Power Surge? Overheating? MP Shut Down!

Discussion in 'Mac Pro' started by Michael73, Nov 6, 2010.

  1. Michael73 macrumors 65816

    Joined:
    Feb 27, 2007
    #1
    Twice this morning out of the blue the entire machine has shut down. The first time my son had been playing some online games using Firefox. He'd been on for maybe an hour or two and he comes running in and tells me the entire machine shut down...not went to sleep, shut down. I waited a couple of minutes and turned the machine back on. It hadn't been running more than 30 or 45 minutes. This time my wife is on the machine and again, I think she was using FireFox checking Facebook. I was in the kitchen heating up some lunch when for a brief second the power to the microwave seemed to fluctuate. I heard my wife from my office say the computer had just shut down. Again, I waited for about 5 minutes and turned the machine back on. I keep using my iStat widget to see if the fans are excessively spinning or the temps are overheating but everything is normal. My first thought has been that the machine is shutting itself down to overheating.

    Any ideas on what's happening? :confused:

    The only commonality between the situations is that both people we using FireFox at the time but I've used it countless times and a complete shutdown has never happened to me.

    Should I call this one into AppleCare on Monday?
     
  2. nanofrog macrumors G4

    Joined:
    May 6, 2008
    #2
    Try resetting the PRAM, and if that doesn't work, repair permissions (i.e. tends to need done when the system isn't shut down properly, as would happen in a power outage) and see what happens (seems like a software problem to me).
     
  3. cutterman macrumors regular

    Joined:
    Apr 27, 2010
    #3
    Power fluctuating on the microwave too? Are you on a UPS? If not I would get one.

    Could be the AC power line.
     
  4. wordoflife macrumors 604

    wordoflife

    Joined:
    Jul 6, 2009
    #4
    You should have a UPS with such an expensive desktop. That should fix your problem too.
     
  5. westom macrumors regular

    Joined:
    Nov 8, 2009
    #5
    Why are you waiting 5 minutes? Does it startup immediately?

    If yes, well, all computers work fine even when incandescent bulbs dim to 40% intensity. Your voltage dropout would be so low and so long as to trip off the computer.

    So, why is power temporarily cut off? The question is found in household wiring or in utility power.

    For example, do incandescent bulbs dim or brighten when a major appliance power cycles? Bulbs must not change intensity; a symptom useful for finding a wiring error.

    Do not ignore these symptoms. Usually it is a trivial problem. In rare cases, this can be a serious human safety problem. Anyone recommending a UPS to cure these symptoms is irresponsible.

    Fix the problem. Because it is an intermittent, you must take care to identify all tiny symptoms. Unfortunately too many assume incandescent bulbs changing intensity is normal; ignore a wiring defect or weakness. That would be one relevant symptom.

    You saw a power fluctuation that coincided with a computer shutdown. That power fluctuation would not exist if wiring (or utility transformer) is operating properly.
     
  6. nanofrog macrumors G4

    Joined:
    May 6, 2008
    #6
    Not necessarily.

    If the PSU is bad for example, this can happen (think bad caps for instance, as the PSU can't take the startup load, and probably the nominal load as well, as it's not performing as it's supposed to due to an internal PSU fault, but not a short). For example, damage can occur to Active PFC units when on a stepped wave output UPS over time that can result in such behavior.

    Assuming the hardware in the system is 100% fine though, then I agree with you (by far likely that the power company's infrastructure is at fault, as most common dwelling problems would result in a tripped circuit; short due to worn insulation, nail through the insulation, bad breaker, ....; assumes the original wiring was done to code enforced at the time of construction/inspection, not something stupid like bypassing breakers with a wire).

    Lights dimming under high load (near, but not exceeding the circuit limit) during a startup isn't uncommon (but would be brief). Anything else needs to be dealt with by a licensed electrician (if the problem lies in the dwelling). Utility co. otherwise.

    Please understand, I get the sense you have sufficent knowledge, but it's not entirely clear, so this post is just for clarification.
     
  7. westom macrumors regular

    Joined:
    Nov 8, 2009
    #7
    How do bad capacitors in a power supply cause a power fluctuation on a microwave elsewhere in the house? It doesn't. Cited are the usual suspects that have no relationship to how electricity works or the OP’s symptoms.

    How does replacing a PRAM eliminate an abrupt power shutoff? It doesn't. If house has a wiring defect that causes that power fluctuation, then no circuit breakers trip. Obviously. But again, it means learning how electricity works and not just blaming the usual suspects.

    Most likely suspect for these symptoms is a loose wire. Often found in a junction box. Other less common problems could also cause it. But anything more is only wild speculation due to almost no information from the OP. For example, do a microwave and computer share a common neutral? One so many details necessary to say anything more. In most every case, a solution means getting an electrician to first identify and then fix the actual problem.

    But again, because it was so irresponsible. That UPS recommendation is a classic most naive curing symptom. Those symptoms suggest a house wiring problem - i.e. a loose wire. Fix the problem. That UPS says how many will cure the usual suspects rather than first learn technology.

