• July 21, 2015, 2:25 a.m.

    I've noticed, or maybe just with my computer shitting the bed it makes me question more, that of late (for a good while) I've had odd fan noises: the fans whisper down, speed up, make weird click/gnarly sounds, and it's never the same fans.

    All of the fans are working, however. So is that potentially a sign of a failing power supply? Power supply failure might explain why when the comp was crashing it was generating no mini.dmp file either at the end, if this is caused by a progressively weakening power supply.

  • Retired 124 posts
    July 21, 2015, 10:24 a.m.

    No boots, random reboots, blue screens, loud & weird noises, smells and sparks are a few symptoms of a dying PSU. But a lot can be similar with other hardware failure. Hardware or software is probably the first thing to figure out. Have you recently upgraded any part of it? The excess draw could cause it to give up. And are you making the switch to Windows 10 anyway when it launches at the end of this month? I am going to it and it rules the software out anyway if you are making the change.

    I had a PSU that was maxed out in my old build that would shutdown when going to full load, if it didn't it would be fine all night. But the first big pull when a game fires up would either shut me down or be fine. Happened when I logged into TSW, I was fine until the game first came on. I would usually get a hello out to the Cabal and then go offline lol. If it is up to the PSU's max you could pull the GPU and use onboard if applicable and see if the power drop helps things out. Memory failure can also cause random problems, if using dual sticks you could possibly pull one at a time.

    It could be the HDD failing too on the hardware side. Without bits to swap out its hard to solve the problem really.

    I guess the cheapest option is a full re-install, sometimes welcome anyway for general speed up. But a pain in the butt re-installing and setting up all accounts etc again. I bought a decent Antec PSU this time, a part of the system you really shouldn't cheap on but something I am guilty of often. Really as well it is nearly the only part that you move over to a new system build as well.

  • Retired 124 posts
    July 21, 2015, 10:32 a.m.

    The fan speed changing could be a setting but unlikely on case fans, which could be a sign the 12v is fluctuating which in turn may be causing your GPU to throw a wobbler.

  • Retired 56 posts
    July 21, 2015, 12:55 p.m.

    I also think it can be the PSU, it may be time for a new one...
    You also need to calculate the WATT usage vs PSU output...

  • July 21, 2015, 4:11 p.m.

    @Knoll - oh, yeah, didn't post all that's happening- driver power state failure was forst bsod, no mini.dmp files generated anymore. Changed some drivers, stable, bsod. Rcpss.dll potentially corrupted, replaced with valid one, crash, every time it brought back old version (potential rootkit).

    Win 8 clean I stall, stable for a bit but 100mb win 7 partition 0, crash again, format partions 0 & 1, still crash.

    Memory diagnostic, malwarebytes, mb anti rootkit (is what found rpcss.dll), Microsoft security essentials run, mbar only thing that found anything. Tdsskiller found nothing.

    8 passes memory diagnostic, no errors.

  • July 21, 2015, 4:16 p.m.

    Sys specs:
    I7-950 stock 3.06ghz, very lightly undervolted for 4 years, stable until now (stock hsu was hitting 80c, so I dropped it very very slightly)
    Hyper212+ hsu
    Asus sabertooth x58
    1tb wd hdd
    12 gb ram, tri-channel, 6 2gb sticks.
    Msi 660ti 2gb fac oc
    Corsair 750watt silver rated.

    All components but 660ti: 4 years old.

    Real temp runs 24/7 under load +ambient highest is 73c

    Crashes loading in, no discernable pattern. Many event viewer critical events, no mini.dmp or BSOD even with autoreboot turned off.

  • July 21, 2015, 4:19 p.m.

    Oh, last thing: bios two years out of date (2010 version, new bios is tagged 2012), with insecurity and instability refuse to flash bios at this time, reasonably enough.

  • Retired 124 posts
    July 21, 2015, 5:20 p.m.

