Memory failed Memtest now works???

Status
Not open for further replies.

footballcoach

Honorable
Apr 27, 2012
3
0
10,510
Hi all,

I have a Maximus III Formula and have a technical question regarding memory.

Relevant specs:

ASUS Maximus III Formula LGA 1156 Intel P55 ATX Intel Motherboard
G.SKILL Ripjaws Series 8GB (4 x 2GB) 240-Pin DDR3 SDRAM DDR3 1600 (PC3 12800) F3-12800CL9D-4GBRL
Intel Core i5-750 Lynnfield 2.66GHz 8MB L3 Cache LGA 1156 95W Quad-Core Processor
Corsair Cooling Hydro Series CWCH50-1 120mm High Performance CPU Cooler


So, over the last 3-4 days I started to experience frequent BSOD and/or restarts. After a system restore did not resolve the issue I ran a debugger on the .dmp files and the BSODs were caused by different applications, but all pointed to having an issue with memory R/W.

I had been overclocking the CPU for about 4 months with the Asus BIOS overclocker Auto-tuning feature. However, the memory has always been running at 1600 rated spec @ 1.51v. After BSODs started I removed all CPU overclocking and am now running at stock speeds. Memory is still at 1600 and 1.51v.

So, I set about testing the memory. I booted Memtest x86 ISO and immediately racked up 1700+ errors. So, I thought probably bad ram. I took the PC outside and blew the dust out with my air compressor and then re-seated the 4 memory sticks in the original slots so I could keep track of which stick might be bad. Ran Memtest again and it failed immediately in similar fashion.

So, I pulled the two original black sticks [A1B1=Red, A2,B2=Black]. I ran Memtest with the red sticks and all tests were OK. So then I swapped the black sticks into the red banks. Ran Memtest and all tests were OK again. So, at this point I am a bit puzzled and considering that one of the A2D2 slots may be bad.

I then took the red sticks and put those in the black banks. So, at this point I am fully populated again but the original sticks have reversed banks (i.e. red in black, black in red). Ran Memtest and all tests are now OK. WTF?

Could maybe during the re-seating process I accidentally adjusted the heat spreaders when pushing the memory into the banks? Maybe a spreader that was loose? Would that cause something like this?

Thanks for any help or suggestions. I went out and bought a 16GB upgrade which I am tempted to use, but if the memory I have is going to be OK I will probably save the cash.

Thanks!
 
Solution
I have had the exact same problem with my newly bought asus p7p55d a year and a half ago,

Put four sticks of 4GB in the black and blue slots, the thing wouldn't even boot.
Ran memtest+ and got loads of errors. troubleshooted with one stick at a time and everything was fine, tried 2 sticks in blue slots, booted just fine, same with 2 sticks in black, worked too. For some reason when I changed sticks from the black to blue slots it worked like a charm with a full complement of RAM.

Never experienced RAM orientated BSOD's after that.

Then a while ago I flashed my BIOS to the newest version and switched the sticks back for checking purposes, now it doesn't matter where I put them, but it's good to know that fixed the error cause it's...

footballcoach

Honorable
Apr 27, 2012
3
0
10,510
Right, that's kind of where I am at too. I expected to find a bad stick. Keep in mind the memory tested bad *after* I reseated them in the original slots.

It was only when I moved them to different slots and tested them in pairs that everything is OK now.

I think it's got something to do with the heat spreaders. I didn't notice anything abnormal, but I didn't look too closely either.
 
I've had lots of experiences with RAM where all it takes to fix it is a gentle tap on the heat spreader. The days of quality DDR2 DIMMs with tank armor level heat spreaders (Patriot extreme comes to mind) are long gone. Even extremely reputable manufacturers such as Corsair and Mushkin use this flimsy pressure fit spreaders with no paste or anything
 

blackmidian

Honorable
Apr 27, 2012
2
0
10,520
I have had the exact same problem with my newly bought asus p7p55d a year and a half ago,

Put four sticks of 4GB in the black and blue slots, the thing wouldn't even boot.
Ran memtest+ and got loads of errors. troubleshooted with one stick at a time and everything was fine, tried 2 sticks in blue slots, booted just fine, same with 2 sticks in black, worked too. For some reason when I changed sticks from the black to blue slots it worked like a charm with a full complement of RAM.

Never experienced RAM orientated BSOD's after that.

Then a while ago I flashed my BIOS to the newest version and switched the sticks back for checking purposes, now it doesn't matter where I put them, but it's good to know that fixed the error cause it's been a mindboggling issue for a long time.

P.s. praise yourself lucky with that watercooling, it's small on the cpu block. With a Scythe Mugen 2 it's a complete different story.
 
Solution
Status
Not open for further replies.