Computer goes into 5-second reboot cycle after sleep mode

K

kevineugenius

[H]ard|Gawd
Joined
Dec 9, 2006
Messages
1,415
Seemingly at random, though I think I updated my Catalyst and drivers at roughly the same time, my computer will no longer wake up out of sleep mode. It displays the motherboard logo screen for 5 seconds, sometimes even gets clear to 'Resuming Windows' ever so briefly then the computer turns off. As if by magic, it then turns itself back on after 5 seconds and repeats this cycle until I somehow tell it to discard the remembered settings and boot fresh. I don't really think it's a hardware issue at all, but I'm open to suggestions as to where I should look for a solution before I reinstall Windows.

Gigabyte Z68 UD4
2600k
2x4gb G.Skill Ripjaws 1333
2x XFX 6870 (removed one thinking maybe power supply problems, didn't make a difference)
OS on 300gb Raptor
700w Hipro PSU
Win7 Ult x64
 
Well if you noticed this issue right around the time you updated your drivers, then my suggestion would be to revert back to the older drivers to see if the problem persists.
 
Haven't been able to do extensive testing yet, but I started getting BSODs in Dungeon Siege 3 also, so I did what I thought was uninstalling all AMD stuff from the system. The utility only removed the actual software, not the driver, so my Eyefinity is still working but there's no CCC to turn it off or back on again. BSODs are gone and it's recovered from sleep twice so far.

If this does end up being a fix... the question becomes: what the heck is CCC doing that would cause the system to not be able to boot from a sleep state? Really not instilling a lot of confidence in their programming ability.
 
Nope, problem persists. I don't think it can possibly be software related because I was mashing on the F12 key for 'boot menu' and it doesn't even reach that point before resetting most times. I removed one GPU, added a 2nd power supply for the remaining GPU, removed 1 RAM stick, put RAM stick in a different slot, basically got the computer down to barebones and it still does this. If I can break the cycle and get it to boot, it works fine. What could be the cause of this?
 
I would recommend you to reinstall system. To exclude hardware problem with booting try to run some live CD os, where you also can try sleep mode. If that will pass probably your windows installation has gone wild.
 
Update: running with 1 RAM stick in slot 2 with no hangs so far. 1 RAM stick in slot 1 did have hangs (same stick). Could the entire problem be/have been that I put my RAM in the channel B slots by mistake?
 
I don't know which slots you used but gigabyte recommends using slot 1 which is closest slot to edge of board and slot 2 which is the second slot from the cpu socket for dual channel.
 
Yeah, those are the slots I used. It did power cycle on me again last night so I swapped the single RAM stick to the other one just for giggles. Will swap the GPU next. Then pull the mobo from the case and desk run it... then stomp on my foot because it'll feel better than the headache... ugh. Was hoping this 'symptom' was related to a particular problem but I guess it isn't. Also going to remove all fan headers from the mobo and run them straight off the PSU. A zillion steps I can think of at an average of 4 days per step; good thing my warranties are long.
 
Updating just so I don't have to remember all this stuff myself. Problem seems to happen more often with 2 RAM sticks installed, location doesn't matter, each RAM stick has the problem when installed by itself again location not mattering. Talked with Gigabyte who recommended enabling XMP which is not in the BIOS anywhere and they then told me if it isn't in the BIOS the RAM doesn't support it. Pre-sale I talked to G.Skill and they said it did, I dunno if that matters at all but it struck me as odd. Per Gigabyte suggestion now running with memory in 1066 mode instead of 1333.
 
First boot after F8 BIOS "successfully" installed, computer went into endless reboot loop. I should mention that when the PC is 'reset' it powers down fully for x seconds, not a quick 1/4 second click like most computers. So, it could be an endless reset loop, not an endless power off/on loop...

Anyway, that's the first time it has ever had the issue when not recovering from sleep mode, which shifts my suspicion slightly away from the RAM being faulty.

In F8, reset RAM to 1066.
 
Try clocking the RAM speed down. Or bump the voltage for them to 1.55, or 1.5. I'd try 1.55 first. I'd lean towards memory personally just based on what you've said. Check to see if your BIOS has set the timings correctly too, and if not, manually set them. Maybe bump up the voltage to the CPU as well, but I'd try the memory first.
 
I've lowered the RAM speed from 1333 to 1066 and the timings are correct at 9-9-9-24 and they've been at 1.5v. It's "guaranteed" to run on this motherboard at 1333@1.5 and the problem persists whether using stick 1, stick 2, or both sticks and in various slots on the motherboard. While it's possible that both sticks are faulty, it's not super likely is it? On the other hand, last time I had weird inexplicable problems it was the RAM's fault so I can't ignore that.
 
