I am a computer-repair professional. I installed a new motherboard / processor / memory combo and Windows 7 Pro x64 in my own shop in December 2013.
The installation had one notable intermittent problem. Standby (aka "suspend" or "sleep") was configured to occur after 30 minutes with the "Allow hybrid sleep" option checked. Intermittently, the PC wouldn't wake up from standby. When the mouse was moved or a key depressed, the PC attempted to resume and went dark. It resumed from the hybrid sleep image after pushing the power button.
The problem was caused by a defective motherboard, an ASUS Z87-A. I still have many questions about the defect and the troubleshooting procedure I used. This post is designed to help answer those questions. Skip to the bottom for my questions.
Reliable resume-from-standby was a requirement for this PC because scripts are run nightly to perform various tasks. Whenever resume-from-standby failed, the scripts didn't run.
The system was not overclocked. Processor temperature was steady and acceptable (38°C = 100°F). There were no thermal shutdowns.
BIOS was updated as new versions were released. This had no effect on the problem.
The ASUS user forum for this motherboard had no reports of resume-from-standby or power management problems. A post about this problem received no responses.
There were no BSODs. No error related to this problem was reported in the Event Viewer in the System or Application log. When resume-from-standby failed, SysLog Event ID 1 from Source "Power-Troubleshooter" reported "Wake Source: Unknown" instead of "Device - Intel(R) USB 3.0 eXtensible Host Controller" or the name of the invoked script.
The PC is connected to a UPS; the battery is fresh and resume occurs normally after power is turned off and then back on.
The power supply was replaced with one from a major brand rated with significant reserve capacity -- it had no effect on the problem.
There are 4 identical DDR3 modules that were picked from the qualified vendor list. The resume problem was unaffected if two or three modules were removed or if one or two modules were used that were located in other slots. The RAM passed the Windows Memory Diagnostic test and no errors were reported after MEMTEST86 ran overnight. RAM timing was switched from "Auto" to "XMP" in the BIOS, but this had no effect.
The graphics controller is an embedded Intel HD 4600. Games are not used on this system.
The hard disks (one SSD for the OS and a mechanical drive for data) had no known problems. CHKDSK, run weekly, showed nothing unusual.
Intel Rapid Start was used for several weeks. It had no effect on the problem. The PC would resume from the image on the Smart Response partition whenever resume-from-standby failed. (Intel Rapid Start was subsequently uninstalled.)
Windows 7 is run with a limited-attack-surface strategy. The only third-party services that are running are for the antivirus and an Intel NIC. The antivirus was uninstalled as a test. There was no effect on the problem.
The motherboard was finally replaced with an identical board. This solved the problem.
The resume problem occurred 24 times in 8½ months (18 times in the last 6 months). The minimum time between incidents was 2½ hours; the maximum time was 35 days. There was no common element. Resume could fail at any time of day or night. There were approximately 1150 standby events (SysLog Event ID 42) over the last 6 months. The problem occurred, then, in roughly 1.5% of the resume attempts.
I now know that this problem was caused by a defective motherboard. It just slipped out of warranty.
I have a number of questions:
regards, Andy
The installation had one notable intermittent problem. Standby (aka "suspend" or "sleep") was configured to occur after 30 minutes with the "Allow hybrid sleep" option checked. Intermittently, the PC wouldn't wake up from standby. When the mouse was moved or a key depressed, the PC attempted to resume and went dark. It resumed from the hybrid sleep image after pushing the power button.
The problem was caused by a defective motherboard, an ASUS Z87-A. I still have many questions about the defect and the troubleshooting procedure I used. This post is designed to help answer those questions. Skip to the bottom for my questions.
Reliable resume-from-standby was a requirement for this PC because scripts are run nightly to perform various tasks. Whenever resume-from-standby failed, the scripts didn't run.
The system was not overclocked. Processor temperature was steady and acceptable (38°C = 100°F). There were no thermal shutdowns.
BIOS was updated as new versions were released. This had no effect on the problem.
The ASUS user forum for this motherboard had no reports of resume-from-standby or power management problems. A post about this problem received no responses.
There were no BSODs. No error related to this problem was reported in the Event Viewer in the System or Application log. When resume-from-standby failed, SysLog Event ID 1 from Source "Power-Troubleshooter" reported "Wake Source: Unknown" instead of "Device - Intel(R) USB 3.0 eXtensible Host Controller" or the name of the invoked script.
The PC is connected to a UPS; the battery is fresh and resume occurs normally after power is turned off and then back on.
The power supply was replaced with one from a major brand rated with significant reserve capacity -- it had no effect on the problem.
There are 4 identical DDR3 modules that were picked from the qualified vendor list. The resume problem was unaffected if two or three modules were removed or if one or two modules were used that were located in other slots. The RAM passed the Windows Memory Diagnostic test and no errors were reported after MEMTEST86 ran overnight. RAM timing was switched from "Auto" to "XMP" in the BIOS, but this had no effect.
The graphics controller is an embedded Intel HD 4600. Games are not used on this system.
The hard disks (one SSD for the OS and a mechanical drive for data) had no known problems. CHKDSK, run weekly, showed nothing unusual.
Intel Rapid Start was used for several weeks. It had no effect on the problem. The PC would resume from the image on the Smart Response partition whenever resume-from-standby failed. (Intel Rapid Start was subsequently uninstalled.)
Windows 7 is run with a limited-attack-surface strategy. The only third-party services that are running are for the antivirus and an Intel NIC. The antivirus was uninstalled as a test. There was no effect on the problem.
The motherboard was finally replaced with an identical board. This solved the problem.
The resume problem occurred 24 times in 8½ months (18 times in the last 6 months). The minimum time between incidents was 2½ hours; the maximum time was 35 days. There was no common element. Resume could fail at any time of day or night. There were approximately 1150 standby events (SysLog Event ID 42) over the last 6 months. The problem occurred, then, in roughly 1.5% of the resume attempts.
I now know that this problem was caused by a defective motherboard. It just slipped out of warranty.
I have a number of questions:
- What motherboard component was most likely responsible?
- Is there any way to confirm it?
- What kind of a problem did this component have?
- Would this component have provoked other symptoms that, perhaps, weren't noticed?
- Could this problem have been diagnosed by software?
- What other troubleshooting steps would have been useful?
- Can a defective motherboard that interferes with resume-from-standby be more easily diagnosed without replacing the board?
regards, Andy
