Super User is a question and answer site for computer enthusiasts and power users. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

When I first built this machine in mid-Feb, I had some problems with stability, which I traced to a particular stick of faulty RAM (failed memory tests, wouldn't boot if it was the only stick). Of course, I replaced the stick.

After getting everything stable, things were fine, until about 2 months later.

Now, about every 6-8 weeks, the system will get unstable; it'll crash about once per day at first, finally getting to the point where it's going down almost constantly. Windows BugCheck codes are almost exclusively related to memory errors (or occasional NTFS errors which can be doubly-caused by memory errors). Using memory diagnostic tools results in errors being detected with the memory.

So far, the "solution" is to, when it gets unstable, take the following steps:

  1. Power down the machine.
  2. Remove case cover.
  3. Remove all sticks of RAM; shuffle them (in the static bag) so that they are not replaced in the same order as they were removed. Not sure if this is vital, just what I've always done.
  4. Thoroughly dust, with canned air, all the components, falling just short of removing the HSF.
  5. Replace the RAM.
  6. Reboot the machine.

After doing this, I'll be fine for another 6-8 weeks and memory diagnostics will show No Errors.

Things I have not tested

  • Just powering down the machine for 30 minutes or so (machine is usually slept and not reset).
  • Dusting without removing the RAM DIMMs

I am fairly competent technically (programmer by day, hobbyist by night), but so far this stumps me. Does anyone know if it is the dust, the extended on-time, or something (seemingly) trivial I might have missed somewhere during construction? Parts list follows:

Most Relevant Parts:

  • Asus M3A79-T DX Motherboard
  • AMD Phenom II X4 940 Deneb (Stock 3.0GHz, OC'd to 3.5 via 17.5x200 [stock FSB], although the OC has been ruled out as a cause; it was doing it months before I overclocked)
  • 8GB (4x2048MB Corsair Dominator PC8500/DDR2-1066)
  • CoolerMaster HAF 932 (included because the case is basically a giant mesh cage)

Additional Parts (Just In Case):

  • 2x Sapphire ATI Radeon 4870 1GB, PCIe x16 Dual CrossfireX
  • Corsair TX850 850 power supply with single +12v rail at 70 amps
  • Thermaltake SpinQ Aluminum HSF
  • 2x Afermarket 120mm Thermaltake Case fans, cause there was room.

Any and all thoughts are appreciated!

share|improve this question
up vote 2 down vote accepted

I had a similar issue that turned out to be heat expansion/contraction. Heating and cooling of electrical components cause them move by minute amounts, which over time could 'loosen' them enough that the contacts start failing.

I'm not sure if humidity plays a part too, in that case the contacts could also be a little corroded, adding to the bad contacting.

Extended on-time shouldn't be a problem, how often do you reboot the system? If at least once a week, then I'll go on a limb and say I wouldn't blame the OS.

Try bring the temp inside your case down, you already have 2 fans, can you fit more or improve airflow through the case? Minimizing the min/max temperatures when the PC is off/on should reduce the chance of heat shift.

I can't guarantee this is your issue, but I experienced this before, living in a hot + humid area at the time, without A/C.

share|improve this answer
I have noticed that there sometimes seems to be room to press the DIMMs in slightly further now and again, though I can't recall if that's before or after I remove them. However, the HAF 932 comes with 3x230mm fans (front and side intake, top exhaust) and 1 140mm rear exhaust; I added a 120mm bottom intake (to take advantage of its generous ground clearance) and a 120mm top exhaust, since there was room. Moreover, all cables are kept behind the right case wall, so I don't believe it to be an airflow issue. But, it might be heat, this is FL and the hottest room in the house. Will check. – Dereleased Nov 11 '09 at 15:46
I wonder if some main boards are just more susceptible to temperature changes. Could it be that your memory module clips are worn or don't latch as they should? – invert Nov 12 '09 at 13:15
I'm not sure, maybe my screws are also in too tight, causing the motherboard to flex a little bit. I didn't do my usual ritual of dusting and whatnot and just turning the thing off and pressing components in did the trick. – Dereleased Nov 13 '09 at 18:34

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .