of faulty sensors on the Motherboard? Because if anything is faulty, I have no problem replacing it. I spent so much money on this build that I want to make sure everything is working 100% as it should.

Thanks! :)
 

fil1p

Distinguished
Nov 29, 2010
944
0
19,360
I wouldn't worry about the crazy reading, it could be that Hardware Monitor misread the sensor. However from the latest screenshot I would say that it is ok, just ignore the two reading AUXTIN, and TMPIN3, they are most probably not crucial. What you should look at is CPUTIN, or in other words CPU temperature, don't worry now it is fine. To read the temperatures and voltages on HW Monitor, you look at the three columns on top, one says "Value", the next one says "min", and then "max". Value is the current temperature, voltage, or fanspeed, it displays the data that is current. Then you have min, it displays the lowest value (temperature, voltage, etc) that a specific component has reached since HW Monitor has been opened. Then there is Max, and that displays the highest temperature that your component has reached since the openin of HW Monitor. So lets say you turn on your PC, and run HW Monitor, the CPU (in your case CPUTIN) will start out at 20C, and then it will warm up, and reach a temperature of lets say 24C. HW monitor will now display 24C as the "Value" or in other words current temperature, and 20C will be your min, then your max will be 24C, as it was the highest temperature that has occurred throughout the time that HW monitor was on.

The main things you want to look at is your CPU temperature, although it is recommended to look at the individual core temperatures. Core temperatures provide you with more detail (as normal CPU temp comes from the sensor mounted under the socket, while core temps are from the CPU and are more accurate), and it is normal that there is a difference of a few degrees on each core. Keep your temps there more or less under 80C, but that rule of thumb varies from CPU model, and usually you hit that when overclocking, but I looked at your load temps, and the hottest core at 100% usage was at 58C, so you are more than fine. I do not have much expierence with your processor, but I think if you keep your temperatures under 70-75c and you will be fine, and your processor doesn't come close in reaching them, so your good.

Finally you have your voltages, now you do not necessarily have to worry about those, but lets go over them anyway.

Then you have your SSD temperature, but you don't have to worry about it.

Then you have your GPU temperature, and it should stay under 80C when gaming, it usually sits bellow that even while gaming.

Those are the essentials

Now lets look at some voltages, however HWMonitor gives you a general idea, although on yours it doesn't display all the rail voltages, but you have the VCORE, which is the voltage of your CPU, and it can be seen in HW Monitor, or it is better to use CPU-Z, but Hw Monitor will give you the general idea. Earlier I said that HW Monitor doesn't want to display the voltage for all the rails except the 3.3v rail. However from the "voltages" screenshot in your first post you can see that your 12v rail (the rail where all the main components like the CPU, and GPU take power from). Now your is @12.096v, which is good, it is very close to 12v (none are perfectly 12.0v) and there is a tolerance to what is acceptable, but you are well within the range so you are more than fine. Your 5v rail (Hard drives, and other devices are on this one) is @5.08v which is also well within the tolerance range, and is fine. The closer the actual voltage is to rated voltage on the rail the better, but yours are very good. So from all those readouts you are fine, it just seems that in certain cases HW Monitor misreads the CPUTIN max temp, but that was under heavy load, and you would still probably use the core temps, so I would not worry about that.

Hope this helps/
 

TRENDING THREADS