Logic board death, and how to prevent it

Discussion in 'iMac' started by Argali, Jun 5, 2012.

  1. Argali macrumors newbie

    Joined:
    Jun 5, 2012
    #1
    So I took my 2006 iMac in to the Geniuses at the Apple Store yesterday after suffering some serious Grey Screen of Death, and as I feared, the prognosis was fatal: fried logic board. This is also what felled my old G4 tower.

    I seem to get about 5-6 years life from my Macs. Right now I'm in a holding pattern using my wife's MacBook and my PC laptop for work, waiting for the new Macs to be unveiled this month (hopefully). My hunch is that I'll go for another iMac, depending on how the prices and specs fall.

    Anyway, in mourning the death of a computer that served me very well for five years, I got to thinking about the issue of logic board malfunctions and what can be done to avoid them. Genius at the Apple Store talked about the importance of turning the computer off vs. putting it to Sleep, saying the idea that powering a hard drive on and off taxes your system is a myth. I Googled that and of course opinions on this are sharply divided. I also got to thinking if I need to invest in a better surge protector, or possible even a power conditioner, like the Furmans I see used by musicians.

    Any thoughts on all this? Or is the general feeling that logic boards on most Macs are destined to fail in computers that are used every day?
     
  2. kitsunestudios macrumors regular

    Joined:
    Apr 10, 2012
    #2
    I've only had the logic board/motherboard go on one of my systems (early 2004 ibook died summer of 2011) although all of them have run into hardware failure issues of various severity at some point.
     
  3. zackkmac macrumors 6502a

    Joined:
    Jul 7, 2008
    Location:
    Tulsa
    #3
    I had a 15" MacBook Pro (early 2011 model) and used it for just about a year, and then the logic board went downhill. AppleCare ended up replacing the board, display, AirPort card and cable when I sent it in for repairs. Yet here I am typing on an iMac G4 that seems to have no issues despite being 9 years old. So I guess it just varies.

    I suspect that my MBP's logic board failed because of the generic-brand RAM I installed, because it only showed symptoms after that. Later on I read that the 13" and 15" models only support 1.35-volt RAM while the 17" supports 1.35 and 1.5-volt, and the RAM I had installed was indeed 1.5-volt itself.

    But, enjoy whatever you end up purchasing. :) My suggestion is to thoroughly check the small details if you plan on upgrading things such as the RAM, because I think the fact that I did not do that when I upgraded my RAM is what caused my MBP to die at less than a year old.
     
  4. yadmonkey, Jun 5, 2012
    Last edited: Jun 5, 2012

    yadmonkey macrumors 65816

    yadmonkey

    Joined:
    Aug 13, 2002
    Location:
    Western Spiral
    #4
    After years of Mac consulting, I think a major factor is dirty power. Brownouts and voltage dips/spikes are hard on circuit boards.

    If you see the lights dip when the garbage disposal comes on or the washing machine starts the spin cycle, I'm talking about you. Apartments, townhouses, urban homes, and older homes tend to have some degree of dirty power. Even new homes do sometimes.

    There are inexpensive battery backup units that offer some degree of AVR (automatic voltage regulation). Cyberpower Systems offers some good ones.

    I question that genius' advice. The initial surge of power into a circuit board is also taxing. Look at it this way - how many times have you seen a lightbulb spontaneously stop emitting light versus how many times they burn out when you first turn them on? I've never seen the former. I think overall Macs that sleep live longer than Macs that are powered down every night, but obviously my evidence is more anecdotal than scientific.
     
  5. leman macrumors 604

    Joined:
    Oct 14, 2008
    #5
    It is actually a rather rare thing to have a failed mainboard. The most likely component to fail in a computer is either the HDD or the PSU. The computer won't die just because you use it regularly. Of course, if your power network is bad, this may tax the components. At our department we have lots of macs (some of them as old as 2002) and I have yet to experience a single mainboard failure. That said, why would you want to own a computer longer than 6 years? After 6 years, its long obsolete.

    This is very true.
     
  6. Macman756 macrumors 6502a

    Macman756

    Joined:
    Sep 24, 2008
    Location:
    Atlanta, GA
    #6
    Is yours a 24" model from 2006?
     
  7. TyroneShoes2, Jun 5, 2012
    Last edited: Jun 5, 2012

    TyroneShoes2 macrumors regular

    Joined:
    Aug 17, 2011
    #7
    That may be anecdotal, but there is a point where enough anecdotal evidence combined stands on its own and matches the credibility of empirical evidence. I think your theory may almost be at that point. I have decades of anecdotal experiences that support this also.

