Jump to content
Corsair Community

I think I've got a faulty module


Recommended Posts

Hey folks,

 

I've been struggling with crashes and after running Memtest86 a few times I've discovered repeating faults in one or both of my Dominator sticks (CM2X2048-8500C5D).

 

I can run Memtest86 for around 13 minutes, up until Test 5, when I'll begin receiving errors.

 

Thousands of them!

 

They seem to be two specific addresses, 00009fc18e8 (159.0Mb) and 000a4c0e685 (2636.8Mb)

 

They also showed some sort of error code at those addresses:

 

00009fc18e8 = 7FFFFFFF

000a4c0e685 = 02000000

 

I've tried running the memory with SLI Ready Memory disabled and enabled, with voltage at both 2.1 and 2.2 (as recommended in another thread) and with timings at 5-5-5-15-2T and 6-6-6-16-2T (I wondered if maybe my sticks weren't quite capable of the speeds they were advertised at) to no avail, errors at the same test every time.

 

Am I right in assuming these sticks are faulty or is there something else I can attempt?

 

One more thing, Memtest86 claimed the faults would be ECC correctable, but I can't find any way of turning ECC on in my bios (I'm not too knowledgeable about memory to be honest so I'm not sure if that's even possible anyway?) nor can I find anything relating to ECC and the memory I own.

 

If anyone can help me out with some advice I'd really appreciate it, as crashing every few hours isn't exactly fun or productive.

 

KsE

Link to comment
Share on other sites

Hey Ram Guy,

 

Got my memory shipped off yesterday. I popped both sticks in an anti-static bag, wrapped that in bubblewrap and stuck that in a padded envelope, so *hopefully* I've been paranoid enough to prevent any sort of physical damage :P

 

One thing I did however forget to do was put the RMA number on the outside of the package. I knew I was forgetting something the second I handed it over at the post office to send it away.

 

Will that be a big problem? The RMA number is on the letter inside the package, along with a return address both on the outside if undelivered and on the inside too, so I'm hoping it wont be a big deal for me forgetting the RMA number on the outside of the package, but I just wanted to ask your thoughts on the matter.

 

Cheers mate,

 

Chris/KsE

Link to comment
Share on other sites

  • 4 weeks later...

Hey Ramguy,

 

I received my replacement sticks a few weeks ago and popped them in, everything seemed to be fine and dandy, but now I've started experiencing the same crashes again. Upon running Memtest86 again, It's reporting thousands of errors again, however this time rather than it being two separate places, it appears to be dozens, if not hundreds of different places.

 

I'm not sure what the problem is, my system is ventilated enough, I've tried different timings (I even lowered them this time, since I don't have one of those Dominator air coolers for my RAM) and somehow my machine appears to keep killing my RAM.

 

Got any ideas please? I'm at my wits end with this damn machine lately...

 

Cheers,

 

Chris/KsE

Link to comment
Share on other sites

UPDATE:

 

Okay, now this is weird.

 

I had my timings set at 1066mhz 5-5-5-15-2T with a voltage of 2.1v, as recommended, and memtest goes berserk, but if I set all my settings to auto (the RAM drops to 800mhz), my system stabilises (or at least appears to). I dont suppose this sheds any light on the matter does it? My motherboard is supposed to support up to 1066mhz, but I'm starting to wonder if maybe EVGA were exaggerating a little bit, as this board has been nothing but a pain in the backside. I don't mind running the sticks a little slower than they were designed to be ran at, as long as the system doesn't crash on me the way it has today, but I'm just wondering what your thoughts on the matter are please Ramguy.

 

It feels some days like my PC is a sentient, malevolant demon that just enjoys making me suffer :P

Link to comment
Share on other sites

  • Corsair Employees
Have you tried each stick one at a time in each slot to eliminate if this might be an issue with the motherboard? Make sure you set it to its correct settings and run memtest with each stick installed on its own. Let me know the results!
Link to comment
Share on other sites

Hey Ramguy,

 

I ran Windows Memory Diagnostics in Windows Vista and it came up with the simple message "Hardware fault discovered", so I took the stick out of slot 2 and ran it again and got an all clear. I then switched the sticks and tried the one I'd removed in Slot 1 and that gave an all clear too. I popped both back in and it *appears* to be working fine now, leading me to think it may have been the 2nd stick that needed re-seating in Slot 2. I may be wrong, but I'll keep you updated if anything happens.

 

One more thing, I did get a BSOD that had "MEMORY_MANAGEMENT" in it, Googling that led me to believe the stick could either be faulty, not seated correctly or perhaps the slot on the motherboard itself was faulty, but as of this moment everything seems fine and dandy.

 

I'm still running at auto settings though (800mhz, 5-5-5-18 2T 1.8v) but I'll leave it at that for now just to see how it goes. If I'm still stable in a week without a crash I'll pop it back up to the proper settings.

