Travel back in your time machine's with me to the socket-775 era!

[The_Lanbone]

Guys I feel like a total newbie for having to resort to this but I'm at a total loss of what to do next, so I turn to the community to help me with my (probably simple) problem! I'm going to attempt to be as detailed as possible so this might be pretty long/boring/confusing. Any questions that might help you diagnose my problems please just ask. Here's the whole story:

Earlier this week my buddy sent me some old parts he had sitting around his house: an EVGA 790i SLI FTW motherboard, and an old Tuniq cooler. He upgraded to a UD3/2500K setup and pulled the board/cpu/cooler out as a straight swap. It was used as his main PC and worked without issues. I had picked up a second-hand Q6600 quad-core for next to nothing a while back and had popped that into my old and outdated Dell XPS 720 where it seemed to work just fine. It ran for at least a month or two playing the occasional game and doing a lot of F@H without issues. Here is where my problems start...

I felt it was time to put that board to use so I popped it in a little mid-tower case along with the following parts that were pulled out of my functioning XPS 720:

EVGA 790i SLI FTW

Q6600 G0

8GB Kingston 1333Mhz 9-9-9-24 Memory (This my buddy just tossed in with the parts he sent me)

GTX 465 1GB (x2 - Only had one running in the Dell due to a shitty PSU)

Corsair AX850 (New as of Friday 02/10/2012)

Tuniq 120

10000 RPM WD VelociRaptor 300GB (Sata 3 drive)

Noctua NT-H1 TIM

I'm probably forgetting something but those are the main players. Now here's one of my big concerns: I know what most of you are going to say, and there isn't a whole lot I can do to change it right now, but I am using a stolen 'borrowed' copy of Windows 7 Ultimate x64. I am all but convinced this is not the issue as I have tried multiple versions, 32 and 64-bit, and the problems still persist. It's very hard to explain in exact detail what all is happening but I have been getting a TON of random freezes, lockups, 'Windows 7 failed to load' errors. They have been happening at all different times; boot, installing the OS, loading the OS, while working in the OS. I have reseated the CPU twice (after about 7hrs of troubleshooting), checked all connections, and up until 3 days ago, thought the CPU was working as it should. I have got to the point where W7 installs, loaded the OS, and begun my installation of software (Mobo drivers, GPU drivers, CPU-Z, HWMonitor, Prime95, etc etc). Another thing that concerns me are temps. Idle temps with the Tuniq are roughly 28-30c which is slightly lower than it was when I had the stock Dell XPS720 cooler and cheap TIM on, but if the CPU is doing anything other than idling (mind you these are at stock voltages and stock 2.4Ghz) the temps sit at roughly 46-52c on the hottest core. This is at any load above idle. 5mins into a Prime95 run and the temps hit 62c with ease and were on their way up. The old Dell with the Q6600 only hit that number (and nothing higher) after a 15-min blast running OCCT. Temp issues I think were caused by a crappy fan in place of the Tuniq one which was not powering on for some reason (since fixed the issue, now running standard fan included with heatsink). I have no way of looking at the temps other than the idle temps showing in the BIOS (which is where I'm getting the 28-30c) as now I can't even get the system to boot and stay stable enough to either reformat & install windows, or just get into windows long enough to install HWMonitor and view temps.

I'm sure I'm leaving out a ton of details but I'm beginning to feel like I'm talking in circles and not really explaining things correctly. I'm all but convinced this board is bad somehow but the only way of testing for sure would be to swap out the quad-core for an old dual-core I have lying around. The other kicker is this: I'm running pretty low on TIM, I think I have enough for one more heatsink install, maybe two. I have a very strong feeling I messed something up or I'm just forgetting something really stupid and simple. I know just enough to get myself in a fair bit of trouble, but not enough to fix whatever problems I create! (Clearly!!!) I started this project at 7pm EST this Friday and have been dealing with it since I gave up at about 2am this morning so I could go to bed and come to work. I humbly turn to the OC3D crew to help me fix whatever mess I've created and get back to F@H and playing racing games poorly.

Help me OC3D, you're my only hope!!! (Yes, I just went there hahahaha)

Michael

 

[nepas]

Didn't the first 790i's have a problem with data corruption,that seems to ring a bell somewhere.

