Asus Sabertooth 990fx (v1) and RAM issues? (BSoD in a 2.5 year old build)

rhase

Honorable
Jan 19, 2014
4
0
10,510
Start right off with the system specs:

AMD Phenom II x6 3.3ghz 1100T (Thuban)
Asus Sabertooth 990FX (First generation of the board, NOT r2)
Corsair Vengeance 16gb (4x4g) - although currently only half the kit (2x4GB)
EVGA Nvidia GeForce 560 (ftw+)
Samsung 128GB SSD
1TB 6gb/s 7200rpm Seagate HDD
Corsair 650watt PSU (Enthusiast series)
Windows 7 Ultimate Edition (64 bit)

I built my rig over 2.5 years ago. It ran smooth for about a year, then suddenly I started having BSoD errors. Ran memtest, and sure enough had a bad stick of RAM. Tested them individually, narrowed it down to just one stick. It failed in multiple slots. I figured that was the end of it; RMAd the entire kit, installed the completely new kit and tested it extensively. It passed without any errors.

Two weeks later, same BSoDs. Run memtest again on a whim, and sure enough, it is the same slot that failed last time; hung up and was unable to finish in multiple slots. (the test was still running, but reporting thousands of errors per second and unable to complete, so I manually stopped it).

Started to suspect the mobo/PSU; nobody's luck is THAT bad. RMA'd the ram kit again (thanks corsair!) and installed only half of the third kit after testing, avoiding the "bad slot" all the while. By only using half of my kit, I achieved stability for about a year. I figured it was that slot frying the RAM somehow, and if I avoided it all was well.

Today, after a year of stability, I'm playing skyrim when suddenly BSoD. I haven't tested my RAM yet, as it's sunday and I am so not in the mood to troubleshoot (I wanna game!), but I suspect it'll fail again. I'll update, but that isn't exactly the point of this post, anyway. There is an underlying issue I need to stop ignoring.

I do not overclock. All RAM is set to auto (And it auto'd to the proper settings: 1333 and 1.5v), and the "ez BIOS" setting is on normal; no board-regulated under or over clocking. I have never flashed my BIOS. I'm posting from the machine in question; it still runs fine. The BSoD was very random, as it was before. If it follows pattern, though, they'll become more frequent now that one happened.

My question(s):

Should I RMA my motherboard? I don't want to do this lightly; I hear Asus has fallen hard in customer service, and I worry I'll get a broken "refurbished" mobo or even worse troubles than I have now. If I RMA I want to be beyond a shadow of a doubt certain it is the mobo.

OR (more preferably) is there a way I can re-stabilize my system?

Is there a known error with the board and RAM that the BIOS updates have fixed? And does flashing your BIOS void your warranty?

Perhaps there is an issue with the board's automatic settings, and manually setting voltages and timing would help?

Is there something I am completely overlooking? Can the CPU or PSU still be the culprit?

Admittedly, this is my first build so I am very inexperienced compared to most!

Any thoughts appreciated. Will do suggested testing tomorrow after I've enjoyed my day off. c:
 
Solution
BIOS updates are not only used to fix RAM settings, they are used to workaround known bugs in the CPU electronics, and chipsets used on the motherboard. If you have issues that look like hardware issues picking up the BIOS updates is one of the first things you want to do.

Most bugchecks are not caused by hardware defects. Even a bugcheck that indicates memory corruption does not mean the system is complaining about your RAM. It is complaining about a location in memory that has the wrong contents in it.


Should I RMA my motherboard?

I RMA'd my 990fx R 2.0 with no issues whatsoever. Took a couple weeks and I had use to use a Pentium III while I waiting.

Did you set the BIOS to the correct RAM settings, or did you just plug and play? You need to manually enter the RAM speed in the BIOS.
 

rhase

Honorable
Jan 19, 2014
4
0
10,510


Thanks for the report on an Asus RMA going smoothly, firstly!

I let the board auto-detect everything then double checked it. It had recognized everything just fine, and set everything to what it was meant to be set at, so I let it be. Would the board randomly recognize the RAM improperly on a boot and change all the settings or something? As a point of curiosity.
 
I doubt. I was overclocking the RAM on my board and bricked it somehow. It RMA'd fine.

I'm thinking there is a voltage issue with the DIMM slots that's causing the RAM to become damaged. Did you confirm the voltages are set correctly for the RAM as well? Switch the UEFI BIOS to advanced mode if you haven't yet.
 

rhase

Honorable
Jan 19, 2014
4
0
10,510


I have. They are all set to 1.5V. I believe this is correct; it's what the vengeance kit claims to run at and what the mobo's compatible memory table puts them at anyway.
 
BIOS updates are not only used to fix RAM settings, they are used to workaround known bugs in the CPU electronics, and chipsets used on the motherboard. If you have issues that look like hardware issues picking up the BIOS updates is one of the first things you want to do.

Most bugchecks are not caused by hardware defects. Even a bugcheck that indicates memory corruption does not mean the system is complaining about your RAM. It is complaining about a location in memory that has the wrong contents in it.


 
Solution

rhase

Honorable
Jan 19, 2014
4
0
10,510


Thanks for the reply! Assuming it doesn't void my warranty (I don't think it does, but a quick call can confirm that for me) I'll try flashing my BIOS tomorrow and update. I bought the board when it was fairly new, so it's running one of the very first versions. Maybe something this simple really is all it needs.

And Alec, I don't think they make my board new anymore so if I RMA'd I would likely get a refurb. That, or I would get incredibly lucky and they'd send me an R2! I don't mind troubleshooting, really. I'm pretty patient and enjoy the experience it gives me. My issue is not particularly frustrating... yet. I just ran out of ideas. And I still have another 2.5 years on the warranty. I'll definitely RMA if it becomes more frequent and nothing I do fixes it, but I want to exhaust other non-warranty-voiding options first. I appreciate your help! I'll be sure to manually set my RAM's timing, etc. Allowing the mobo to auto detect it was pretty lazy on my part, in hindsight!
 

raja@asus

Distinguished
Sep 28, 2011
891
2
19,360
1) Updated UEFI to latest version if not done so already.

2) Place a fan over the memory. Memory can be prone to producing errors if the module temp is too high. Up to 80C should technically be fine, but there are insatnces where DIMMs get too hot in some PC cases, especially when there is heat pouring out of other devices thus raising the ambient temps. Check other system temps under load as well and add cooling as required.

3) You can also try increasing CPU/NB voltage to about 1.20V (NOT the other Northbridge voltage) and see if that helps. Setting Read to Read Delay to 6 in the DRAM timing page and Write to Write Delay to 6.