Just opened a Jones Soda and under the lid it says "There is new hope for projects you had almost given up on." WOOT!! :p

2nd occurrence of the boot recycle from power off mode since updating to F8.

Reinstalled latest CCC and almost immediately got a driver crash followed by BSOD. Uninstalled 11.8, now trying 10.12.

Later that day, reverted to F6. First reboot after F6 did a stupid #$@%ing power cycle again, which never happened on F6 before. Going to RMA motherboard tomorrow.

I take it as proof that no software is causing this particular problem, the BSODs and CCC being loony are a separate issue, that the fresh load BIOS has the wrong HD selected and cannot boot to Windows at all, yet the problem persisted. Woohoo, some data... Now I get to pay UPS and restocking fees a couple hundred bucks over the next month to figure out the rest of this.
 
Last edited:
10.12 CCC fixed the BSOD it seems, it was 100% crash rate while playing Dungeon Siege 3 and it didn't crash after 10.12 so that's a plus. Irrelevant to the larger problem though.

As usual, troubleshooting a problem that only shows itself every few days has taken so long that I can't do a newegg advanced replacement anymore, so probably have to tear the computer down and ship crap and be without a working system at all for two weeks for each troubleshooting step from here on out... Argh.
 
Last edited:
I never did RMA motherboard, but I haven't had the issue coming out of sleep or out of a full power off in ~5 days. They've been long time periods of off/sleep, so I need to try with some 15 minute naps and see if I can replicate the problem.
 
Issues persist today both from sleep mode with 2 vid cards and from power off mode with first vid card removed and 2nd vid card in the primary slot by itself. It seems to do the power off mode recycle every time something gets changed and usually only happens once, so I can't say for certain the vid cards are 100% excluded from scrutiny just yet.
 
It seems that the problem only happens about 10% of the time with one mem stick installed and more like 50% of the time with both sticks. I have lowered the mem speed in both modes to see if that's the issue but with no effect. Does this point toward it being a motherboard issue?
 
It could also be a CPU issue since the memory controller is now on the CPU rather than the motherboard.
 
That's true. It occurred to me that since Windows is what is writing the sleep state to the memory that maybe I should to a fresh install just to eliminate that variable completely. I did a full PC rebuild basically without reinstalling Windows and was surprised at how well it worked. Maybe it didn't.

Is there any way to test a CPU mem controller?
 
memtest would have tested both the ram and the controller. how long did you run it for?
 
kevineugenius said:
Is there any way to test a CPU mem controller?
Click to expand...

Like enginurd said, Memtest would be the test for your RAM and the memory controller itself. The way to isolate would be if you had errors with your RAM, and then swapped in known working RAM and rerun Memtest and had errors again, you would have good cause to suspect the memory controller.

And on a sidenote, I always run Memtest at least 3 iterations, usually overnight. I've had some bad RAM sticks that will test good for one loop or sometimes two, then all of a sudden there will be a flurry of errors on test three or four.
 
heh, i've had ram that will pass memtest for 24+hrs but fail in the 36+hr range - this is most likely due to heat and/or instability on the mobo/chipset, rather than the RAM itself; but since RAM has a lifetime warranty, I get it replaced if it ever fails on stock/advertised settings.
 
I am so pro. Finally remembered to run memtest overnight... forgot to put the 2nd ram stick in... so it ran 20 passes in single channel, proving nothing. D'oh.
 
Hasn't actually done the boot loop in quite a while. Generally while in dual channel it would be like every other start attempt. Yesterday, however, a new and exciting symptom occurred.

I'd hit sleep. It would do its Windowsey stuff and shut everything down. Screens go black and say no input, USB devices all power off. Then there's a click and it comes back out of sleep. Haven't tried it today but it did it 100% of the time yesterday so I did a full shut down last night. Weird.
 
Original problem persists, this time after a restart. I guess sleep mode isn't involved at all, which means the RAM is probably not involved at all. Guess I'll RMA the motherboard in December when winter break starts, then Giga will test it and say "I don't see any problem with this" and send it back to me. Then I'll be mad. Etc.
 
Well, I'm trying to replicate the problem as best I can, I thought putting in the other memory stick would set it off like it always used to, but so far there's been almost no problems. Would running memtest for a long while give the system like a burn-in or something?
 
Yah, but I dont think it's gonna help. Sorry, I'm out of ideas right now.
 
What I meant was I already ran memtest and maybe it did a burn-in, which is why I'm getting the hugely noticeable increase in stability. I dunno.
 
Hasn't happened in a long while except the first reboot after running memtest. I have every single component installed. Maybe it fixed itself, maybe it'll come back. Idk
 
Yesterday and today it came out of sleep mode without me asking it to... Nothing on the keyboard, just randomly decided to power up.
 
You must log in or register to reply here.
Back
Top