CPU Going Bad?

Alright so I recently had a problem with my PC; The I/O Panel was causing a short. Popped the I/O Panel off, and was good to go for about 2 days I think. Then I hit a dead end. The computer would boot on for a second, and shut back off. On, then off. I didnt have any spare parts to test, nor did any one I know, so I had no choice but to pay for a diagnostic test. I took the mobo,psu,cpu,ram and video card(but not my SSD RAID Array my OS is on) into the shop. I did not take my cpu cooling fan which a a Scythe 21E with a 3 pin female adapter for power.

The tech used a diagnostic card, and told me my RAM was bad. He used some Kingston value brand and it posted fine. Get home put everything back together in my case. I purchased some Kingston Value RAM from them(1066 mhz) because they were the only place in town with DDR3 RAM. I put the Kingston in planning to use them as temp ram till my G.Skills were sent back to me after doing an RMA to G.Skill(who's customer service sucks btw). Well oddly enough, I couldnt get a post with the Kingston. Same on/off problem as before. The only difference between when I put them in and when the tech used the exact same type and speed of RAM during testing, was my SSDs,HDDs,Case Fans and CPU Cooling fan run of Sys Fan 3 on the mobo. I finally disconnected the CPU Fan from the mobo as my heat sink can keep the cpu just cool enough not to over heat and shut down. It runs about 70*C with just the heat sink and a case fan blowing on it. Sure enough, it posted, so I tried running the cpu fan off Sys Fan 2 instead of Sys Fan 3 and everything seemed fine, or so I thought.

A day or two latter, I was right back where I started from. with the PC doing the on/off with the Kingston Ram. Well I just happen to get my G.Skill RMAs that next day. Put the G.Skills in, and guess what? System was posting fine.

Then, maybe another day or two latter, right back to the on/off. Seeing as the mobo had now killed 3 sets of RAM, I knew the diagnostic test missed something, but they refused to test it for free. So I RMAed my G.Skills for a second time, and RMAed the mobo. Received them Monday of this week, but theres still something wrong as I keep getting complete system freezes.

I checked the logs and both comodo firewall and vmware reported errors. So I assumed it was vmware. Removed vmware, still got a system freeze during windows. When I rebooted, it acted like CMOS was cleared as my RAM timings,voltage and speed settings had been wiped clean. Set everything up again, and had another freeze and another incident where cmos seemed to be cleared. So I though for sure it was Comodo, but before I even got to try test this theory, the computer once again froze up, this time in the bios. So that pretty much eliminated Comodo as the problem.

Since Monday, I had noticed the hang was initially during gaming, so I decided to test another game, which I was able to play for about 2 hours with no problems.

So I dont know what else to do. My CPU,Mobo,RAM,and PSU are all under warranty. The PSU passes the paper clip test, but I have received mixed opinions that my PSU was not powerful enough for my rig. The girl working sales in the store even said I needed an 800 watt psu, to which I just smirked and said sure I do. Then I realized she didnt even know how to take the video card off the mobo, and had to help here...so I put no faith in her opinion what so ever as I know for a fact I dont need an 800 watt psu. However, the multi(3) 12+V Rails in my PSU were brought up as concern as after powering the video card, only 5 amps or so would be left.


So, any suggestions on what I should test, or how to proceed seeing as most everything is under warranty?

 

Dont have one, planing to buy one soon, but dont have the money for another week or two.

 

Tried that to. Any one I know with a pc has some OEM thats several years old and the PSU would likley be to weak, and wouldnt have PCI-E power cables anyways.

For what its worth the systems been stable for 3 days except for a few reboots/power offs for testing purposes.

 

Im leaning towards a cpu problem. Seems that both sticks work fine in DIMM 3, and it seems DIMM 1 isnt registering.

 

But this is basically a brand new mobo, and the second one to suffer an issue now. The only constants are the CPU,Video Card,PSU and case. It doesnt apear to be a gpu problem. Also I think something that has to be considered is the i7 800 series is the first cpu, that I know of, to have the memorey controller in the cpu. So I *assume* its possible if the memory controller goes bad, it could cause some sort of issue with the ram in the mobo. One thing I may have forgot to mention earlier was the symptoms were a result of a bad oc, or it may have been the I/O panel short I reference earlier. Both of which were experienced on the original mobo, not the RMA mobo.

 

Im not quite sure how to take that comment as it could be interpreted as being a bit smart coming as text on a screen, so I ask just to be safe, do you still think it is the mobo, or would you agree it is the cpu with the possibility of a bad psu?

Also the cpu was using the 4pin at the time as my blind..didnt realize the mobo was capable of 8-pin.

FYI, I tested every possible ram combination, none of the failed attempts produced any warning beeps.

As we know DIMM2, black furthest from cpu socket, works. Long story short, every other DIMM failed in one of two ways. The most common, power on, stays on, but no post. The only combination that did not produce this results was DIMM4, Blue furthest from cpu socket. This test produced a scenario I have seen before. Power on for a few seconds, shut down, power back on, repeat pattern. Now I admit I may not have tested each individual stick for each scenario, but seeing as both RAM sticks seem fine, seems unlikely to be a ram issue.

 

Slot points as in cpu pins?

 

i don't know about the i7 but i had an issue with the first generation Phenoms that when you would up the core voltage it would also screw around with the Memory Controller. however this screw up would only happen on DIMM 0 and DIMM slot 1 DIMM slots 234&5 would act normally.

as it later turned out Core 0 was the only one getting the voltage boost why'll Core's 12&3 where still at default Voltages. it just so happens Core 0 in the unganged Memory Mode gets access to DIMM slots 0&1.

the CPU was fine however my little OC session Ruined my DIMM slots removing both ram sticks from slots 0&1 stopped my post hangs.

if i used DIMM slots 0&1 after they went bad, i would have to restart my computer at least 10 times before it would load into window after it loaded into windows it works fine.

 

im with DDP on this one it sounds like a motherboard Issue.