    For instance we once had a TV control room full of CRTs of various ages going back as far as 20 years or more, all running 24/7 on an enterprise-level giant UPS/generator, meaning many of them had not ever once been turned off or experienced inconsistent power in all of that time. When we decommissioned the room, and removed all the normally-working monitors, about 2/3rds of them would not power back up, at all. Keeping them powered up kept them working; powering them down just once, after all that time, was a death blow to most of them. That is not uncommon.

    But it is not "dirty" power, it is inconsistent power that is problematic. A good UPS is probably the best defense for a desktop computer. Just not dropping it is the best defense for a laptop.

    "Dirty" power implies power which may or may not be inconsistent in voltage level, but which may have other frequencies impressed upon the 60 Hz AC. This is typically not a problem because all AC becomes DC before it is used in a computer, which means there are no variations in the level, and what was "dirty" about the incoming AC is now a moot point, as far as the DC circuitry of the LB is concerned, at least.

    But inconsistent AC power can actually create dirty DC power, if the level falls below spec, and then the filter caps in the power supply can't keep the AC ripple out of the DC. That's not a good thing, but it isn't all that problematic. CPUs abhor this and the OS will probably lock up or crash if this happens, but it is usually of no real threat to the logic board/MB.

    What is dangerous about inconsistent power is the spikes, the overvoltages. And they are typically also filtered out by the voltage regulator in the PS, but if too high or too sudden, that can make the PS fail, and rarely the LB/MB. That is why a UPS is a good idea, not because it also filters out crap at other frequencies (the "dirty" stuff) which would be filtered out naturally by the DC PS anyway.

    Logic board failures are rarely caused by dirty power (since it really can't get beyond the DC power supply) or even by inconsistent power (since the PS regulates the DC level if over voltage, and under voltage is equivalent to running that light bulb with LESS power as well, a situation that actually greatly EXTENDS its lifespan).

    LB failures are mostly either due to deteriorization of cold solder joints, or physical failure of traces on the board, due to continued flexing, heat, a catastrphic physical blow, or just age. A LB failure is then very similar to a HDD failure; it is not so much a matter of if but of when, and with luck, "when" will be after it is no longer in use, or will be in the infant mortality period covered by the basic warranty (which is why we have them).

    So if that informs the advice, it would be take care of your computer, avoid heat and banging it around, and don't worry beyond that about the MB/LB, because you otherwise honestly have little control over it. Everything returns to dust at some point; entropy is not avoidable forever. Its like the 96 year-old guy said when confronted with a question of mortality about the welfare of his 24 year-old wife: "if she dies, she dies".
     
  8. Argali thread starter macrumors newbie

    Joined:
    Jun 5, 2012
    #8
    I believe so. I'll check tomorrow.
     
  9. MacTech68 macrumors 68000

    MacTech68

    Joined:
    Mar 16, 2008
    Location:
    Australia, Perth
    #9
    This is probably the best example of what normally causes failures.

    So I proffer the following:

    Excessive heat due to insufficient airflow. Causes can be physical blockage like placing the computer in an 'alcove' type desk or a build up of dust throughout the airflow path inside the computer or it's air intake.

    I love the example of the CRTs failing after being powered down once after years of use. Moreover, it is often the constant cycling of temperature extremes that cause faults. Think of the thermoelastic strain theory of geological exfoliation.

    The iMacs of recent years are a nightmare of engineering. Removing the heat from the various sources in these machines seems to prove difficult. The graphics processors in these machines seem to be the weak point, and most versions do not feature a removable video card, so once the video IC begins to physically detach from it's soldered ball grid array the entire logic board requires replacement.

    Added to this is the change to lead free solder, due to health and environmental concerns. The long term strength of lead free solder seems to be much weaker than the older lead containing solders. This combined with BGA seems to be the cause of iBook G3, Xbox 360 and other BGA failures.

    So, for my 2006 iMac, I keep the fans running slightly higher than Apple deems "safe". :)
     
  10. Macman756 macrumors 6502a

    Macman756

    Joined:
    Sep 24, 2008
    Location:
    Atlanta, GA
    #10
    If it is a 24" 2006 either 2.16Ghz or 2.33Ghz please PM me! Thanks
     
  11. Argali thread starter macrumors newbie

    Joined:
    Jun 5, 2012
    #11
    I checked the serial number. Mine is a 17" 2 Ghz dual-core, made October 2006. Has 3 GB RAM now.
     
