Won't wake up from sleep ("comatose Mac"), the Odyssey; possible progress

Folks:

iMac First Gen 17-inch G5 1.6GHz , USB keyboard and mouse, stock (Apple) RAM, MacOS10.3x --> 10.4 fully updated

My G5 suddenly stopped waking up from sleep. Just to be certain: I mean "sleep" as the condition achieved by Apple Menu-->Sleep or Control Panels-->Energy Saver after the preset time elapses. In this state, the machine is dark and quiet except for the throbbing white light on the lower right. Agreed?

Up until last week, the machine would ALWAYS wake up successfully from sleep in a few seconds on mouse movement or key-press. Now, wake-up from keyboard or mouse fails most of the time, not always, but seemingly more often the longer the machine has been asleep.

All that happens now when I try to wake it with the keyboard or mouse is the throbbing white indicator lamp goes dark. There is sometimes briefly the suggestion of a rotational noise, not clear if it is disk or fan. The machine stays quiescent except for a slight hum. Holding down the power button has NO EFFECT in this state, at least not in the up to 30-40 seconds I've held it down before giving up. Maybe we should call this "comatose Mac" state?

The ONLY recovery is to remove power from the machine, e.g. by pulling the AC power plug. I wait 10 - 20 seconds, re-apply power, and use the power button to boot. Boot is always <sigh-of-relief> normal.

I've read lots of traffic on Apple Support Discussions and other forums about what appears to be the same problem, and I believe I've attempted all solutions and diagnostic procedures offered and one --upgrading the OS-- that wasn't:

--reseting NVRAM and PRAM, separately and together
--repairing permissions
--trashing the power management preferences file, com.apple.PowerManagement.plist
--checking the hard disk
--upgrading from 10.3.x to 10.4.x (which I wanted to do for other reasons)
--reseting the System Management Unit (SMU)
--creating a new user for test purposes
--disabling the screen saver
--logging out before enter sleep
--changed Energy Saver to not put the display or hard drive to sleep separately
--removing all non-essential USB connections, all but keyboard and mouse.

None had any effect or provided any further clues. Yes, of course, it is possible to work around this problem by preventing the system from going to sleep or by shutting down whenever it isn't going to be used for a while. Neither of these is acceptable beyond the short term.

It's a bit disturbing that discussions of similar problems go back as far as 2005, at least, on Apple and other discussion boards with no definitive solution and a lot of flailing around, people suggesting improbably broad solutions like erasing the hard drive and re-installing the OS, or specific but seemingly superstitious ones, like adding system RAM. No way I can prove it, but I have a feeling the comatose state sets in before the processor itself starts executing code, so it seems quite unlikely that many of the measures described above could have any effect at all.

Another possible red herring: I've seen multiple referrals to the G5 iMac Repair Extension program, people saying that this will solve wake-up problems --and my G5 is in one of the affected serial number groups-- but the official write-up ( http://www.apple.com/support/imac/powersupply/repairextensionprogram/) seems clear that the program is intended to fix a different set of issues. I guess I'll have to spend the time to find out directly from Apple.

Finally, here is some possible progress: I read somewhere that briefly pressing the power button is an alternative way of waking up iMacs, so I experimented. For the past several days, this method has worked ALMOST ALWAYS --the one failure may be a fluke-- while keyboard or mouse wake-up continues to fail most of the time. Using the power button is not an ideal solution --it is hard to retrain one's self to avoid the keyboard and mouse for waking up a particular Mac! It's also not near as convenient.

I hope this works for other people. If confirmed, this result would seem to offer the prospect of isolating the problem to a very small portion of the system. I hope so.

Thanks,

Henry

First Gen 17-inch G5 1.6GHz, Mac OS X (10.4), USB keyboard and mouse, stock (Apple) RAM,

Posted on Feb 27, 2007 9:34 PM

Reply
85 replies

Feb 27, 2007 9:53 PM in response to Hen3ry

Are your Mouse and Keyboard Apple ones, Henry?

Had you done anything immediately before you first became aware of the problem? Upgraded your OS, for example? If you are using a "third party" mouse and keyboard , then it is possible that an OS upgrade may have introduced an incompatibility with the keyboard drivers. Kensington and Logitech keyboard / mouse drivers produced this issue at one stage, I seem to remember)

There are quite a number of different causes of this problem (hence the apparent "flailing around" that you mention, and the wide variety of suggested processes for overcoming it). You have, indeed, indeed, tried many of them already.

USB devices and / or their drivers are probably the most common cause, but sometimes firewire or even ethernet devices can also produce this symptom.

Have you had a look at your console / system logs to see if they provide any useful hints?

Cheers

Rod

Mar 1, 2007 9:31 PM in response to Rod Hagen

Rod Hagen:

Sorry for the delay in my replying, I've been buried.

The problem occurred after no particular change that I noticed. It is possible it happened after a routine Software Update, but I cannot tell.

