Random BSOD Mystery

[gphantom]

It's a re-post though nobody answered in the other section. I thought to bump it here.

Please help out, im sure thousands of people have this problem in common.
-
Hello everyone.
As you can see I'm actually stuck in solving the mystery until i found a thread on your forum that has the title -
-BSOD Random Reboots, Mystery Solved
https://www.techspot.com/vb/topic108859.html

I read the whole post and i have EXACTLY the same problem

But i dont understand one thing. How did the guy fix this issue?

I have ONE stick of 2gb RAM.

What should i do to it in order to fix my problem, Like its fixed in the above thread.

If you can give me his EMAIL address or if he's present on this forum still. I would like him to help me out. I'll be grateful...
x(

EDIT: Let me further mention that. A month ago my system crashed rarely, Graphic freezes etc. That time i had 2 sticks of 2gb corsair on a 32 bit system. For 32 bit OS i thought to give away one of the stick to my friend. Happens to be, the BSOD came down like cats and dogs ROFL. Until i followed this thread where the guy says he switched the stick into another slot. THAT did help TEMPORARILY, i mean it was AMAZING for a week i didnt get a single crash..until it crashed again. though not like it was before.
still i want to make sure what really causes this. It really has something to do with the swapping of RAM slots or something.

Please solve this once and for all. I'll forward the link all over the internet.

 

[dividebyzero]

I don't see how swapping RAM amongst the DIMM's can cause this.
First things first. You haven't mentioned in either thread your system specs- they would be rather handy.
Have you tested the RAM for stability (Memtest86) ?
Have you checked your BIOS to see if your timings and sub-timings are correct (either for JEDEC or XMP) and DRAM voltage ?
While you're in the BIOS check the MCH/Northbridge voltage setting. If it's set to "Auto" then may be able to access it's true reading through the BIOS's hardware monitor.

Breaking up a Corsair RAM module kit is probably not the wisest procedure as you'll note that in addition the the model number (say TWIN2X4096-6400C5 for arguments sake) there is a version number-this denotes what manufacturers memory chips are under the heatspreader. Most of Corsair's modules have between 3 and 8 different versions, and many are incompatible when run together. Finding v1.x and v2.x (often Micron/Qimonda and Elpida) can prove very difficult.

 

[gphantom]

SPECIFICATIONS:
CPU: c2d e6850 3.0GHz @ 3.6GHz [BSOD even at stock frequencies]
Motherboard: Asus Strike Extreme nForce 680i [all driver updated after BSOD, still no use]
GPU: gtx260 192 @ latest nvidia drivers [tried other stable drivers too]
RAM: Corsair XMS-2 DDR2 800 Mhz 5-5-5-18-2T (1.8v) [tried every stable timings/voltages, no use]
PSU: A-45GA 450 watts
No extra peripherals.

Memtest passed every test..never had errors..3 days back when i switched the RAM memtest ran overnight with 8 passes 0 errors. + crashes happened on every stock settings.

(either for JEDEC or XMP) and DRAM voltage

I dont know about those. I never tweaked any of these things/never noticed those variables at all. Would be glad to test a tweak. I believe these are gonna be my last hopes for fixing it.

While you're in the BIOS check the MCH/Northbridge voltage setting. If it's set to "Auto" then may be able to access it's true reading through the BIOS's hardware monitor.

I'll do that in a minute, will edit the post.

And i just have 1 stick of 2gb RAM.

Thanks for the reply.

Here's a picture of the BIOS voltages. If you cant read them clearly, i'll type it for you.
http://dl.dropbox.com/u/1518581/img076.jpg

 

[gphantom]

Fail,
I was wrong about swapping RAM slot fixing the issue. It started crashing again right on my face. It was all a coincidence so far then..

But i think i'm getting the idea for why all this is happening. As i mentioned in my first post. It started in Winters. when temperatures are LOW. When i turn on the heater its fine then.

+

I get alot of vDroop. e.g

Voltage in bios – vcore – 1.45
while
Voltage in cpuZ – vcore – 1.344~1.360

If you can please look at the screen-shot i've linked in the above post and confirm that my PSU is weak. One more query, Do PSUs grow weaker with time? if it does then thats my best bet.

P.S. I'll try getting another RAM from a friend just to check if the crashes occur on that.