  12. maflynn Moderator

    maflynn

    Staff Member

    Joined:
    May 3, 2009
    Location:
    Boston
    #12
    I have a 2001 G4 cube that still going strong, heck even the original HD works. My point is there's no way to extend the life (short of keeping the logic board clean of dust and dirt)
     
  13. Argali thread starter macrumors newbie

    Joined:
    Jun 5, 2012
    #13
    I do think "dirty power" was a big factor in the death of my G4. Back then I lived in a condo that experienced various power problems. Things seem more stable where I live now, but it's clear that a good AVR is a worthy investment. And I'll admit, I definitely could have done more to keep my iMac vents dust-free. So I'll keep that in mind with my next one...hell I'll even schedule it inot my calendar to dust it every two weeks! :)
     
  14. Spike88 macrumors 6502a

    Joined:
    Jan 25, 2010
    #14
    I do the same for my 2010 iMac as well. re: Having fans spin 200-300 RPMs faster (which is a minor speed increase) keeps its insides 5C-8C cooler. Cooler under heavy loads = good thing in the long run…..

    .
     
  15. Melbourne Park, Jun 6, 2012
    Last edited: Jun 6, 2012

    Melbourne Park macrumors regular

    Joined:
    Mar 5, 2012
    #15
    Also check the underneath of the machine for things blocking the air inlets, and use a vacuum cleaner at the exhaust to try to suck out air at a faster flow-through rate.

    The pity of it is - that Powermacs are so easy to keep clean (although the power supplies in the ones I have looked at, seem to hide their dirt, you have to get to them, although I am not sure about the recent ones though )... But iMacs have not been designed for removing dust. Which IMO is somewhat outrageous.

    I wonder though, how much Apple would charge to flush out the dust. Probably nothing - because they do say they'll inspect an issue for free, and then quote on it. So ... perhaps under that guise, you'd get a free clean.

    Just don't drop the heavy iMac getting it there though !!!!!
     
  16. Intell macrumors P6

    Intell

    Joined:
    Jan 24, 2010
    Location:
    Inside
    #16
    Intell based iMacs actually have a very well designed cooling system as evidenced by there rare need to spin up the fans to keep the internal parts within the individual part manufactures' specifications. If something did over heat or get near the upper limit of maximum heat threshold, the fans do spin up.

    Every iMac since the mid-2007 iMac has a separate removable video card except one or two models with the NVIDIA GeForce 9400M GPU. Overall, I've only had to replace one NVIDIA GeForce 8800 GS video card, due to the 8800 die flaw. The majority of the problems that kill Intell iMacs are hard drive failures. Second are power supply failures.
     
  17. MacTech68 macrumors 68000

    MacTech68

    Joined:
    Mar 16, 2008
    Location:
    Australia, Perth
    #17
    Indeed you are correct - I should have checked (my bad). The 9400M models without a removable video card are the 20" Early & Mid 2009 models (but only the 20").

    I don't agree that Apple has set the threshold for increasing the fan speeds at a level that gives the machine the maximum longevity. I can't argue that Apple didn't do a lot of testing to arrive at those thresholds, but I do believe that "quietness" would have factored in their calculations.

    I will agree that Apple has done a great engineering job with the recent iMacs, and my comment of it being a 'nightmare' referred to the complexities of such a task.
     
  18. Mr. Retrofire macrumors 601

    Mr. Retrofire

    Joined:
    Mar 2, 2010
    Location:
    www.emiliana.cl/en
    #18
    Intell? ;)
     
  19. TyroneShoes2 macrumors regular

    Joined:
    Aug 17, 2011
    #19
    Very cool. So how do you and MacTech68 accomplish that?

    Heat is definitely the nemesis of electronics, so I think your advice is sound.

    On a related topic, Leo Laporte has said at least twice now, recently on his podcasts, that the Ivy Bridge chips will run hotter than Sandy Bridge. I thought the entire idea of a smaller process was to decrease the heat created, so I find this very puzzling.

    Can anyone here explain what Leo means?
     
  20. Intell macrumors P6

    Intell

    Joined:
    Jan 24, 2010
    Location:
    Inside
    #20
    Yes?

    smcFanControl or FanControl. I suggest smcFanControl.
     
  21. MacTech68 macrumors 68000

    MacTech68

    Joined:
    Mar 16, 2008
    Location:
    Australia, Perth
    #21
    I use smcFanControl though it's last update was mid 2011

    http://www.eidac.de/?p=207
     

Share This Page