The keyboard and mouse are standard Apple issue.

Good idea to check the log. I haven't had a chance to check the failing system log, but here is a log excerpt from a successful sleep and wakeup on a working system:

...
Mar 1 21:09:23 Espresso kernel[0]: AFPSleepWakeHandler: going to sleep
Mar 1 21:10:12 Espresso kernel[0]: System Sleep
Mar 1 21:10:12 Espresso kernel[0]: Enabling XMM register save/restore and SSE/SSE2 opcodes
Mar 1 21:10:12 Espresso kernel[0]: Started CPU 01
Mar 1 21:10:12 Espresso kernel[0]: IOBluetoothHCIController::restartShutdownWL this is a wake from sleep
Mar 1 21:10:12 Espresso kernel[0]: System Wake
...

I'll run some tests on the failing system as soon as I can and post back to you. Do you have anything in particular in mind that you expect me to see?

As a guy who has designed a bit of microprocessor hardware, I know that powering up a processor from system standby is a very well-defined process that necessarily occurs in carefully-defined steps with a chain of electronic events. It seems clear enough that something goes "flaky", probably on the USB side of the wake-up signal.

I read through maybe 100 posts in my search a few days ago Most of the information was so diffuse and non-systematic thatvery little is really clear except people were having trouble powering up their iMacs. There isn't enough to come close to concluding anything about "causes".

In further tests, the "power button solution" seems to work almost all the time, but not always. One problem is that if you are not careful to look for the throbbing white light, you may attempt to wake a system that isn't actually sleeping. This seems to kill the system and requiring powerdown followed by a reboot, at least in the failing system. It isn't an experiment I want to repeat often on any system!

The suggestion by Sachi to swap keyboard and mouse with another system holds promise. I'll try that next. If it works, this is starting to smell like a component drift problem. As time goes on and environmental changes occur, components values can drift, thresholds will change. With bad luck, you get ...flaky wakeup. (Who knows, room temperatures could contribute. Does this problem occur more in winter months?) Component drift isn't often an issue with digital systems, but the power circuitry is a special case.

I'll post back as soon as I can.

Thanks,

Henry

Mar 1, 2007 9:37 PM in response to sachi

Sachi-San:

Thank you very much for your suggestions. I'm sorry that I took so long to respond.

I continued my discussion with Mr. Hagen, as you can see, and I agreed that your suggestion is a good one. After I get some data on the failing system as it is, I will try to trade the keyboard and mouse with another system. All working Apple peripherals.

Out of curiosity: Are you having a cold winter in Japan? Are you operating your Mac in an unusually cold room? As you see, I am wondering if room temperature could have an effect. We in California have had unusually cold temperatures in the last month, I realize. Our heating is not very effective, so this may have had an effect.

Thanks/Arigato gozaimasu,

Henry

Mar 1, 2007 10:03 PM in response to Hen3ry

Hajimemashite Hen3ry san!

It's worked fine 'till now 😟 It couldn't wake up from sleep at lunchtime again. I gave up and had a contact with my Apple dealer where I bought it then they will take it.
BTW, I'm in Nagoya and the weather in the winter not to cold at all, like 0- +5 celsius around in the morning, but this year it was hotter than last year. I don't think so this temperature has an effect at all.
Just at the summer time is very very hot in my room and my Power Mac's fan usually running much often than I turn on the aircon.

Jane

Mar 1, 2007 10:04 PM in response to Hen3ry

I'll run some tests on the failing system as soon as I can and post back to you. Do you have anything in particular in mind that you expect me to see?


Signs of problems accessing particular USB devices or their drivers would be pretty high on my list to look for here, Henry. Network issues would also be worth keeping an eye out for.

When you upgraded from 10.3 to 10.4 how did you do it? With a "clean" installation of 10.4 or by simply upgrading the original installation?

As you have one of the "first generation" iMac G5's it may also be wise to pull the back off and check whether there is any sign of bulging or leaking capacitors. They should all have nice flat tops. (see http://en.wikipedia.org/wiki/Capacitor_plague ). Many of the "first generation" model suffered from this. If so, you should be eligible for the extended repair program, even if your "issues" are not quite the same.

Cheers

Rod

Mar 2, 2007 3:51 PM in response to Rod Hagen

Rod Hagen:

Thanks for your response.

The machine has no problems with USB devices or drivers as far as I can tell. No network issues, either. It operates normally in every respect except powerup.

I think I did a clean install of 10.4, but I am not certain. It's simple enough to repeat the install, but I still strongly doubt that the OS version has anything to do with this problem.

I'll also look for bulging capacitors. Certainly if I see anything abnormal, I'll contact Apple and, hopefully, they'll take care of the problem.

As you'll see elsewhere in this thread, I had no success substituting a keyboard and mouse.

Thanks,

Henry

Mar 2, 2007 3:57 PM in response to sachi

Jane-san:

Thanks for your response.

