Q: 27-inch / 27" iMac Random Shut down

I also have a 27" but with 4GB of factory RAM.. Nothing altered on the computer, factory original RAM.  I have about one unexplained shutdown every 1-1.5 weeks on average.  It is a I3 3.2GHZ.  I am starting to investigate this as we just spent alot of money on what we thought was a very nice machine.  I have the factory 1 year warranty only and am waiting on a call from Apple on Friday.

I ran MEMTEST and it didn't come up with any errors... Any tips on dealing with Apple on this matter?  The gist I get from reading these posts is that nobody knows what the solution is (including Apple).  The SMC reset didn't do anything for me either.

The Mac is used VERY lightly, mostly for web browsing.. No VMs or parallels, FC, anything like that.  I'm a little dissapointed if I am being honest.  Paying a premium on hardware shouldn't end up with more frustration than a $500 Hackintosh with a fancy screen.

I had upgraded my 2010 27" i7 iMac to 16GB with OWC's 'house brand' memory and it started the random shutdown as described in this thread.  After multiple attempts at resolving (reducing 16GB to 12GB or 8GB, reset SMC, reinstall, etc...) the real fix was to replace with more compatible memory. 

Since Apple used Samsung memory originally, I worked with OWC to replace their house-brand memory with genuine Samsung memory (which OWC also offers, but at a slightly higher price) and once done the system has been 100% random shutdown free with 12GB of memory since then (6 or more months ago).  As pointed out here, the key is to go with what has proven to be compatible; if that's Hyinx or whatever, go for it.  But since Apple uses Samsung, I use Samsung...

I don't know if the Samsung RAM at 16GB configuration would be stable or if there is really some innate instability when running at 16GB regardles of memory brand.  To be safe, I went down to 12GB.

Looking through all the posts I can find on this issue, it seems that everyone is experiencing a similar problem but attributing it to different causes. My own 27" 2.66 i5 has frequent random blackouts but is completely factory spec, with the original Apple 4Gb of RAM. I can generate the blackouts really easily - just use any software that heats up the GPU. So running EyeTV, playing movies in iTunes, viewing YouTube or any Photoshop work. Movies in iTunes are a very reliable way to generate a shut down. My own fix has been to download SMCFan Control and create a range of settings - fast, medium, slow etc - and if I remember to select the right one, I use a fan setting to match the software. The results are pretty clear - with the fans turned up, there are no blackouts in iTunes, EyeTV etc whatsoever, no matter how long they run. But with the fans turned down, blackout can happen within a few minutes of startup. I can pretty much guarantee these results every time and so I'm convinced it's a heat-related, hardware problem.

There's a lot of posts that report Apple have tried power supply replacements and RAM replacements etc but the results are of varied successes. So it's likely that these were lucky fixes, if they've worked.

Since my Mac ran perfectly for a year till about a month over the warranty period , my own theory is that at some point, something inside - probably a chip on the graphics card - seriously overheated and the solder got melted or became dry/cracked. Once cool, the solder contracts and it reconnects, so everything runs fine. But as soon as a specific critical temp is reached it's likely that the solder expands and re-opens any breaks that have occurred - it would only take a hairline crack, on a single connection of a chip's soldering, to create a problem. If you consider the millions of soldered connections inside, and the billions of instructions whizzing round, just one faulty joint is bound to cause a problem pretty quickly.

As everyone seems to have a similar result - the same black screen and seemingly dead Mac, but with the fans still running - I suspect that it's the same soldered connection or the same component in every case that's getting heat damaged. And I'm guessing again, but I bet all of the iMacs with this problem have a Radius graphics card. Looking around various posts, this appears to be a common link and there are also plenty of reports of heat-related faults with Radius cards out there, from PC users as well as Macs.

This is not the first time Macs have had overheating problems. A friend of mine's Powerbook was replaced recently - despite being out of warranty - when he took it in to Apple, because on that particular model, Apple was aware that the solder on one of the chips was known to overheat and melt.

There are also posts around regarding Apple's free replacement of the graphics card on one of the earlier iMac models, if you took it in to Apple (even if the machine was well beyond warranty) but in this case it was nvidia who had admitted supplying faulty cards. In each case though, it seems that you only get to know this if you take the machine in and report the fault.

I suspect that Apple is fully aware of this problem - it's too widely reported - but that they are waiting for Radius to admit a problem with its cards so that they can foot the bill. We may all have a long wait. Shame really, as I adore my iMac and am already considering buying another as a backup machine, despite the problems. I'm just hoping that I get a lucky one that doesn't break down.

Actually, I don't think you have the same symptom we're discussing.  Our iMacs shut down as if you pulled the power cord out of them - that is, no 'fans still running' like you describe.

I'm waiting on the replacement memory, but since popping out the memory I have not had a single shutdown

as noted in Terminal: (uptime)

13:24  up 6 days, 18:43, 2 users, load averages: 0.10 0.11 0.08

iMac-i7

I been doing exactly the same stuff with my Mac, so bottom line I'm sure it's the 3rd party memory. I know Apple from experience has always been very very particular about memory. I also know this from being a certified Apple service tech for many years. Once I have the replacement memory installed and I'm satisfied it's good I'll post.

Akitaguy wrote:

 

I been doing exactly the same stuff with my Mac, so bottom line I'm sure it's the 3rd party memory.

