PC won't properly wake from sleep (sometimes)

[adambean]

Hey all,

Have been fighting an ongoing issue that I'm simply unsure how to resolve. My sleep settings are default (minus PCI Express Link State Power Management - which is off to help this issue) and yet my machine regularly won't wake up from sleep.

I come back to my machine after a period of inactivity and upon pressing mouse/keyboard, monitor and PC turn on; however, no display is shown.

I have to physically power down the machine; however, upon turning it back on and booting through a normal sequence, my Windows session will resume where I left off, almost like a hibernate.

I'm assuming the answer is right in front of me; however, curious to learn from others what they think of the sequence of events I see below in my system logs:
- 2017.01.11 10:58AM: The system is entering sleep.Sleep Reason: System Idle
- 2017.01.11 2:52PM: The system uptime is 70581 seconds.
- 2017.01.11 10:58AM: The system has resumed from sleep.
- 2017.01.11 2:52PM: The system time has changed to ‎2017‎-‎01‎-‎11T19:52:09.500000000Z from ‎2017‎-‎01‎-‎11T15:58:13.096615000Z. Change Reason: System time synchronized with the hardware clock.
- 2017.01.11 2:52PM: Intel(R) Ethernet Connection (2) I218-V. Network link is disconnected.
- 2017.01.11 2:52PM: The firmware reported boot metrics.
- 2017.01.11 2:52PM: There are 0x1 boot options on this system.
- 2017.01.11 2:52PM: The boot menu policy was 0xF0570.
- 2017.01.11 2:52PM: The boot type was 0x2.
- 2017.01.11 2:52PM: Windows cannot store Bluetooth authentication codes (link keys) on the local adapter. Bluetooth keyboards might not work in the system BIOS during startup.
- 2017.01.11 2:52PM: The Windows Error Reporting Service service terminated with the following error: Windows Error Reporting Service is not a valid Win32 application.
- 2017.01.11 2:52PM: The system has returned from a low power state.
Sleep Time: ‎2017‎-‎01‎-‎11T15:58:08.919219600Z
Wake Time: ‎2017‎-‎01‎-‎11T19:52:10.283284700Z
Wake Source: Unknown
- 2017.01.11 2:52PM: The access history in hive \??\C:\Users\xxxx\AppData\Local\Packages\Microsoft.WindowsStore_8wekyb3d8bbwe\Settings\settings.dat was cleared updating 5 keys and creating 2 modified pages.
- 2017.01.11 2:52PM: Intel(R) Ethernet Connection (2) I218-V. Network link has been established at 100Mbps full duplex.

Thanks!

 

[dextermat]

The best way to get rid of the problem is to turn off sleep/hybernation function.

 

[bigpinkdragon286]

A boot type of 0x2 means it resumed from hibernation, so it sounds like your system is performing a hybrid sleep. That would explain why it boots back where you left off in the event of a failed wake from sleep.

As far as the reason for the failed wake from sleep, I don't see it in the information present in your post.

 

[adambean]

@dex - well that's not really a solution, would prefer to fix as I'd prefer to allow the machine to sleep when not in use.
@pinkdragon - any insight to where else I can/should look?

 

[reedo_43]

http://techreport.com/review/24897/the-big-haswell-psu-compatibility-list

 

[adambean]

Good find reedo. Here's my build: https://pcpartpicker.com/b/ckYrxr

I have 5930K CPU and a EVGA SuperNOVA 1000 P2 PSU ... which is not on your list. So do you think that is my issue? Is there any resolution?

 

[SgtScream]

Were you ever able to wake from sleep or has it always been an issue?

 

[reedo_43]

http://forums.evga.com/EVGA-SuperNOVA-1000-P2-m2125489.aspx
Sorry but it was a good guess.