 

[The_Lanbone]

I was wondering about that actually. I had been doing some reading and seen people complaining about BIOS issues when overclocking, but I can't even get that far! A lot of this article talks about how they were only using 2 then down to 1 stick to avoid the FSB hole, but that they still lost some OS installs. I never tried going to just 1 stick of RAM. Guess that's the next step... http://www.anandtech.com/show/2660/7

 

[nepas]

I was wondering about that actually. I had been doing some reading and seen people complaining about BIOS issues when overclocking, but I can't even get that far! A lot of this article talks about how they were only using 2 then down to 1 stick to avoid the FSB hole, but that they still lost some OS installs. I never tried going to just 1 stick of RAM. Guess that's the next step... http://www.anandtech.com/show/2660/7

Good point,I knew it was to with that board.did,nt they release a revision/Bios update that kinda solved it(am drunk so did not read the link,sorry)

 

[S_I_N]

you say 8GB of memory but not its configuration. is it 2x4gb sticks 4x2gb sticks??? I ask because I'm getting the same board next week for my backup with an E8400 and the guy I'm getting it from says the 790i SLI FTw doesnt play nicely with 4gb sticks

 

[The_Lanbone]

Sorry, its 4x2GB sticks. I just tried yanking 2 then 3 sticks out to see if it would boot. It got through about 2/3rds of the windows install. When it needed to restart to finish I got the windows failed to start' error.

 

[Lollipop]

Have you tried Linux to make sure it isn't your particular w7 copy?

 

[S_I_N]

ok do this take power cord from back of psu flip off switch if there is one. remove battery/jump clear cmos jumper if applicable for 2 minutes. I mean a full 2 minutes longer is ok shorter is not. put jumper back if applicable replace battery plug back in cord flip switch on psu back on restart go into bios make your setting changes and try again.

 

[The_Lanbone]

So I didn't want to say anything last night in fear of jinxing it. I think I did anyway. I gave what you said a shot S_I_N and it actually worked! Got W7 installed and working just fine. Updated, drivers installed, programs installed, everything running fine. Temps stayed below 40c for normal use (that Tuniq is LOUD at 12V) ran Prime95 to see how stable it was and it was running like a champ. Or so I thought. I played around for another hour or so and all was well, everything seemed like it was back to normal but I wanted to be sure so I planned on leaving prime running overnight (that meant 4hrs last night) and at sometime after the 1hr mark it shut down and seems to be back to it's old tricks. When I tried booting it back up this morning I started to get the same little glitches, eventually followed by a windows lock-up. I'm going to try and clear the CMOS again tonight before booting up to see if this helps anything.

 

[Techie103]

I would throw a new battery in there for shits and giggles.

 

[Grizzly]

Make sure you're running the latest BIOS revision once you get the board working again.

If the problem is as Nepas said a BIOS update could solve the problem.

 

[The_Lanbone]

Just wanted to update you guys real quick. All is not still well. I tried a few more long-shots last night with no luck: replaced CMOS battery just to be on the safe side, pulled out 2nd GPU, reset CMOS, restored default BIOS settings, pulled out 2 sticks of RAM leaving only 4GB @1333Mhz. I also heard some talk on the EVGA website about the main problem being the JMicron eSATA controllers causing data corruption and disabling them might help; it didn't in my case.

Oddly I tried unlinking the FSB and memory and upping the FSB to 1333Mhz with an 8x multi and was able to get through an entire W7 install, loaded settings, and when it finally booted into windows (This was tried with both GPU's installed but still only the 2x2GB sticks) it worked fine for all of 5 minutes before freezing. Once I did this I completely started over from scratch: 1 GPU, 2x2GB RAM, CMOS batt out, clear CMOS for 60secs, disabled JMicron eSATA, and got windows installed, flashed EVGA E179_SZ16 BIOS, and was able to install nVidia display drivers, windows updates, HWMonitor (showing mid/low 30c CPU temp at idle). Another freeze, and I haven't been able to get windows to load, repair, or install correctly since.

I figured I'd post up my BIOS settings to see if anyone sees anything wrong. They are all but completely stock settings, so I can't see why any of the settings would be causing a problem.

