crwdns2933423:0crwdne2933423:0

Model A1312 / Mid 2011 / 2.7 & 3.1 GHz Core i5 or 3.4 GHz Core i7 Processor, ID iMac12,2

crwdns2934591:0708crwdne2934591:0 crwdns2934593:0crwdne2934593:0

iMac random restarts - still stumped? The definitive diagnosis?

Dear All,

The issue of random restarts on my (now old) iMac mid-2011 continues, as does the struggle for a proper diagnosis of the problem. This used to be my main desktop some years ago; later I sold it to a family member, who after a couple of years started reporting this problem. Then I took it back so I could try to fix it for good (I love a challenge). Yet the issue remains.

So I am writing to this knowledgeable community so i can try to, once and for all, find the answers I need. Now let me explain again what the problem is and what has been done already (I had already posted the original question containing “AHT” elsewhere here):

PROBLEM: Random restarts in the iMac mid-2011 (3.4GHz i7), 16GB RAM (Crucial) - they happen virtually anytime between/during boot time and normal operation, and most often right after the login is entered or a little after the Finder is active. Please note that this is NOT a random SHUTDOWN issue; but just random RESTARTS.

NOTABLE EXCEPTIONS: it NEVER happens when you invoke the AHT test (D boot key) or when acting as target in target disk mode (for instance, through FireWire); it also does not happen when booting in ctrl-R (recovery mode), except on a few occasions in the middle of a macOS reinstall on the main drive.

ACTIONS TAKEN (OR EXCLUDING FACTORS) - Virtually everything has been tried so far:

  • Changing households (to exclude power outlet issues) - it has happened both in my house and my relatives';
  • changing power cord - since this was originally bought in the US (and I currently live in Europe), I've tried swapping power cords, also to no avail;
  • erasing/reinstalling macOS - several times with High Sierra; apparently this seems to stabilize the functioning of the iMac for a while, but then the problem comes back again;
  • replacing the PSU - even though the first one I ordered had an issue with the temperature sensor, a good replacement part came and the problem still continues;
  • changing the boot drive/using a different SATA connection - thinking that perhaps the SATA cable may cause the problem, I reinstalled macOS on another SSD with the second SATA connection available in the iMac - it seemed to resolve the issue for a few minutes/hours, but now it is back again;
  • removing/reseating the added RAM - the problem continues;
  • Apple Hardware Test - with the original and then bad replacement PSU, an issue was reported with the temperature sensor; afterwards, with a good PSU, no trouble is indicated; this EVEN after running the overnight extended tests;
  • GPU - I can confirm that no graphical artifacts/banding have ever occurred with this iMac; not to mention that the AHT shows no trouble at all;
  • fans/sensors - all seems to work fine now, so this does not seem to be the cause of the problem;

WHAT ELSE, THEN?

Given the above, the only remaining factors that I could remotely consider are:

  • Major logic board issue - improbable, since nothing has ever been reported on AHT;
  • bad primary SATA connection still plugged to the first SSD - even if it is no longer the boot drive, it might be that the bad connection causes random restarts just because it is still active somehow;
  • extreme sensitivity of the PSU to power spikes/oscillations, thus requiring use of a UPC unit - highly improbable since the problem has been reported in different households; yet not totally impossible;
  • software issue somewhere? Funnily this is what I tend to believe in right now, since the problem does NOT happen at all when doing the AHT or acting as target disk mode. But who would be the culprit? I have not tried installing Windows/Linux on it, so this is unclear.

IN CONCLUSION: I hope this helps others facing the same issue, and assists us in collectively debating the root of the problem. Thanks!

crwdns2934081:0crwdne2934081:0 crwdns2934083:0crwdne2934083:0 crwdns2934093:0crwdne2934093:0

crwdns2934109:0crwdne2934109:0

crwdns2889612:0crwdne2889612:0 2
crwdns2934285:0crwdne2934285:0

crwdns2933315:05crwdne2933315:0

crwdns2934051:0crwdne2934051:0

What does the console logs tell you? Open your Utilities folder to locate the app.