    I wouldn't think you would have to under volt with the Hyper on there, I use that too and my i5-2500k is running around 4.3 ghz OC at not much more than 60 degrees in Prime, I think I fell lucky with the "golden child" of chips though. I took it to 5ghz comfortably on air but played safe. That and the MB is all I didn't upgrade entirely due to OS licenses. But I don't think it should be going above 75, I am personally happy in the 60's but low. I wouldn't expect an i7-950 to be any higher than 65 degrees stock with the Hyper. But you can fall unlucky with a bad one I guess.

    Did your undervolt survive a decent length Prime or Intel Burn test? I guess all electronic devices lose efficiency and will draw more as they age. It would potentially be my first port of call but I wouldn't be ruling out the PSU either. I would return it to stock or marginally increase it especially if you haven't tested it since installing the Hyper. I intended on going water in my build and used the Hyper as a budget option to get running, I have been nothing but impressed. I put another fan on as a push pull setup. I got a Corsair Air 540 this time as a case, the thing is huge/quiet and cable management like I have never seen. Your ambient temps are high but then I live in Yorkshire in England lol, we see no heat only rain clouds.

    I wouldn't expect a Corsair PSU to fail in such a short time, but then all electronics can. Its probable "actual" output won't be so bad either. There are a lot of memory sticks that could go wayward there though, I haven't looked into how "final" a memtest could be though. You certainly have enough to sacrifice some though if all else fails.

    You can probably rule out software then with a fresh install giving issues.

  • Retired 88 posts
    July 21, 2015, 10:43 p.m.

    CPU temp? That's a bit high imo...

    I wouldn't bother much with flashing the bios btw. Unless there's something specific you really need from the new firmware. I never upgrade my motherboard's firmware unless I have a reason to.

  • July 22, 2015, 3:54 p.m.

    Can't, it doesn't generate BSODs or mini.dmp files anymore. Even with it set in system advanced to show me this. It just crashes and restarts, one of the reasons I think PSU.

    @Kemo- depends on ambient, I'm on third floor in summer, so under load 73c is max I've seen. Lower ambient 60~65c is normal

  • Retired 88 posts
    July 22, 2015, 6:27 p.m.

    I'm in attic and my room temperature in summer is literally hell... sometimes it gets to 40-45ºC. My CPU temperature is about 10-15ºC when idle and 40-50ºC under load. It can get to 70ºC when running specific programs designed to stress the CPU, but never under normal load.

  • Retired 124 posts
    July 22, 2015, 7 p.m.

    I hope you get it all sorted mate, it frustrates the hell out of me and I have several computers in the house to swap parts about. With the latest build I had no display what so ever and suspected the new GPU was bad. Swapped out part after part trying to narrow it down only to discover, the idiot that I am, forget to plug in the 8 pin connector from the PSU to the CPU. Wasted hours..

    And now I have wasted money too. I had a 3x21" monitor setup from the last computer, but then opted for a wall mounted arm arrangement to clean the build up a bit instead of 3 stands. Didn't like it at all, came too far forward. That is now in the loft. Then I decided I wanted a larger 27" monitor above the 3 to use for gaming and video stuff. I have installed that today and now the address bar is over 45 degree high on looking angle. It isn't comfortable. So 2 more stands are now ordered and I am after copying number 12 on the link below, as long as the overall heights match I think it looks ok and functional! I think that is a good compromise. I don't really game in Eyefinity but I couldn't be without multiple screens for all other duties.

    twistedsifter.com/2012/06/amazing-computer-station-multi-monitor-setups/

  • July 23, 2015, 2:26 a.m.

    YESSSSSS

    Bios:

    CPU 1.248v
    3.3v Rail 2.768v
    5v Rail 6.008v
    12v rail 14.048v

    It looks like PSU is shitting bed, shall find out tomorrow if that's true, if I crash after installing a new, overkill (ie, looking into an 850gold) PSU it's my mobo dying.

  • Retired 56 posts
    July 23, 2015, 12:52 p.m.

    I have a CORSAIR RM850 (850W), it is awesome !!!