Driver stopped responding / BSOD crashes

Associate
Joined
23 Oct 2004
Posts
236
Location
London
Greetings.

Having spent lots of cash on my lovely new rig from OcUK several weeks ago, I'm rather frustrated by how unstable it is. I'd really appreciate the support of this forum if you are able to help.

When running games I get crashes to the desktop and usually a BSOD.



Now, I've never heard of a STOP 0x116 before, but it appears to be a graphics problem - nvlddmkm.sys is part of the Forceware drivers I believe.

Spec:
E6850 at 3Ghz stock speed
4Gb (2x2Gb sets) Crucial Ballistix 5300
Gigabyte P35C-DS3R mobo
BFG 8800GTX OC2 768Mb (comes clocked at 626/1000 out of box)
Corsair HX620W PSU
Vista Home Premium 64bit

Drivers:
Same problem using the following drivers:
162.22 Official
163.44 Beta
163.67 Beta
163.69 Beta

Temperatures:
I have run the nVidia monitor to try and log some temps, running Team Fortress 2 beta as an example game.

Fan on automatic (59%):
Idle - 70C
Time to crash - 10 minutes
At moment of crash - 79C

Fan on manual (80%):
Idle - 68C
Time to crash - 45 minutes
At moment of crash - 76C

NB: I am not assuming that GPU core temperature is causing the crashes, although there did seem to be an increase in stability from upping the fan speed (although the noise of the fan above 70% is rather unacceptable for me).

I do however wonder if these temps are rather on the high side. Also, I can't measure graphics RAM temps which I also believe can cause these errors.

Power Supply:
Very soon after having built the rig, there was a fairly loud 'pop' from the machine when I turned it on. There was no odour or flash, but I would guess it came from the PSU. Could this be a sign of a problematic power supply? I don't have any voltage measuring equipment at home so I can't test it directly.

System Cooling:
I have an Antec P192 case with the default 2x120mm exhausts and a Tuniq tower on the processor. This keeps the CPU nice and cool.

However there is not much airflow over the GPU at the moment except that from the built on cooler.

Benchmarking/Testing:
Windows Vista Memory Test shows my RAM is fine.
3DMark06 runs fine and has not caused this error.
The more intense the graphics, the sooner the crash occurs, for example I can run World of Warcraft for many hours, Source games for less than an hour, and BioShock runs for between 5 seconds and 2 minutes before crash occurs.

What now?
I'm pretty stuck and frustrated. I don't know how to pin down the problem, whether it is software or hardware. I will gladly receive any advice or troubleshooting methods you all can offer, and I hope that the General Hardware forum is the place to start this thread given my uncertainty over the nature of the problem.
 
Last edited:
Associate
Joined
2 Jun 2003
Posts
600
Location
GA
OK i have the Gigabyte P35C-DQ6 mobo and had no end of problems with it. First off, take a deep breath, sorting problems like this out is half the fun !

I think your GPU temps are fine for that sort of card on stock cooling.

Visit gigabytes website and download the latest chipset drivers. While your there see if there is a bios update

Uninstall your gfx drivers and reboot. install the newly download chipset drivers and reboot

install gfx drivers.

Also on my gigabyte I get massive vdroop, it was hit and miss whether or not my machine would boot into the login screen or blue screen.

Just check with a proggy what vcore your cpu is recieving....i bet its a tad lower then it should be.

I'd take 1 pair of ram sticks out and give that a try, if its the same swap the pairs around.

Suprised you have 5400 mem, 6400 would've been better.
 
Soldato
Joined
17 Aug 2007
Posts
2,896
Location
London, Moscow & Samara
Greetings.

Power Supply:
After having the rig set up for about a week, there was a fairly loud 'pop' from the machine when I turned it on. There was no odour or flash, but I would guess it came from the PSU. Could this be a sign of a problematic power supply? I don't have any voltage measuring equipment at home so I can't test it directly.