    A bad computer power supply does not cause household power to fluctuate. This problem is most likely located in household wiring or the utility’s transformer.
     
  8. dknightd macrumors 6502

    Joined:
    Mar 7, 2004
    #8
    Sounds to me like it might be a street power problem. If it does not happen again I would not worry too much about it. It could be a local (in house) power problem which is worth getting fixed. Not likely overheating or other computer problem, though I suppose it might be. Has the OP posted back? probably a power company glitch that is already fixed. maybe.
    A UPS could be useful if the glitch is outside the house, but it gives you one more thing to maintain . . .
     
  9. nanofrog macrumors G4

    Joined:
    May 6, 2008
    #9
    I was going on the basis that the microwave + system shut off = once, not both instances, according to the first post. But the software commonality caught my attention (hence the mention of PRAM and permissions, as the microwave issue could have been coincedental, not another symptom of another common issue). That is, start simple, then work your way up, and both of those recommendations are quick and easy (beginning of a thorough diagnostic process that could end up requiring a licensed electrician).

    Now back to the microwavce, Yes, in that one instance, it seems there was a fluctuation that affected the entire dwelling. So suspecting the power from the utility co would be the first thing that makes sense. But there's no indication of a house-wide power fluctuation the second time the system failed, though possible, and it ignores the software commonality (hasn't yet been ruled out).

    So I took the approach that the house wiring may be fine (seems recent, so went with the start simple, then get more complex as things are ruled out). That doesn't mean some critical information's missing, such as the utility co. recently doing work in the area, or if there's been recent electrical work done to the dwelling. Now if the problem's been there for awhile, then a loose wire would be highly suspect. But again, I'm going by the OP's information this is all recent (both instances same day). This is why I'm thinking it's more likely the software or the utility company rather than the structural wiring in the home.

    I mentioned bad caps for the system's PSU (i.e. can't store their full value at working temp), and when the load gets high enough (such as startup, resuming from sleep, when the system requests HDD use, ....), the system will lose power and go down since it can't handle the load presented to it.

    As per the UPS, I agree that other possible sources need to be properly investigated first, as it could mask a serious problem (structural wiring), allowing a saftey issue to go unresolved that could cause a fire, potentially resulting in serious injury or death as well as significant property damage.
     
  10. westom macrumors regular

    Joined:
    Nov 8, 2009
    #10
    Utility power problems would be on multiple homes. A relevant fact would also include a neighborhood survey. Anyone else suffer same?
     
  11. nanofrog macrumors G4

    Joined:
    May 6, 2008
    #11
    Brown-outs are rather common, so you can't leap to conclusions that it's a wiring fault. But I gathered that your interpretation would have wiring fault first as the likely suspect (dwelling or utility in origin), brown-out second. If this was a wrong interpretation, I appologize.

    But thinking statistically, both brown-outs and software are highest on the list in general in my experience (more so than a structural wiring fault), and the software side is diagnosable by the user (takes time, but it's free if they do it themselves). Wiring would require a professional with the necessary skills and tools (which means the OP has to have the funds to pay for that person's time if the fault is located in their home).

    Unfortunately, I'm not accustomed to most people understanding what a brown-out is if you ask them (seems the OP as well from what was posted), so it would be hard to obtain the necessary information (ask the right questions) and interpret it from a neighborhood canvas IMO (not impossible mind you, but rather difficult).

    And I'd be totally amazed if there's a meter on-hand (friend, ...) with data logging capabilities to check.
     
  12. wonderspark macrumors 68030

    wonderspark

    Joined:
    Feb 4, 2010
    Location:
    Oregon
    #12
    I've lived in many different homes over the years, and never experienced wiring problems. I've even lived in homes that were moved several miles from a mine site to a small town in the boonies.

    On the other hand, I've experienced many brown and black-outs. Many. I can't remember the last time I owned a computer without a UPS. It's part of the computer, as far as I'm concerned. Computer, monitor, UPS, mouse, keyboard, printer.

    I know I'm lucky to have lived in areas where building codes are enforced, so maybe this discussion circles around places where all electrical connections are made with those twist-nut thingies like you'll find in ceiling fans... I don't know.
     
  13. westom macrumors regular

    Joined:
    Nov 8, 2009
    #13
    All electronics must work at a normal voltage which is an incandescent bulb at 50% intensity (also called a brownout). The light bulb 'browns'. And that is sufficient power for all electronics.

    Computers are required to be even more robust. Bulb must dim to 40% intensity. And every desktop computer must even start without problems.

    Now, that same low voltage is harmful to motorized appliances. If voltages remain below 5%, then harm is possible to refrigerators and the furnace. So utilities either maintain voltage that accurately. Or disconnect power.

    Low voltages that can be harmful to the dishwasher or washing machine are perfectly ideal power to all electronics.

    Every homeowner has a powerful tool to locate defective house wiring (or a much less common utility transformer problem). The incandescent bulb. If power cycling of any major appliance causes light bulbs to dim or (very serious) brighten, then the homeowner starts with suspect number one - loose wiring.

