Memory issues and optical drive

theta

Distinguished
Oct 4, 2009
37
0
18,530
I seem to be having some odd memory problems. The thing is, these only seem to start when I am using the optical drive.

I got a new motherboard (GA-P55-USB3) and RMAed memory (G.Skill Ripjaws) about a week ago. Everything seemed fine, none of my previous cold boot issues happened. However, I got a blue screen when I inserted the driver CD. I chalked it up to Windows weirdness, because when I rebooted I was able to install all the drivers fine. And everything worked fine for about a week.

Now, a day or so ago, I went to pop in a live cd of Gentoo: It wouldn't boot, citing memory issues. ("NOT AGAIN," I groaned, remembering six months of cold boot issues prior). Ran memtest86 and sure enough, I was seeing red. So I decided to see if one stick was bad, or what:

Popped out one stick from the 2nd DDR3 slot, memtest86: Pass.
Swapped out the sticks: Memtest86 passes again
I try both of them again, and it passes.

To me, it seems like the memory is working. Windows loads up fine, I can boot into the live cd without error. Until I decide to run GParted off another CD: more errors. I repeat the process, this time only bothering to boot once with one stick, then booting with both sticks the next time. From what I can see, the arrangement of sticks doesn't matter. Additionally, it seems that I can reproduce these memory errors just by using the optical drive. The errors don't seem to always occur right away (but sometimes they do).

The optical drive seems to function fine, insofar as reading my discs. Also to be noted that, on my old GA-P55-UD3R, I had no issues with it (granted, I was tearing my hair out trying to figure out my cold boot issues).

I've updated the BIOS, set it to fail-safes, set it to optimized, set it to what G.Skill recommended for my RAM, but it doesn't seem to solve anything. I've also switched SATA ports and I will report (probably very soon) if it still fails--probably will, but who knows.

Anything else I could try? PSU is an antec earthwatts which I place high faith in... Other hardware seems operable without these memory issues. If I had to guess I would say I got a bad motherboard (and I would then curse several varieties of gods for the luck). Anything else that I could test?
 
How long were you letting MemTest86 run:

Stick 1, solo?

Stick 2, solo?

Sticks 1&2 together?

The concern I have here is that MemTest86 actually reported errors. I am a little reluctant myself to blame your optical drives, as they ran MT86 no problem.

Have you tried running Prime95, a stability testing software? If not, try it. P95 is free and when you run the blend test, it will test your RAM and put 100% load on your CPU. Additionally, download and run SpeedFan and/or CPU-Z to monitor your temps while P95 is running.

Lastly, though you have faith in your PSU - and Antec is a very good brand - don't rule it out just yet. The Earthwatts line is designed to be more energy efficient, thus it doesn't put out like non-"green" PSU's.
 
How is the Voltage to the Ram and is it the correct amount for the new modules ?? does it fluctuate when you use the drive ?? Could be a problem with the PSU if the added use of the drive is making the DDR voltage fluctuate leading to the errors. Perhaps try increasing the voltage of the RAM one notch and see if it stops the errors.
 

theta

Distinguished
Oct 4, 2009
37
0
18,530


Granted I didn't do the usually day-long memtest in this instance, when this bug occurs--and for now I'm calling it a bug since I've been running my system (playing games and doing intensive activity) for hours--it takes a second or so for the errors to occur. So leaving it in for 10-20 minutes generally assures me that the errors have resolved. When I first got this RAM I did about a 10 hour memtest overnight, with no errors.

I'll try Prime95. My temperatures are a bit high, despite my applying AS5 with a new heatsink (Freezer 7 Pro)--50C on normal operation. Though note that I'm in Florida and my room, while cool, is not freezing.



I'm not sure what you mean. My BIOS gives me voltage readings (I can't remember what they are at the moment, I'll report that soon), but I'm not sure how I would tell if they fluctuate when using the drive. As far voltage goes, I'm using what is specified here: http://gskill.us/forum/showthread.php?t=2206 I did seem to remember some of the voltages being a bit off, I'll check again.

Here is what my BIOS says about voltage:

MB Intelligent Tweaker:

Vcore: 1.264V
DRAM Voltage: 1.536V

MB Intelligent Tweaker > Advanced Voltage Setting:
Load-line calibration: Disabled
CPU Vcore 1.30000V [1.30000V]
QPI/Vtt Voltage 1.100V [1.190V]
PCH Core 1.050V [Auto]
CPU PLL 1.800V [1.900V]
DRAM Voltage 1.500V [1.500V]
DRAM Termination, Ch-A Data VRef., Ch-B Data VRef., Ch-A Address Ref., Ch-B Address VRef. 0.750V [Auto] (all of these values are the same)

PC Health Status:
Vcore 1.284V
DDR15V 1.536V
+5V 4.945V
+12V 11.985V


First value, I think, being what the current value is, the second brackets being the one that I have chosen. Seems a bit lower than my choices, not sure what to make of it.
 

bilbat

Splendid
Out of curiosity, what brand is it? Someone, a couple days back, reported 'fatal difficulties' with a Plextor - which surprised the hell out of me, as I used to buy Plextors, exclusively, back in the days of SCSI chains - they made the best in the business!
 

bilbat

Splendid
Hmmph! That should work - I've gone through, between those and the LightScribe version, about three quarter dozen in the last ten months! Got another question - how are you powering it? Molex to SATA adapter, or does your PSU provide direct SATA plug in? Reason I ask, is I've a tidy pile of defective MOLEX adapters. Haven't run into one that would 'down' a machine, but am generally suspicous of the lot! Any time I need to use one on a system for anyone, I usually order three - in the hopes I get at least one good one! Worst part - most common defect is that they work in some position; let 'em 'relax' for an hour, or re-stress the cabling by adding something else - bye bye drive!
 

theta

Distinguished
Oct 4, 2009
37
0
18,530
No, they are directly SATA. However, it occurs to me that I probably didn't use the same power cord (that is, the same series of wires protuding from my PSU) for the optical drive as opposed to what I used with the other motherboard.

I will swap the two SATA power cords, to see if perhaps one is defective.
 

theta

Distinguished
Oct 4, 2009
37
0
18,530
Well, thanks for the help but I think this is going to be my course of action: Seeing that I have a spare (defective) mobo due for an RMA, I'll RMA it and swap it out to see if the problem persists, and if it doesn't I'll use the RMA'd motherboard while I RMA the newer one. If it does persist, I'll be buying a new PSU (I want a better one anyways, something modular), and possibly replacing the optical drive.

Since there is no problem when I don't use the optical drive, and I rarely use it as is, I don't mind taking a few months to troubleshoot. School will start anyways, and I can't be worrying about this.

Thanks.