Faulty psu?, combined amps on 12v rails?, and future proofing my rig?

Hi all, first post to tomshardware, but have ready many articles on this site. Love this website.

Although I’ve done a lot with computer hardware in the past, this is the first time I’ve built my own system and had some questions/problems. I’d appreciate everyone's help.

System specs:

CPU: Intel i7-3770 (not the k version), using the intel stock cooler
RAM: 8gb corsair vengeance 1600mhz ram (2 x 4) (not overclocked)
MB: Intel DZ77BH-55K motherboard (with two pci express 3.0 slots available)
Disk1: Corsair force gt 90gb SSD (this is the boot drive)
Disk2: Western digital 500gb sata 7200 rpm drive (for downloads, movies, extra storage)
Disk3: Asus 24x dvd burner
Video: EVGA Nvidia GTX 560 SC with 1gb ddr5 (slightly factory overclocked out of the box), the specs says it requires 450W PSU and 24 amps on the 12 volt rail, and requires 2 pci-e 6pin connectors.
Fans: Four 120mm SILENX fans , two intake, two exhaust
Case: Coolermaster elite 430 case
OS: Windows 7 64 bit ultimate
Other: Integrated sound and Ethernet from the mobo
Various usb peripherals (wireless mouse, keyboard, headphones)

PSU that I replaced: Coolermaster Extreme Power Plus 500W, 12 volt #1 rail 1 has 18 amps, 12 volt rail #2 has 18 amps, and there is 1 pci-e 6pin connector. (it says the combined 12v rails is 360 watts)

PSU that I have now: OCZ ModXStream Pro 700W, 12 volt #1 rail has 25 amps, 12 volt rail #2 has 25 amps, and there are two pci-e 6 pin connectors. (says its sli ready, and is 80+, and combined 12v rails is 552 watts)

Here’s the story. I built this gaming rig (which I love) after saving my pennies for quite a while and taking advantage of some deals through friends. It’s been running for about one month. All was running well for the whole month, until last week. I started to experience random reboots and shutdowns. In the span of 3 days it happened many times on day 1, none at all on day 2, and many times on day 3. Usually the system was not under load. It was usually just after startup, but sometimes later. Not usually when I was playing a game (only once when I was playing diablo 3).

This made me nervous, so I started monitoring and reviewing everything. No noticeable heat issues with the cpu/video card (using core temp and gpu-z). everything was nice and cool. In my bios I noticed something strange. The intel visual bios is pretty great and it was showing me real time monitoring of all the voltages on the power supply (the coolermaster psu shown above). The 5v, 1.5v and 3.3v all looked fine, little to no fluctuations and all within +/- 5% of the spec. But the 12 volt was awfully funny. Sometimes it was 11.7, other times 11.4, then sometimes 10.7 and even 10.4. It was fluctuating a lot, and way beyond 5%.

So I ran out and quickly bought another power supply. After only a very short while to research a new power supply, I found a good deal on the OCS ModXstream 700 shown above. It’s only been a few days, but at last check, it all seemed to be working fine again. Hopefully it will continue to do so. The new PSU has been rock solid in the bios on the 12V, always at 12.41, no fluctuations that I have seen. So, here are my questions:

- Does it seem reasonable that the random reboots and shutdowns (usually while mostly idling but sometimes not) and the bios indicating that the 12v seemed to be off (but not the 3.3, 5, etc) indicate the power supply was faulty? Do you think that was it and I am in the clear with a new PSU?

- The old power supply was in the computer for about a month, but only did the weirdness over the course of the last three days before it removed it, a couple of hours each day while I was testing stuff. Do you think I may have damaged the other hardware because of this? No visible smoke, burns, damage, or smell, and the computer seems ok so far with the new power supply, but still kind of worried.

- Later on I realized that the old psu had two 12v rails, both at 18 amps each. The EVGA GTX 560 superclocked said a minimum power supply of 450 watts and 24 amps on the 12v rail is required. Did I hurt the video card or the psu because it recommended 24 amps, but I have two rails at 18 each? I thought the rail amps strength is supposed to be combined, but now I am not so sure. The whole psu was 500 watts, but the side of the psu said the combined 12v rails is 360 watts. The psu had one pci-e 6 pin connector, which I connected to the 560. The 560 needs two 6 pin connects, so I used the molex adapter that came with the video card for the second one. Did I somehow unde rpower the gtx 560 and cause all this myself?

- Finally, and the most important question of all, now I have this nice new OCZ 700W power supply, it has two 12V rails, of 25 amps each. It also has two pci-e 6 pin connectors, which I used both on the gtx 560. So, I imagine the 560 is getting enough juice now. And a few years from now, if I happen to find a gtx 680 cheap (ha ha) and wanted to go with that, could I?, using this psu. The gtx 680 needs 38 amps on the 12 volt rail, minimum of 550 watt psu. I have two 12V rails of 25 each, for a combined strength of 552 watts (out of the 700 watt psu). Is that enough? Do the two 25 amp rails really combine (it says combined watts is 552 on the psu unit), or should I return the unit to get one with a single rail with more then 38 amps? I don’t want to, as this seems like a pretty nice psu and would like to keep it, but not if its not going to be future proof. I know the concept of these separate rails is really just for system protection, but I don’t want to have to buy yet another power supply if I don’t have to.

- Ok, I lied, one more question. The other alternative is that I may someday find another gtx 560 cheap, and drop a second in there. Could I do that with this psu? The psu says it is sli ready, and each rail is 25 amps, but I would then need 4 pci-e 6 pin connectors, so I would have to use two molex 6 pin adapters.

Mostly I’m just worried that I diagnosed everything correctly, that my system is healthy now, and that I can use this new psu for future, beefier video cards with it and not freak out that I am going to kill/short something out.

Thanks for all of your time and patience on this ridiculously long post.

3 answers Last reply Best Answer
More about faulty combined amps rails future proofing
  1. Best answer
    The Cooler Master eXtreme Power Plus really is a very bad PSU, so it would not surprise me if it was the cause of your problems.

    No you do not need to replace your OCZ PSU due to it having multiple +12V rails, unless you are serious about being able to SLI in the future (see below).

    Yes you will be able to install a GTX 680, which isn't that much more power hungry than a GTX 560.

    You might be able to install two GTX 560s on a MSXP700, but I wouldn't recommend it because not having enough PCIe connectors. Also because both graphics cards would be on +12V2, when using adapters, which on load would probably overload a 25A +12V rail.
    Multiple +12V rail units are capable of supporting multi card setups, it's just that the OCZ has a fairly bad rail layout and is also short of +12V rail capacity in general considering that it is supposed to be a 700W PSU, most people look at ~700W PSUs and think they should have no problem SLI/CFing.
  2. Best answer selected by jjcttx.
  3. Thank you! That's exactly the info I was looking for. Now at least I will feel good that I can upgrade to a better card in the future, like the 670, 680, etc.

    As for the possibility of SLI, its kind of a longshot for me really, as I would have to find another identical EVGA 560 SC (superclocked edition) for a very low price, and at that point better cards like the 570, 580, 670, etc may be more reasonable too.

    It's not impossible that I may try it someday, but i'm not going to worry about it at the moment. Thanks again for your help!

Ask a new question

Read More

Power Supplies Components