Changing keyboard and mouse to another set in all combinations did not help for me. Maybe even made the problem worse. I'm going swap with another set (much newer ones) soon and see if that makes any difference. I will post back with results.

Nagoya? Big city! By coincidence we have a neighbor family from Nagoya on our street here in Northern California.

Well, it was worth asking about temperatur. I have seen very specific system problems due to a temperature-senstive part. (I found it with a can of cooling spray.) Maybe some environmental factor other than temperature is responsible. It is a small chance, but one learns to question everything.

Good luck with the dealer! If you discover anything during the process, please let us know.

Thanks,

Henry

Mar 2, 2007 5:17 PM in response to Hen3ry

Hi Henry,

When mentioning USB and network issues, I was talking about log entries specifically at the time when the "wakeup" problem occurs, rather than problems in general operation at other times.

OS issues can be of some significance here. In fact some of the problems with particular hubs etc only arose when people updated to specific "intermediate" versions (OSX 10.4.2, from memory, caused problems with some 3rd party devices, for example, some of which were resolved by the development of updated drivers by the manufacturers concerned).

Cheers

Rod

Mar 2, 2007 6:52 PM in response to Hen3ry

I hve been experiencing same problem after updating to 10.4.8 The senquence of failure went like this.. first the sleep light would not always blink but it would wake up. then it would not wake if I had my Palm adapter plugged in, then it would not wake with my Canon scanner plugged in (USB) Now I have no light and have turned of the sleep mode. If I put it to sleep it will not wake. I h ave also noticed a dim down of my screen when first coming out of screen saver mode.

Are the capacitors talked about in the power supply circuit? and I live in minnesota - the problem was noticed when warm and now it is very cold and has not made any change in how it works.

Mar 3, 2007 3:36 PM in response to johng5mac

Johng5mac:

Thanks for your post on this thread.

Go ahead, re-install, and let us know if it helps. My guess is it will not.

Earlier in this thread I remarked that the only thing that seems common in all the messages I've read is:

(1) People are having trouble their Macs from sleep.

I'd like to take this opportunity to add this second observation:

(2) These problems seem to start occurring at random times, not necessarily connected to anything done or not done to the computer.

My feeling is it is pure coincidence that the problem started about the time you updated to 10.4.8.

Have you tried removing all of your optional USB connections? (That is, with just Apple keyboard and mouse attached.) That seems simple enough to do, and is one of the most plausible suggestions I've seen.

I'll share one more observation that is only my own:

(3) It seems quite possible that wake-up problems can disappear without any specific fix.

I didn't see any message mentioning this, but --well-- people often don't post when things are working. I encountered what seems to be exactly the same problem on my G4 Sawtooth some years ago. Eventually the problem went away, but I can't claim anything I did was responsible.

I did see posts that said this or that fix solved their problem. I wonder how often the problem went away on its own coincidental with something someone did.

Good luck with the problem and with your weather, which I understand is quite extreme right now. We just got a family weather report from central Mn: blowing snow 12 - 18 inches. Whew. I won't tell you about our weather here in Northern California.

Thanks,

Henry

Mar 6, 2007 6:07 PM in response to Rod Hagen

Rod Hagen:

Thanks for your response and sorry for my delay in responding.

Right, I ran some tests. When the G5 fails to wake up from sleep there are no corresponding system log entries at all. I think this confirms that the system is, well, catatonic. If the processor is running at all, it isn't running Unix.

A successful wakeup using the power button seems to be registered in the system log just like a successful wakeup using a keyboard or mouse on a 100% working system. I didn't examine this in great detail.

Swapping keyboards and mouses with another (3 months-old) iMac changed nothing. The failing system continued to fail in the same way, and the other system continued to operate normally.

There's no external hub connected to the failing G5. For the moment, only a standard Apple keyboard and and Apple mouse.

By the way, when the system is catatonic, it does not respond to pings from another machine on the network, and --of course-- it can't be contacted by high-level networking.

The most frequent user of the system confirms that using the power button to wake up works reliably.

If we knew the technical difference between waking up with mouse or keyboard (which fails most of the time) and waking up with the power button (which works all the time), then we ought to be able to isolate the actual problem with a very small area. I do NOT think it is software, or anything that could be directly influenced by software.

Yes, I know, it may seem that a particular high-level event like a software update may seem to be responsible, but my guess is that people searching rather desperately for solutions may tend to interpret coincidences as cause-and-effect. It certainly could be a high-level cause, but there's no plausible mechanism for such a failure, at least no that I've found proposed.

Thanks,

Henry

This thread has been closed by the system or the community team. You may vote for any posts you find helpful, or search the Community for additional answers.

Won't wake up from sleep ("comatose Mac"), the Odyssey; possible progress

Welcome to Apple Support Community
A forum where Apple customers help each other with their products. Get started with your Apple Account.