    Returning to the OP's problem. Insufficient information has been provided to recommend more. We know that a computer shutdown due to brownout means a supply problem. Such problems are not solved by a UPS (Most every cmoputer I have used for generations - even long before PCs existed - never needed a UPS and worked just fine).

    Any homeowner who needs assistance should consult an electrician (or interview neighbors to confirm that that neighborhood does not have a utility problem).

    All computers must startup even when incandescent bulbs are at 40% intensity. And must operate uninterrupted for a short time even when no power exists (which is why a UPS does not cause a computer shutdown when switching to batteries). The OP's descriptions sound very much like a loose wire. In most every case, a simple solution. In a rare case, also a human safety problem - which should have the OP's immediate attention.
     
  14. rtrt, Nov 7, 2010
    Last edited: Aug 13, 2011
  15. nanofrog macrumors G4

    Joined:
    May 6, 2008
    #15
    It depends on the design of the SMPS, and most of those opted by system vendors are based on lowest cost, which means short cuts were almost certainly made to get the pricing that low (cheaper or lower value parts tend to get used, as well as design compromises). You'd be amazed at how many do not meet spec.

    Many are using the CCFL's now to try and lower their utility bill, and may not have an incandescant on hand any longer (though easily solved).

    But a dimming isn't always an indicator of a loose wire. It depends on the load conditions (startup load can be significant for some devices; enough to cause a brief dimming an incandescent bulb, but not trip a properly functioning circuit - no loose wires, bad breakers, ...). Window AC's come to mind (those with say 10+ k BTU).

    In terms of a low voltage (say under 90VAC, which is common for the minimal wall value a computer PSU can operate), we agree.

    My god there was a difference in the PSU's used then and now (better designs, and especially parts selection).

    But a UPS can solve a brown out (low voltage condition) if the source of the issue is the utility power, not a wiring fault (more likely), as poor power conditions from the utility are rather common.

    That doesn't mean a wiring fault doesn't exist, but poor power is far more common.
     
  16. westom macrumors regular

    Joined:
    Nov 8, 2009
    #16
    Like the power supply in electronic appliances, these fluorescent type bulbs regulate voltage. Therefore no quantifiable change is created.

    Because the incandescent bulb can be such a powerful tool, you might find and store a few of these bulbs.
     
  17. westom macrumors regular

    Joined:
    Nov 8, 2009
    #17
    SMPS did as posted routeinly even before the IBM PC existed. That was when quality standards were so much lower.

    However, the computer industry has many computer assemblers without any electrical knowledge. Who is responsible for an SMPS and other parts meeting industry standards? Not the manufacturer. Meeting standards is completely the responsibility of that computer assembler. IOW power supply manufacturers can dump crap into the market when so many computer assemblers have insufficient training.

    Again, electronics routinely work just fine even when incandescent bulbs dim to 50% intensity. An industry design standard that was routine even 40 years ago. And is both easily and routinely implemented. In fact Tom MacIntyre in "Motheboard Problem? Post Problem?" described one TV that worked even down to 37 volts. Because it had an SMPS.

    When does a supply not meet those requirements? When the supply manufacturer all but admits that supply is defective. And a technically ignorant computer assembler only understands price and watts. Cannot read or comprehend even simplest specificatons.

    This is a Mac forum. Those supplies meet industry standards and say so in numbers.

    I never said dimming was always an indication of a loose wire. If A results in B, then B also results in A? Please do not use such defective logic to misrepresent what was posted. Meanwhile a loose wire can cause incandescent bulbs to change intensity when ... and I also said this ... when another appliance power cycles. That is a strongest indication of defective household wiring.

    Bottom line fact no matter how it gets twisted or respun. Electronics must work just fine even when light bulbs dim to 50% intensity. As has been true for generations. Light bulbs changing intensity due to appliance power cycling implies problems most often in household wires (rarely due to utility failures). An incandescent bulb is a powerful tool to identify such defects. Voltage changes that must not affect electronic appliance operations.

    It is a Mac. That means these brownouts do not cause problems. For power to cause a Mac to crash or to shutdown, a wiring problem would be so severe and completely unacceptable.
     
  18. JavaTheHut macrumors 6502

    JavaTheHut

    Joined:
    Aug 15, 2010
    #18
    As a matter of elimination get a circuit tester $10 to check for wiring config.
    you may have reversed wiring?
     

    Attached Files:

  19. westom macrumors regular

    Joined:
    Nov 8, 2009
    #19
    Reversed polarity does not and must not cause appliance failure. Polarity exists as another layer of human safety protection. Every appliance must work perfectly normal when white and black wires are reversed.
     
  20. rtrt, Nov 8, 2010
    Last edited: Aug 13, 2011
  21. JavaTheHut macrumors 6502

    JavaTheHut

    Joined:
    Aug 15, 2010
    #21
    Thanks for clearing that up
     
  22. pinchu71 macrumors newbie

    Joined:
    Dec 3, 2008
    #22
    Does your Computer shut off or the screen goes black?
    Check if the fan of the graphic card is running properly, mine is defective and dont spin up, after 5 min computer screen goes black.
     

Share This Page