Make sure your systems clock is correctly set and use a high precision clock next to your system so you can catch the time point the system restarts. Now open Console and locate that exact time point.

crwdns2934105:0crwdne2934105:0

crwdns2889612:0crwdne2889612:0 1

crwdns2944067:06crwdne2944067:0:

Dan, I also forgot to tell people this: I have also changed the PRAM battery, just in case. As for Console, nothing really shows apart from boot_time without any specific error code.

crwdns2934271:0crwdnd2934271:0crwdne2934271:0

There needs to be some clues within it. It may not stand out. You might want to get a log cleaner to flush the old logs to make it a bit easier.

If you truly can't see anything that would imply the logic board VRM's are failing to supply the required voltage to the CPU.

You might find someone with deeper skills diagnose the logic board or you could just replace it.

Logic board's (820-2828)

● 2.7GHz - Apple P/N 661-5948

● 3.1GHz - Apple P/N 661-5949

● 3.4GHz - Apple P/N 661-5950

crwdns2934271:0crwdnd2934271:0crwdne2934271:0

Tks - does the VRM hypothesis stand up even considering that there are no such restarts when both AHT and Recovery systems are being used? Not to mention that the logic board shows no sign of trouble under the AHT's extended test.

crwdns2934271:0crwdnd2934271:0crwdne2934271:0

VRM's are reactive to the processing load on the CPU. So as the load is increased they wink out. The process load of AHT and Target disk Mode are not very high. As you don't see any dedicated GPU issues that removes it (are you sure?) I can't explain any other answer.

crwdns2934271:0crwdnd2934271:0crwdne2934271:0

Tks again - but wouldn't this just lead to an automatic shutdown as opposed to a restart? Would you know where these VRM are located in the LB? Another question: would the use of an UPS help in this case?

crwdns2934271:0crwdnd2934271:0crwdne2934271:0

crwdns2934273:01crwdne2934273:0

crwdns2934285:0crwdne2934285:0

Dan, tks a lot for all your comments - I just wanted to add another element here: the diagnostic LEDs, as checked from the low/rear vents, are working as follows:

  • LED 1: on all the time as long as the cord is plugged;
  • all other LEDs: remain on for as long as the computer is on (following a restart).

However, random restarts continue to occur; and when this happens LEDs 2, 3 and 4 all go off until the boot sequence is concluded (which would be normal, I guess). Funnily enough (or perhaps just an anecdotal thing), the iMac seems to restart more often when sitting with the screen down (on the cloth) as opposed to upright. Should I remove all RAM sticks just to double-check if there is a problem of RAM seating?

Update (04/29/2019)

I may be getting close to the culprit here - is it the power button? I was toying with this possibility by pressing it multiple times during the startup process - and was able to have the same restarts even more often. If this is the case, how would I fix it?

crwdns2934105:0crwdne2934105:0

crwdns2889612:0crwdne2889612:0 0

crwdns2944067:021crwdne2944067:0:

I think you're on a bit of a snipe hunt here.

The power logic is a digital design not physical. The power button only causes the power supply to enable via a relay. The onboard diagnostic LED's #1 LED would have turned off. if the power supply was defective. The fact the other LED's go off supports the power logic within the logic board is the problem.

crwdns2934271:0crwdnd2934271:0crwdne2934271:0

But all other LEDs remain on as they should, and only turn off once the random restart happens.

crwdns2934271:0crwdnd2934271:0crwdne2934271:0

I'm not following you?? Thats the reaction I would have expected. When the system has the event the LED's will go out but the first one.

crwdns2934271:0crwdnd2934271:0crwdne2934271:0

My point is that the iMac Technician Guide explains that when LED 2 is on, it means the PSU and voltage regulators are OK - the problem is that the random restarts continue to happen even then.

crwdns2934271:0crwdnd2934271:0crwdne2934271:0

The VRMs are the voltage regulators.

crwdns2934271:0crwdnd2934271:0crwdne2934271:0

crwdns2934275:016crwdne2934275:0

crwdns2934285:0crwdne2934285:0

After a lot of investigation and tinkering the issue still continues, but interestingly to a much lesser extent after I fully erased all drives, removed APFS and reinstalled the latest version of Mavericks (instead of High Sierra).

I have also inserted a new PRAM battery, fixed it with lots of tape (as it seemed to be getting off place without it) and reassembled the whole thing again.