You're sure it's the third party memory, even though the vast majority of those who've had this issue have factory-installed RAM?

Vast majority? In that case Akitaguy and myself must be in the extreme minority because I can confirm that after removing the 3rd party crucial ram and reinstalling the Samsung ram that came with the iMac, the shutdowns have seized to occur. I will note that I haven't changed my computing habits before or since, still running the same software, photoshop/painter/blender, all without a hitch. I believe the extreme minority also includes Tank9, who suggested going with Mushkin ram. It's the 3 of us vs. the "vast majority".

You're sure it's the third party memory, even though the vast majority of those who've had this issue have factory-installed RAM?

From reading the 20 odd pages, the vast majority have added 3rd party memory. The sudden shutdowns we're discussing here are as if the power cable was pulled. The computer won't restart on it's own, we need to press the power button. Some have added memory to  the original, others including myself have replaced all the original memory. Once this new memory has been removed the computer is stable again. Also as I've stated

the iMac has be stable as a rock, no crashing, freezing, slowdowns, I run my computers for weeks only rebooting for system updates. I have another iMac thats 5 years old, it's been running for 5 years steady.  This iMac i7 has been running since it came out of the box at the beginning of Sept last year, the  only issue  3rd party MEMORY!

Guys and gals, check the following link

http://forums.macrumors.com/showthread.php?t=1064187

People are coming to the conclusion that some of these symptoms may be related to a bug in the way in which the kernel deals with the Bluetooth mouse and keyboard.  This definitely matches my problem and more importantly the frequency (1 week-2 week) intermittency.

My console log in the "Diagnostic and Usage" messages yields:

------

11-04-11 8:51:08 PM    configd[15]    Sleep: Drivers Failure - AC - EHC2

com.apple.message.domain: com.apple.powermanagement.sleep

com.apple.message.signature: Drivers Failure

com.apple.message.result: Failure

com.apple.message.uuid: 12A1C207-C888-4176-8F35-5334574EE39D

------

These messages at about the same time the problem occurs.

The key is to TRY YOUR CAPS LOCK KEY.  If the green pilot light does NOT function during the time you are having troubles starting up your iMac, THIS IS THE PROBLEM.  Not a hardware issue!

Your kernel log will also have errors such as:

-------

11-05-01 7:49:20 PM    kernel    [[0x92a4000] Req OpCode 0C1A from: kernel_task (0) status: 00 state: 2 timeout: 5000] Bluetooth warning: An HCI Req timeout occurred.

--------

One "solution" is to not use Bluetooth peripherals, but we are likely looking at a software update hopefully soon, once Apple gets wind of this.

I also looked through the crash logs, seen odd entries related to a number of USB devices, bluetooth and so on

Again once the 3rd party memory was removed and replaced with the original the issues where gone.

Remember all these devices also require access to memory, it's all tied together. As I mentioned in my last post

after going back to the 8 GB of Apple memory the issue was resolved. What's the sense in running a computer without bluetooth, USB devices, or only 12 GB when it was designed for all this. 

IT'S a MEMORY issue plain and simple.

Ok well my imac has virgin factory Samsung ram so there may be more than one issue here.  I am just offering one possible solution with examples of what is in my log.  Your results may vary.

Also, I am not incinuating that it is acceptable to not use Bluetooth at all or that I am conceding that these are satisfactory problems to have in the first place.  I am just as ****** off as you are, as a matter of fact.

This thread isn't about a GPU heat problem shutting down the screen. It's about specific ram that shut down the computer completely. Please read before posting.

Zero shut down for over two months with 16Gb Mushkin/Hynix Apple ram on my 2010 iMac i7.

If you guys have all the answers to the problems already then what is the point in continuing this thread?  Perhaps you should look up the definition of "community" before being rude and preaching to others about posting.  There is nothing in the topic that even remotely refers to RAM.

I want to throw my hand into this issue as well. I have a i3, it worked fine with 8gb. Upgraded to 16gb of Mushkin ram (non hynix), it was good for about a week before I started having the random shutdowns. I contacted Mushkin, and with recommandation from this thread, I asked for them to replace it with the hynix chip version. I had a good 2 days after installing the hynix chips before the random shutdown's happened again. I ran threw memtest with no errors. I haven't contacted mushkin about the recent problem yet. So I don't know if i recieved a bad/defective batch of hynix chips, or if others are just a bit more lucky then I am and the shutdowns will still happen....

I am new here, but that is because all of my Mac products have worked flawlessly until now. I have a 27" i3 iMac (mid 2010) and it worked wonderfully on the 4 GB of RAM that came installed. I bought (it sounds like) similar Mushkin RAM and put in 16 GB like others and as of the day of install it mysteriously shut down that night. It has since done it again. The RAM has only been in for two days.

I have read through every page of this thread and it seems as though the first step is to remove 4 GB and make sure to have matching pairs (including the original Samsung RAM). This really ticks me off, why claim something about a product that appears to be untrue? Has anyone went directly to Apple with this issue, or is this something I should take up with the manufacturer? It sounds as if both have already been done, but neither seems to have satisfactorily resolved the issue?

I am still under the 1 year warranty, and I will check my RAM to see if it is the Hynix(?) chip tomorrow, but if this computer is unable to perform up to it's supposed specifications then I feel as though I was mislead and should be able to return this computer. He11, I am still paying it off .

I am getting no errors and all the hardware checked out fine. It is recognizing all RAM.