**System **



CPU Freq, MHz: 2.4 GHZ

FSB Reference Clock, MHz: 1067mhz

CPU Multiplier: [9X]

PCle x16-1 & 16-2, MHz: [100]

PCle x16-3, MHz: [100]

SPPc->MCP Ref Clock, MHz: [200]

nForce SPP --> nForce MCP: [5X]

nForce SPP c-- nForce MCP: [5X]

**Spread Spectrum**

CPU Spread Spectrum: [Disabled]

HT Spread Spectrum: [Enabled]

PCle Spread Spectrurn(SPP): [Enabled]

PCle Spread Spectrurn(MCP): [Enabled]

SATA Spread Spectrum: [Enabled]

**FSB & Memorv Config**

FSB - Memory Clock Mode: [Auto]

FSB - Memory Ratio:

FSB (QDR), MHz: [****]

Actual FSB (QDR), MHz: 1067mhz

x MEM (DDR), MHz: ****

Actual MEM (DDR), MHz: 1333mhz

P1: [AUTO]

P2: [AUTO]

**Memory Timing Setting**

Memory Timing Setting: [Auto]

tCL (CAS Latency): [Auto] 9

tRCD: [Auto] 9

tRP: [Auto] 9

tRAS: [Auto] 24

Command Per Clock: [*T]

** Advanced Memory Settings **

tRRD: [Auto]

tRC: [Auto]

tWR: [Auto]

tWTR: [Auto]

tFAW: [Auto]

tREF: [Auto]

tRFC:

**System Voltages**

CPU Core: [Auto]

CPU FSB: [Auto]

Memory: [Auto]

nForce SPP: [Auto]

nForce MCP: [Auto]

GTLVREF Lane 0: [Auto]

GTLVREF Lane 1: [Auto]

GTLVREF Lane 2: [Auto]

GTLVREF Lane 3: [Auto]

**CPU Configuration**

Limit CPUlD Maxval: Disabled

Intel Speedstep: Disabled

xPPM Mode: [Native]

CPU Thermal Control: [Disabled]

C1E Enhanced Halt State: [Disabled]

Execute Disable Bit: [Enabled]

Virtualization Technology: [Enabled]

CPU Core 0: [Enabled]

CPU Core 1: [Enabled]

CPU Core 2: [Enabled]

CPU Core 3: [Enabled]



As always guys, any thoughts or suggestions are helpful. I am a defeated and depressed man!

 

[JMMP]

when it freezes does it give you a blue screen?

if so , what kind of BSOD code error appears in the blue screen?

Example:

Stop: 0x00000050 or 0xF9DDB962

 

[The_Lanbone]

No BSODs at all actually. I hadn't realized that till just now, but throughout all the issues over the last 8 days and constant freezing and booting errors I have yet to see a BSOD. Most times when it locks up while in windows it just stays frozen. I've let it sit there for a good 10-15mins before just to see if it would show some type of bsod but it never did.

 

[marsey99]

i would look at ram or hdd tbh as they are what it screams at me.

that data issue was because some of the ref design boards had rubber spacers under the heatsinks in the wrong place and it was the early signs that the chipset was frying itself. with this board being from a running machine i would guess this aint the case here.

 

[The_Lanbone]

Agreed, the only thing that confuses me is that my HDD came directly out of a 110% working system, and put right into this one (with a reformat and fresh install obviously) whereas this board/RAM combo sat around for 3-4 months before my buddy finally got off his butt and sent it to me. H did fail to mention that he was aware the board wouldn't run with all 4 dimms populated. I've got it up and running right now but without 1 stick of RAM, and without the 2nd GPU. Every time I put the 2nd card back in to SLI I start getting boot errors again, even to the point where it won't pass the initial POST memory test. I've swapped out the GPU's and can confirm they both function, as I can boot into windows with the 2nd card as the primary (and only).

 

[JMMP]

hi!

did you try to unistall all nvidia drivers/controlers in safe mode and start a clean nvidia driver instalacion with both cards in place?

 

[The_Lanbone]

Made the mistake of restarting to install the 2nd GPU after 72hrs of stable running, prime95, F@H, IBT, and Heaven (just for fun). The entire goal of this was to be able to run nVidia Surround for some triple monitor racing games as cheaply as possible with the 2 GTX465's. Obviously getting the system stable with both cards in is a high priority, but I am struggling to even get it to post with one card now.

The system constantly hangs at C1 now and gives a 3-4sec long repeating beep. Manual says 'Memory Presence - Base memory detect'

 

[marsey99]

you can get sli working on any chipset pretty much now dude, i have seen people sli on p35/45/x38/48 as well as the later intel chipsets via the right drivers but it can take some time for these drivers to get updated after nv release a new driver family.

anyway, back to this.

does the board feel warm when you are using it? the nb/sb/vrm heatsinks do they all feel warm or hot or cold?

 

[Terrifictez85]

I had an issue like this a while ago, Has the hard drive got a set of jumper pins? If so try install windows on it in slave instead of master. Did the trick for me.