 

[adambean]

@SgtScream - a very good question. I'm quite positive that early on (the rig is a bit over a year old now), I did not have issues with sleep; however, I feel early on that I disabled sleep as I used that as my primary work from home machine. So... not entirely certain to be honest. The odd thing is though that sometimes it wakes up without issues, but I'd say the majority of the time I have to do a shutdown.

@reedo - damn, thanks for finding both, at least I know.

 

[bigpinkdragon286]

I would start by looking in the Windows logs for unexpected reboot scenarios, then peruse the events that immediately preceded those for possible hints. Just because your screen doesn't come back up, doesn't mean Windows hasn't resumed from sleep, only to have a driver that doesn't come back up with it.

Few things I have run into that would cause wake-up issues were, having CPU virtualization turned on in BIOS and RAM that was unreliable. The CPU virtualization was consistent, the RAM reliability is about a 1 in 20 failure rate, so I lived with that, as the system was a non-critical, game playing machine.

How is the grid power in your area? Are there any outages or issues during the time your machine was sleeping?

Another thing I would certainly try is, disconnecting all of your wireless peripherals, and disabling the ability for things to wake the machine from sleep. This is for troubleshooting purposes. If your machine wakes reliably after you've turned off the extra wake options, turn them back on one by one until you find the culprit.

 

[adambean]

@bigpink
- Agreed, I need to pay more attention to the logs when this happens to see if anything sticks out. I've just been lazy and frustrated with this machine. The motherboard has been very problematic for me overall and just sick and tired of Asus.
- Yeah, I enable virtualization as I do run VM's.
- Power is fine, I have 2 desktops and 2 laptops in my house (wife and I both work from home) and no other machine has issues.
- I do not have any wireless peripherals on this machine. Lots of USB stuff though (again stupid mobo, had to downgrade my controller to USB 2.0, ugh).

Thanks

 

[bigpinkdragon286]

Just my own opinion, but I strongly dislike ASUS equipment and find it highly overrated.

Generally speaking, of course laptop computers would not have power issues, as they have built in battery backups, unless run without their batteries. I would advise against using those as any metric for reliable home power. Also, the quality of the power supply you are using in each desktop can make a big difference. What is the hold-up time for each in the event of a power failure? A lot of them don't meet ATX requirements, but how many folks actually pay attention to this? A minor power fluctuation for one could be a major fluctuation to another.

Is the problem machine on any sort of UPS system, or is it directly reliant on the power coming into the home?

 

[adambean]

I agree, this will be my last Asus mobo and router, have had nothing but issues with both.

Yeah, I am on a UPS, but seems to be stable? Unsure what I would really need to check for.

 

[bigpinkdragon286]

The UPS needs to have a switchover time greater than your hold-up time on your PSU, or the PC will exhibit issues after a switchover. Your PSU needs to hold it's voltages to within ATX spec for a certain number of milliseconds after AC loss, and the UPS needs to switch to battery and restore AC power before your PSU is unable to hold-up the voltages. If your system is asleep at the time this occurs, I would assume it's drawing small enough from the PSU that hold-up time should be pretty great, but EVGA power supplies really aren't well regarded in hold-up time, depending on the manufacture making that particular supply. They cheat by letting their voltages go out of spec before they shut down to get better hold-up time measurements.

Also, not all PSU's deliver good results when on battery backup, as many battery backups put out really nasty stepped sine waves. These can cause the active PFC regulation in the PSU to be unhappy, if it's not very good. May work, may not work. I always advocate charging the UPS, then pulling the plug with the system running typical tasks to verify it works correctly. Waiting until a power outage could result in lost work.

You could always swap the power supply with the system that wakes reliably, and see if you can consistently wake either machine after that. I know, it's a pain, but without other directions to go, sometimes you have to take the long shots. I certainly wouldn't rule this out.

I know you mention that you use VM's on the machine that won't wake reliably, but I would certainly disable virtualization for the purposes of the testing. In my case, virtualization had to be disabled in Windows, through the "Turn Windows features on or off" portion of Programs and Features in the Control Panel.