 

Cheers again mate, your help is much appreciated :)

 

Chris/KsE

Link to comment
Share on other sites

Hey Ramguy,

 

The problems began again, I didn't even get chance to try upping the settings again.

 

I started getting BSODs again along the lines of PFN_LIST_CORRUPT, MEMORY_MANAGEMENT and a score of other ones that I didn't write down in Windows 7 64bit (and also Vista 64bit - will get to that below) as well as the erroring in both Windows Memory Diagnostics and Memtest86.

 

I'd been experiencing oddities in the Windows 7 beta client that I still had installed and I'd decided to head back to Vista 64 anyway, but even after a complete format, I managed to BSOD during the Vista 64 installation, which is never a good sign from a hardware perspective. I took out the stick in slot 2 (which had previously been in slot 1 before this all began) and ran Memtest86 to have it error on me almost immediately. I took the stick out and reseated it, but no dice.

 

I swapped the stick for the other one and ran Memtest86 and it ran fine for 30 minutes with no problems, so I put in my other two 1gb sticks of unbranded "backup" memory (that I'd been using last time I RMAd memory to you fine folks) so my system now has

 

Slot 1: Corsair Dominator CM2X2048-8500C5D 2Gb (Running at 800Mhz/PC2-6400)

Slot 2: Unbranded DDR2 PC2-6400 1Gb

Slot 3: Unbranded DDR2 PC2-6400 1Gb

 

So far I've been up and running for around 8 hours without a single problem, leading me to believe that Slot 2 on the board is fine, as is Slot 1 (and apparently Slot 3 too, which I've not used before), and that the stick is faulty.

 

The thing I'm concerned about is why this happened. It seems a bit odd that my memory would keep dying this way. My case has multiple fans, a passive-cooled CPU, the side off (which admittedly will hinder overall airflow but due to the size of the CPU heatsink I literally have to run it with the side off as it causes problems if I don't), about the only thing I don't have is the Corsair Dominator airflow fans, as with the way my case fits I don't have room above my memory to have anything attached, so I didn't buy the matched pair set with the additional airflow fan system. I didn't think that was an absolute necessity for running it at it's standard settings, I just assumed it was for if you overclock the RAM, but now I'm beginning to have my doubts.

 

Anyway, if you can advise in any way I'd be very grateful please mate.

 

It's getting to the point where I'm considering buying a different cooler and a larger case so I can have a bit more space to play with.

Link to comment
Share on other sites

Hey RAMGUY,

 

My system's been running under the current setup (Working Dominator module in Slot 1 and the two cheap sticks in slots 2 and 3) for a week now with zero problems, other than it being a bit slower than I'm used to due to obvious reasons.

 

The slots seem to be fine, leading me to think perhaps I've just been horrendously unlucky and just received two bad packs of RAM.

 

I am however going to invest in a Corsair Airflow fan for my RAM and a new smaller cooler for my CPU to free up the space required to attach it to my Dominator sticks, but would it be possible to arrange another RMA please?

 

Hopefully the problems wont continue.

Link to comment
Share on other sites

  • 1 month later...

Hey Ramguy,

 

I've started experiencing errors again with the newest batch of RAM that I received from Corsair's RMA department...

 

The thing is, I've done everything you've asked of me, and my memory continues to keep dying on me. I just forked over £88 for a brand new motherboard, I replaced my CPU cooler to increase airflow in my case, I bought a Corsair Dominator Airflow fan and still, 1-2 months after receiving the replacement RAM the sticks begin erroring with BSODs along the lines of "PAGE_FAULT_IN_NONPAGED_AREA", etc.

 

I checked my Computer Management logs and under my System logs I'm getting an 1801 Event ID describing "The hardware has reported an uncorrectable memory error." This was the last event to happen before the BSOD that crashed my system last. The weirdest thing about it though is that it happened around 4 minutes prior to the BSOD actually taking place. The error is listed at 23:56, but the PC BSOD'd at exactly midnight.

 

Is there a possibility some of this could be maybe PSU or Hard Drive related at all? I've had my doubts about the HDD for a while, but I've had no data loss or corrupted sectors and I keep it regularly defragmented, and as for the PSU it's far more powerful than my system requires and I've had no tell-tale signs of PSU trouble such as automatic reboots and failing to power on.

 

If you have any ideas at all I'd be glad of your input please, as I really dont know what else to do short of avoiding Corsair RAM in the future, and your customer service has been so exceptionally good that I'd really rather not avoid the brand all-together.

 

That, and it's kinda pricey sending it back every time at £10 a pop recorded delivery ;)

Link to comment
Share on other sites

Hey again Ramguy,

 

I ran Memtest overnight and recorded a whopping 890 million errors on both sticks, so I tried reseating the RAM, altering the voltage as high as 2.2v and as low as 1.8v and none of it made any difference. Every time, the memory errored within 2 minutes, sometimes as early as 20 seconds in.

