2 pairs of TWIN2X4096-8500C5D issues

hi. for the last month of so, i have had one hell of a time sorting out the problem in my pc, with the shop that i bought the components from being unable to find a fault. through trial and error i have figured out that it is the RAM (it doesnt fail memtest86+ if you just run the test, but if it bluescreens, and you then run the test, it fails horribly).

suffice to say, the shop wont do anything, since they cant reproduce the fault.

i must warn you however, that it wont fail memtest if tested straight away, it needs to bluescreen the pc first, which can take several hours, and then run the memtest, without powering off the pc.

the motherboard has been replaced, graphics card has been tested, hard drives and OS have been tested, CPU has been tested, PSU has been replaced. essentially everything has been replaced except the RAM.

the modules are running at their listed settings (2.1v, 5-5-5-15, 1066).

the settings check out fine as they worked fine for the first 2 months.

hence, i am here, requesting an RMA. i have an image of the memtest 86+ report that is attached here: image of memtest

i thank you for any assistance.

i tried that too, it still crashes every now and then.

plus, it worked perfectly for the first 2 months, before it started to BSOD.

so tried nearly everything, and they still BSOD.

the motherboard has already been replaced, and according to memtest86+, all of the RAM modules are at fault:

http://9363wa.bay.livefilestore.com/y1p7D0_2NNxdrnW7mPfCU8tMrSZAyzmzuvQs0KsFOydBIJHrTHU3Z198epDJqdTIlam2K3YQx7Ige7k2SofddChXQ/DSC00098.JPG

apparently its 1, 3, 5, and 7 GB areas that are faulty, which would imply all of them.

processor, GPU, PSU, and hdds have been tested and are fine.

That test it failed on was a random pattern over the 8 GB, so there's no way to determine which stick / slot has an issue.

Try reseating the memory.

That test it failed on was a random pattern over the 8 GB, so there's no way to determine which stick / slot has an issue.

 

Try reseating the memory.

its been re-seated about 4 times. will try testing 2 at a time.

for some reason, dont know whether it was the NB voltage increase, but it seems to be stable now, even at 1066mhz. hasnt crashed once, in 2 days.

now i would just like to know, why it worked perfectly for the first 2 months at stock NB volts (memory at 2.1v, 1066, 5-5-5-15), but now the NB needs to be boosted (motherboard is new)?

As I stated before when mixing memory you should set the next lower speed grade and officially mixing memory like this is not supported.

that still doesnt answer my question, and im not mixing any. they are all the same.

again, as said before, the motherboard was replaced!

OK. I read that and it does not matter. That does not mean that the above scenario is wrong. And, no 2 boards are the same.

Your board comes from the manufacturer with the memory controller voltage set to run 2 modules at JEDEC specifications. Any time you go beyond JEDEC spec or add modules, the load on the memory controller increases. This is why we only validate a single kit of 2 modules on boards like this. This is also why you may need to increase the memory controller voltage with 4 modules.

What RAM GUY meant about mixing memory is that you have put 2 seperate kits of memory together which means you are running them in an untested configuration. The fact that they are the same modules is not relevent here. This issue resides with the limits of your memory controller.

its still crashing at 800 with 2.1v 5-5-5-15, and NB voltage at 1.3v (stock is 1.1v).

somebody else had the exact same problem as me on here.

everything else has either been replaced, or has been tested fine.

EDIT: crashed again, and ran memtest, came back fine. must be the app.

With that MB and CPU I would set the memory frequency at DDR667 if you are still having problems. But it sounds as there may be some problem with the MB Memory controller if they all pass one at a time.

they all pass at the same time, with the NB voltage boosted. and as i said, the motherboard has been replaced, so it cant be a faulty memory controller.

the only thing that has fixed it, is incresing the NB voltage, and as said before, the memory is the only thing that hasnt been replaced.

so im stuck?

i am 100% sure that it isnt the motherboard, since it was replaced already, and it was still happening, leaving the RAM as the only component that hasnt been replaced (the shop were unable to help), or replaced, and it has started blue-screening again.

quoting "yellowbeard":"Electronics degrade over time. 4 modules is quite a heavy load on the memory controller. It's possible that as you use the system, and the motherboard wears in, it becomes slightly less efficient and you need the voltage bump."

and when i said that the motherboard was new, i was told it doesnt matter, contrary to what was said.

somebody else had the exact same problem on this very forum, if i can find it again, then i will post the link here.

so im stuck?

i am 100% sure that it isnt the motherboard, since it was replaced already, and it was still happening, leaving the RAM as the only component that hasnt been tested properly (the shop were no help at all), or replaced, and it has started blue-screening again.

 

quoting "yellowbeard":"Electronics degrade over time. 4 modules is quite a heavy load on the memory controller. It's possible that as you use the system, and the motherboard wears in, it becomes slightly less efficient and you need the voltage bump."

and when i said that the motherboard was new, i was told it doesnt matter, contrary to what was said.

 

somebody else had the exact same problem on this very forum, if i can find it again, then i will post the link here.

I'm confused as to what the current question or issue is. You added 4 modules, not stable. You raised the memory controller voltage, it's stable.

theres your answer. i already got the motherboard changed, upped the NB voltage at ram guy's recommendation, and it has crashed again, with memory management as the cause on the blue-screen.

i did run mem-test, but decreased the frequency back to 667 from 800, but if the power is cut from the RAM between blue-screening, and the mem-test, it doesnt show a fault.

With 4 modules I would suggest setting the memory frequency at DDR667 and set the memory Voltage to 2.1 Volts and set the NB/MCH/SPP Voltage to +.2 Volts as well and test the system with http://www.memtest.org. In addition, with some MB's (Mostly ASUS) you have to disable legacy USB in the bios when running any memory test.

In your post above, # 19, I thought you said it is now stable :confused:

Have you disabled Legacy USB Support while running Memtest?

Can you also describe what specifically is happening when you have system instability?

In your post above, # 19, I thought you said it is now stable :confused:

 

Have you disabled Legacy USB Support while running Memtest?

 

Can you also describe what specifically is happening when you have system instability?

it was stable, but evidently, if the RAM modules are to blame (the motherboard is new), then they must have degraded a little more, or i hadnt left it on for long enough to bring it about.

i dont think disabling legacy usb will make any difference to the stability since ive had the same configuration since i started using this motherboard, and it worked fine to begin with (2 months), and memtest seems to run fine with it on anyway, but i have disabled it regardless.

as for what is happening, if i leave it on long enough (usually between 8-10 hours) it will blue-screen, and if i run mem-test immediately after the blue-screen, i will get a sea of red (see here. i know this is with 1066, but i got the same thing at 800mhz).

the motherboard has already been replaced, the operating system has been reinstalled (4 times), the CPU has been checked (its fine), the hard drives have been checked (fine), the graphics card has been replaced, and the PSU has been replaced. the only thing that hasnt been passing all tests, or hasnt been replaced is the RAM.

please help, this is the only thing that it can be now.