I would personally look at the power supply even though no smell of burning ect. I think this may be the culprit
 
Associate
OP
Joined
23 Oct 2004
Posts
236
Location
London
I've done a full clean reinstall of Vista, chipset, sound, and graphics drivers (all latest release from manufacturers).

Still getting the same problem. About to try looking at hardware. RAM is first to be swapped around etc.
 
Associate
Joined
27 Dec 2004
Posts
1,199
Either your RAM is faulty or you need to increase the voltage (what are your sticks rated for?). It is highly likely the cause of your problems though, as even a single error isn't acceptable.
 
Associate
OP
Joined
23 Oct 2004
Posts
236
Location
London
Crucial Ballistix 2GB (2x1GB) DDR2 PC2-5300C3 667MHz Dual Channel Kit

I have 2 of these sets. Their specs are:
- Speed: DDR2-667
- Voltage: 2.2V
- Memory Timings: CL=3

I'm not sure why the RAM would be causing an error related to a graphics driver though...?

What program should I use to measure the voltages being supplied to my processor/RAM etc?
 
Associate
Joined
27 Dec 2004
Posts
1,199
Are you running those sticks at 2.2v then? Stock setting in BIOS will be 1.8v. On your Gigabyte board it probably only shows a list of increases and not actual voltage settings for DRAM (i.e. instead of 1.9v, 2.0v it'll show +0.1v, +0.2v etc). So a setting of +0.4v could solve your problem... if you've already done this though then disregard what I've just said ;)

Regarding as to why you'd be getting a graphics driver error... it's probably because the driver is low level and makes different demands on memory than other applications, which sorta explains why some of your games run fine and others do not. It could still be a driver error of course, but until you've got your RAM stable I wouldn't bother looking at that again. On the same note once your RAM is confirmed stable I would uninstall and reinstall the latest driver package, just on the offchance that some files were corrupted during your previous installs.

http://hcidesign.com/memtest/

I would recommend testing with this app too, run a couple of instances with each testing about 700 meg each and see if you get any errors. This particular program saved me a lot of time in the past, when I had RAM issues on my last PC memtest86+ couldn't find anything wrong, yet this program found my RAM was duff within literally seconds.

Hope this helps! :)
 
Associate
OP
Joined
23 Oct 2004
Posts
236
Location
London
I have the following results in SpeedFan, could anyone interpret them? Is Vcore2 my RAM voltage?

Vcore1: 1.12V
Vcore2: 1.84V
+3.3V: 3.34V
+5V: 4.92V
+12V: 0.51V
-12V: -16.97V

-5V: -6.46V
+5V: 5.03V
Vbat: 3.26V

Um, the 12V ones look very odd (in that they are not ~12V). The 12V line powers the graphics card, right? This was done just idle in Windows, does that make a difference?
 
Last edited:
Associate
OP
Joined
23 Oct 2004
Posts
236
Location
London
I have increased my RAM voltage to 2.0 from 1.8, will report back on whether this makes a difference on Memtest86+ and actual game stability.

I would however really like to know about those 12V values I am seeing. Anyone know about that?
 
Soldato
Joined
17 Jun 2007
Posts
9,273
I rma`d some of those a week or so ago(not had them back yet)

If stock voltage is 2.2v then 2.0v may still give errors
However i have read that at 2.2v they may not last that long mine didnt but with the amount they produce they will get manufacturing defects so i wasnt worried plus lifetime warranty:D
 
Associate
OP
Joined
23 Oct 2004
Posts
236
Location
London
At 2.0V I ran Memtest86+ for 1hr50m:

Test 8 (the one that was giving errors) ten times
Full tests 1-8 twice through

No errors! Time for some game testing.

I'd still be interested in some info on the voltages I am reading on the 12V lines though...
 
Permabanned
Joined
29 Dec 2006
Posts
3,817
Hey BlackEyedAngel, first of all, just leave one Memory module in and then make sure you have set your Memory Frequency, Timmings and Voltages correctly.