 

[dividebyzero]

First off -Theres not a 450watt PSU on earth that will provide enough power to keep your system stable- the good news is the fix is at hand.
The 680i SLI chipset is one of the most power hungry ever made -closely followed by the 780 and 790. The Striker is somewhat worse than it's competitors ( I owned a EVGA 680i SLI, 790i Ultra and, very briefly....an Asus P5N-T Deluxe (feel free to laugh).
Your E6850 is a 65watt CPU -running it at 1.45 Vcore/3.6G turns it into a 105 watt part.
nVidia's specs list the GTX260 (216SP) has having a 25 watt idle and 182 watt max power draw (if it's a 65nm part) or 171 watts (55nm part)- nVidia recommend a 500 watt PSU with 38 (or more) amps on the 12 volt rail/s.
A quick calculation says you will be drawing a little over 300 watts on the 12 volt rails under load (assuming you have a 1-2 harddrives, a DVD burner a CPU fan and a few case fans)- my guess is thats iffy in the extreme with your PSU -especially if it's a few years old as PSU's lose a percentage of their rated power over time.
I think upgrading the PSU to a good brand in the 550-650 watt range (@ 40 amps) would most likely solve most (if not all ) of the problem.
It may not solve it entirely- Asus's nVidia chipset mobo's are quite prone to instability and the P5N-T Deluxe (780), Striker Extreme, Striker II Extreme and NSE (790i Ultra and SLI), Striker II Formula (780i SLI) and P5N32-E SLI (680) in that order are the trickiest to maintain (The P5N-T Deluxe ceased production in record time). There's a couple of reasons why but I think you'd be best served by tackling the obvious-and I think you're pretty much there already judging by your post- and thats buying another PSU.
Judging by your screenshot, all three voltages are a bit on the low side anyway (+3.3, +5 and +12) -nothing radical as the ATX spec is +/- 5% on each. You just get enough amperage into the system for stability- although excessive ripple, even on a seemingly well specced PSU will cause a system to crash. You can test the ripple with OCCT's PSU test. OCCT here: http://www.ocbase.com/perestroika_en/index.php?Download
Memtest86 tests the RAM independently of the the rest of the system so thats obviously ok.
Running a torture test would most likely pass too (OCCT, IBT, Orthos, Prime95 etc), but I think the moment you ran the OCCT GPU test you'd most likely crash the system, because that where the big power draw is.

 

[gphantom]

I'm in safe-mode atm. I cant go in Normal mode for it uses the GPU and thats what causes the pc to crash.

I'm grateful for the info you've put here. I just noticed prime failing with an error saying "hardware failure". I'm sure OCCT would crash the system for sure.

I just remembered when my system crashed at my friends place i had another ram. and now i have a different ram and still it crashed.
+
the psu was a second-hand..argg...

I tried a link for some power-calculator and it gave me 404 watts on a 12 volt rail..this is disappointing..nonetheless a relief..I think its time i get another psu..

Once again i'm thankful for your help.

Just one last request. Do u think this psu:
http://www.coolermaster.com/product.php?category_id=30&product_id=5906
would be suffice for the current system i'm running?

 

[dividebyzero]

I'd say the CoolerMaster PSU is probably borderline at best. CoolerMaster PSU's don't have great efficiency-and I can't say I'm a fan. They are generally cheap-unfortunately, there's a reason for that. It's definitely better than what you have- but I personally would go for a better build since a bad PSU can take out a good chunk of your system if it goes.
Depending where you are and what etail/retail you have access to (and budget of course), any of these would be a good choice (Amps, Watts and connectivity)
Corsair VX550 http://www.newegg.com/Product/Product.aspx?Item=N82E16817139004
Tt 650w http://www.newegg.com/Product/Product.aspx?Item=N82E16817153040
Antec TruePower 550 http://www.newegg.com/Product/Product.aspx?Item=N82E16817371020
OCZ 600w http://www.newegg.com/Product/Product.aspx?Item=N82E16817341010
Silverstone Strider 600w http://www.newegg.com/Product/Product.aspx?Item=N82E16817163109
Seasonic 550w http://www.newegg.com/Product/Product.aspx?Item=N82E16817151092
Antec BP550+ http://www.newegg.com/Product/Product.aspx?Item=N82E16817371016
OCZ ModXstream 600w http://www.newegg.com/Product/Product.aspx?Item=N82E16817341017
OCZ Z-Series 550 http://www.newegg.com/Product/Product.aspx?Item=N82E16817341037
OCZ Fatality 550 http://www.newegg.com/Product/Product.aspx?Item=N82E16817341022
Raidmax RX-600AF http://www.newegg.com/Product/Product.aspx?Item=N82E16817152041
Tt TR2 600 http://www.newegg.com/Product/Product.aspx?Item=N82E16817153114
Sparkle 600w http://www.newegg.com/Product/Product.aspx?Item=N82E16817103038
Seasonic 520W http://www.newegg.com/Product/Product.aspx?Item=N82E16817151094
Ideally I would look for 80+% efficiency and 40 amps or more on the 12v rail/s
The CoolerMaster : http://www.newegg.com/Product/Product.aspx?Item=N82E16817171039
is only a 70% efficiency unit 600w x 70% = 420w actual) and is a little light in the current dept. TBH - especially for 600watts.
I'd say the power calculator is probably in the right ballpark, figure wise

 

[gphantom]

aH. I think the best i can get among those is just the Corsair one. and all others are coolermaster here.
But hey, If you're suggesting these, then maybe I'll just wait and get it from abroad.
Because I just got the current PSU to work after getting every over voltage to its stock values and i cant describe it but its a wet dream..my next upgrade would surely be a PSU. God willingly ^^

Btw i'm from Pakistan.