 

I decided on a whim to flash my BIOS, hoping to find a later version of BIOS for my board, and while the program said there was no update, it reflashed the current BIOS fresh and rebooted.

 

I immediately ran Memtest on a hunch and it ran for 24 minutes with no errors before I decided to end the tests early and log into Windows to search for answers.

 

I'm wondering if all of this could be a nasty BIOS related error, perhaps a virus or something along those lines, I don't suppose you've got any ideas do you? Can you recommend any software I can use to prevent my BIOS from being infected? I've got Avira antivirus running around the clock, but apparently it missed it (that, or it doesnt check the BIOS), and I've not altered or tampered with my BIOS in any way since putting the newest sticks in, making me unable to suggest it was something I did manually.

 

I feel more positive this morning about the whole thing and I apologise for the tone of my previous post (if there was a tone), and hopefully we can resolve this issue without any more expense on either of our parts ;)

 

A BIOS virus would explain in theory why my memory was erroring, but it wouldnt explain why the other memory (the cheap samsung chipped DDR2-800 I own) runs fine on it despite being both covered in dust and only worth around £15 a stick per 1Gb hehe.

 

Anyway, if there's any advice you can offer I'd be deeply grateful.

 

Chris/KsE

 

EDIT: Started erroring in Memtest again around a minute into the tests. Officially out of ideas.

Link to comment
Share on other sites

  • 4 weeks later...

Hi Ramguy,

 

Well, the Gigabyte MB I just bought is brand new and has zero faults, as well as being able to handle way faster RAM (DDR2-1336+, according to Gigabyte's website: http://www.gigabyte.com.tw/Products/Motherboard/Products_Overview.aspx?ProductID=2951 ) than these sticks require, leading me to believe it's more than likely the PSU as that's the only thing I haven't replaced since this memory death problem began. I've also updated the BIOS to the latest revision also, which hasn't resulted in any changes to how the RAM is behaving.

 

I've put an order in for a Corsair HX650W PSU as I've read that it's a hell of a piece of kit (and kinda covers my bases a little hehe ;) ), hopefully that will solve all of the problems, but I'm imagining that this RAM I currently possess is already dead to the world. It's still weird though, that this cheap Samsung RAM that I bought from Ebay 2 years ago is still working fine.

 

Is it at all possible that perhaps the Corsair RAM I'm getting is just of a faulty design? I don't want to put you on the spot or anything in front of your higher-ups, especially with this job climate at the moment, but a lot of companies do have the odd batch or brand of product that either shouldn't have passed the QC labs - but did - or didn't and the company in question has just figured "sod it" and put it out anyway. I don't mean to point a finger at Corsair and say that you guys do that, but it does strike me as odd that these expensive and supposedly well-built sticks that I paid over £80 for keep dying, yet the sticks I paid £15 for 2 years ago can stand the test of time as well as supposedly faulty motherboards and PSUs...

 

Anyway, enough of my conspiracy theories. Thanks again for the help Ramguy and sorry for the late response, been trying to work out when I could afford to replace my PSU (I'm on a very limited budget, hence my apprehension to replace parts "willy-nilly") and that time just arrived so I figured I'd let you know.

 

Is there any chance I could have your blessing (hopefully) one last time to get these replaced again please? All being well, if the PSU is the culprit then the problem should be solved relatively soon.

 

Thanks again Ramguy,

 

Chris/KsE

Link to comment
Share on other sites

  • Corsair Employees
With these modules and that MB you will have to set the tested settings manually if you leave it set to AUTO it will run them at DDR1066 with out setting the Voltage and that will not work and may be why you are having problems, did you set the memory Voltage manually to 2.1 Volts?
Link to comment
Share on other sites

With these modules and that MB you will have to set the tested settings manually if you leave it set to AUTO it will run them at DDR1066 with out setting the Voltage and that will not work and may be why you are having problems, did you set the memory Voltage manually to 2.1 Volts?

 

Yep, I set the voltage to 2.1v manually in the BIOS and CPU-Z indicated it was definitely running at 2.1v after setup with all the correct timings at 1066.

Link to comment
Share on other sites

Hey Ramguy,

 

Thanks again for all your help, just wanted to let you know that I've requested an RMA, but I think I entered the wrong part number so I've just re-requested an RMA with the correct one (I'd chosen CM2X2048-8500C5D instead of TWIN2X4096-8500C5D - that is the correct designation for the CM2 matched pair, right?), hopefully there wont be too much confusion and I do apologise again for increasing your guys' workload this holiday season hehe.

 

Thanks again,

 

Chris

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...