Your frequency needs to be set at 667MHZ. Your timmings should be set as follows:

CAS - 3
tRCD-3
tRPD-3
tRAS-12

And as for the Memory voltage set it to 2.1v.

Then make sure everything else is set to Auto as it should including things like your North/South Bridge etc...

After you have done all of that run Memtest. Then report back your results if you dont mind. :)
 
Associate
Joined
28 Dec 2006
Posts
587
Location
Epsom Downs
i found that fanspeed said that on of my fans was going 32,000 rpm, which is rather fast !!
also my case/motherboard was at -55 and the -12 volts was -5.
To be honest i got rid of it, cause it's rather poor at the best of times.

Can u loosen the memory timing at all ?
Also installed CPU-Z and have a look at the volts going into the CPU, as fanspeed says 1.2 volts, which is a bit strange to me !

Most of the testing software for overclockers is great and finding problems with even standard pc's.

hope this helps mate, i know how annoying it can be if ** new pc don't work.
 
Last edited:
Associate
OP
Joined
23 Oct 2004
Posts
236
Location
London
@ cooljimy84: The memory timings are quite loose already I believe (5 3 3 15). Here are my voltages per CPU-Z:

Voltage sensor 0 1.12 Volts [0x46] (CPU VCORE)
Voltage sensor 1 2.08 Volts [0x82] (VIN1)
Voltage sensor 2 3.34 Volts [0xD1] (+3.3V)
Voltage sensor 3 4.92 Volts [0xB7] (+5V)
Voltage sensor 4 0.51 Volts [0x8] (+12V)
Voltage sensor 6 -4.61 Volts [0x48] (-5V)
Voltage sensor 7 5.03 Volts [0xBB] (+5V VCCH)
Voltage sensor 8 3.26 Volts [0xCC] (VBAT)
Temperature sensor 0 46°C (114°F) [0x2E] (TMPIN0)
Temperature sensor 1 29°C (84°F) [0x1D] (TMPIN1)
Fan sensor 0 1152 RPM [0x24A] (FANIN0)

Those look pretty much in line with the ones reported by FanSpeed which I have posted further up the thread (except I have now got 2.0V on RAM).

@ Stickroad: See 2 posts up from your post. I have now stabilised by RAM in Memtest86+ by increasing the voltage from 1.8V to 2.0V. I cannot see any way to change the timings in my BIOS, and from CPU-Z's timing tables I believe that the ones I posted above are simply the default when running at the 333Mhz speed.

So, I can now get no errors in Memtest86+, but when running a game I still crash to the BSOD. This would suggest that it is not my RAM that is the problem, would you agree?

Perhaps the problem is therefore with the graphics card, software, or the power supply.

Graphics card: I have no way to test this as I don't have another card to swap in, or an alternative PCI-E rig to try this card in. Heat values discussed above, they're not ridiculously high it seems.

Software: I have done a complete reformat etc. All my drivers (BIOS, mobo chipset, nvidia, sound) are the latest versions from the manufacturers' websites.

Power supply: This made a suspicious pop soon after building the rig. My 12V line is not outputting 12V. I have no idea if this is normal, please can anyone who knows about power supplies advise?
 
Associate
Joined
27 Dec 2004
Posts
1,199
@ Stickroad: See 2 posts up from your post. I have now stabilised by RAM in Memtest86+ by increasing the voltage from 1.8V to 2.0V. I cannot see any way to change the timings in my BIOS, and from CPU-Z's timing tables I believe that the ones I posted above are simply the default when running at the 333Mhz speed.

So, I can now get no errors in Memtest86+, but when running a game I still crash to the BSOD. This would suggest that it is not my RAM that is the problem, would you agree?

But the sticks are rated for 2.2v, not 2.0v. Until you set it to 2.2v, it seems kinda pointless looking at other possible causes.
 
Back
Top Bottom