You get guns for free..and gadgets for gold here.. :dead:

Thank you so much for all the help.

 

[gphantom]

Hey, I think I'm having second thoughts.

After I got the overclock to stock voltages and all. The IRQL and all those bunch of random BSOD are gone, Completely.

Plus, I ran the OCCT Power-supply torture test for 2 hours and it passed without a single BSOD.

Problem is, I'm getting a very NEW BSOD this time, and its the same every time, Not random at all.
That says

*** Hardware Malfunction

Call your hardware vendor for support

*** The system has halted ***

Now I don't know if i really should spend my budget over the PSU, Since it passed every test even the Combo test.

I Just installed the Bluescreen viewer in hope to capture some faulty drivers. and i dont know why but i don't have any dumpfiles in the minidump folder. Will the bluescreen viewer capture the future ones?

Thanks

 

[gphantom]

I think the problem persists, At least not all of it but,

and what i know about this new BSOD is that,

1 - its the same every time

*** Hardware Malfunction

Call your hardware vendor for support

*** The system has halted ***

2 - No logs/dump-files created. Event viewer shows nothing

3 - Most importantly, It will happen in the first 2, 3 boots of the day Edit: First 5 minutes, It might crash during booting or on desktop.

Like i mentioned before. Turning the pc on after its been off for quite some time, the more the time its off, the more is the chance to crash on the first boot.
But after a few boots it feels as if the pc warms up and afterwards it runs perfect for the rest of the day.

 

[Route44]

Overclocking will cause your issue.

Question: Is the RAM that you have installed on the recommended vendor list from Asus for your particular motherboard? Corsair makes great RAM but what I am specifically asking is if your particular Corsair model on that list?

 

[gphantom]

Thanks for the reply,

Current RAM:

DDR2 800 CM2X2048-6400C5

On vendor list I found only two models of 800 Frequency:

DDR2 800 1024MB CORSAIR Heat-Sink Package N/A N/A DS CM2X1024-6400C4 V V V
DDR2 800 1024MB CORSAIR Heat-Sink Package N/A N/A DS CM2X1024-6400C3 V V V

 

[dividebyzero]

Listed as compatible with Corsair's memory configurator but not on Asus's QVL. Even if you had the C4 kit (which is just a faster binned version of what you have) Asus don't list version numbers so compatibility isn't a given, and judging by a quick Google search....
Your problem is not an isolated case
Corsair forum
Driverguide forums
Asus support forum

Recent BIOS updates address specific issues with Corsair RAM -here

I don't know if these problems still exist or not. Personally I'd ditch the Corsair and buy a kit of Crucial Ballistix - they are on the Asus QVL and have worked with virtually every DDR2 board around.

 

[gphantom]

That is Great help brother,

But I don't know why I don't feel like its the RAM anymore. I read on another forum that the Crashing of PC on the first BOOTs, after leaving the PC OFF for long is actually BAD CAPS.

BAD CAPS either in the MOTHERBOARD or in the PSU. And I really am about to bet on the PSU since the Motherboard isnt that old and no apparent signs on BADCAPS + the psu is a second-hand and it weak and too old.

When I remember, the first time it crashed was also due to COLD temperatures in SUMMERS, at my friends place where he had some extreme AC ON. and that BADCAPS are bad when COLD. + I had another RAM at that time.

I'll wait for one more week, observe and finalize my hypothesis about the PSU. Today the PC crashed Just once on its FIRST BOOT. and no BSOD for the rest of the day.

So I guess I just want you to approve this and then I should go and take a new PSU rather than a new RAM.

Also, I'd like you to look at this graph from my OCCT torture test. Look at the voltage drop, Is that normal?

 

[dividebyzero]

That drop in Vcore is pretty normal. As soon as the CPU goes under full load the Vcore drops a little -this is Vdroop explained here (beware uber-tech http://www.thetechrepository.com/showthread.php?t=126)
Sounds like the board definitely has a cold boot problem. Your BIOS is the latest available?

 

[gphantom]

Yes its the latest 1902 something. Actually i cant look into the PSU just like that since i have a clean setup of the cables inside the casing and Maybe i open it in a day or two just to look for badcaps.

 

[dividebyzero]

Very nice clean looking build. Impressive cable management.

 

[gphantom]

Solved

I'm here again just to tell you guys that I'm grateful for helping me out with the problem, and to tell you my story how it ended up for me.

A week ago my PSU couldn't take the load anymore. It would crash on little loads and hardly reach the desktop. I removed the PSU and opened it to look for bad caps and what I found was some GOO on the RING CAPS and but NO signs of bad caps on the CYLINDER CAPS. no bulging caps nothing pouring out of its bottom.

But anyways. I bought a corsair TX 750w PSU just today and I'm glad it finally solved all the above issues. I overclocked to my desired freq/timings again and its all well.

I'm attaching a picture of the Inside of the old PSU. Just to confirm if its the Electrolyte from a bad ring-capacitors.


This psu was fitted upside down in the casing. what you are looking at was facing the ground.