Dan’s suspicion that this has something to do with VRMs still stands; however, I still struggle with the fact that such random restarts are no longer occurring under Mavericks in the same manner they did before under High Sierra - is it because Mavericks has lower energy requirements? Not clear.

So as a sort of “intermediary” conclusion - random restarts DO NOT occur when:

  • the Apple Hardware Test is invoked and run with the D key;
  • the iMac is used in Target Disk Mode or Target Display Mode;
  • the recovery partition is invoked with cmd-R.

The problem seemed ALMOST solved under Mavericks - only a couple of restarts happened so far (under normal usage/installing programs, etc.).

crwdns2934105:0crwdne2934105:0

crwdns2889612:0crwdne2889612:0 0
crwdns2934285:0crwdne2934285:0

I’m having the exact same issue on my mid 2011 imac 27”,

it seems to be the end game issue for even well looked after iMacs of this age, I also had less of the issues on snow leopard but they still occurred, so i don’t think its a software issue

I suspect its a power supply under load issue, getting brand new power supply’s is very hard except perhaps from China as most are just ripped from an equally old iMacs etc and the 310w apple specced supply has very little head room with age to keep the machine going, no bulging caps on mine either but meh.

I’m going to run AHT, can I do that from optical media? Anyone know where I can download for the 27” mid 2011. I think i have the original optical media that came with the machine do they contain an AHT partition / disk? I don’t want to use the disks I had in it, as the random restarts manged to damage 3 drives and I dont want to harm them more, so I can’t use the recovery partition on my SSD for the moment I’ll have to try a sacrificial install on an older disk I guess?

crwdns2934105:0crwdne2934105:0

crwdns2889612:0crwdne2889612:0 0

crwdns2942205:01crwdne2942205:0:

The best way is run the diagnostics from the grey disk.

crwdns2934271:0crwdnd2934271:0crwdne2934271:0

crwdns2934285:0crwdne2934285:0

I’m getting exactly the same issue on my recently rebuilt 2009 27” i7 iMac. I had had to replace the graphics card, so installed one from a 2011 model which is compatible, having taken the time to renew the thermal past on the GPU and CPUs.

I have read that Sierra and High Sierra have some sort of bug that can cause random restarts - this was introduced in the last version of Sierra (10.12.6) Apparently, it is possible to circumvent this when installing from a USB stick by swapping out to an earlier installer on a separate USB stick, which somehow fools Sierra and installs correctly. Sorry it’s a bit vague - but if you google around for “Sierra random restart bug” and you should be able to find out more.

crwdns2934105:0crwdne2934105:0

crwdns2889612:0crwdne2889612:0 0

crwdns2942205:01crwdne2942205:0:

boblowes - There is no random restart bug in the macOS Mavericks or Sierra! I get a kick of people posting stuff without thinking things through. There is so much mis-information on the web it gets hard to be rational.

So let's tear into the mis information!

1 - Apple enabled automatic update services, so the system if left in sleep mode will update during the night when needed. This freaked people out.

2 - Memory leaks! There was a bunch of apps which were using a retired memory call (including Apples apps and OS), so these apps would let the memory space dirty. As part of the housecleaning services if found the system would flush - Reboot!

I still use Sierra on my 2011 27" iMac without any restarts after I found the memory leak. It was tough to spot!

Re-installing the OS tends to hide the issue if you refresh your OS installer and apps with the later versions as Apple did fix a memory leak they had in the newer update.

The issues most people encounter:

● Discreet GPU fails! This happens when you push the system beyond what it was designed to support for graphics load. Heavy graphics games and on online gaming, deep image or video editing. Need a new GPU board.

● VRM failure! Logic board issue as your CPU load increases the VRM's can't supply the needed power so the system crashes.

crwdns2934271:0crwdnd2934271:0crwdne2934271:0

crwdns2934285:0crwdne2934285:0

crwdns2934229:0crwdne2934229:0

Ricardo GF crwdns2934231:0crwdne2934231:0
crwdns2936625:0crwdne2936625:0:

crwdns2936751:024crwdne2936751:0 0

crwdns2936753:07crwdne2936753:0 1

crwdns2936753:030crwdne2936753:0 6

crwdns2942667:0crwdne2942667:0 1,923