10-04-2013 08:35 PM
first of all, a want to apologize for the bad english. It's obviously not mine native language, and all mine knowing is mostly from reading some documentation etc. Never had to write something on it, but not this time. So i will try to use simple words, hope its will be understandable and you'll forgive me
Ok, mine problem is with notebook HP ENVY 14 Spectre 3100er. Its unexpectedly shuts down and restarts with no reason. Its not looks like just reboots, but exactly like power-off and immediately on. In Windows Event viewer appears event: source:Kernel-Power, event id:41 "The system has rebooted without cleanly shutting down first. This error could be caused if the system stopped responding, crashed, or lost power unexpectedly."
This can occur for example once in 48hours(all that time notebook is on) - it is almost 100% guaranteed there will be shuts down during 48hours(but with important correction - see below) or several times in just few hours - rare.
There is no any minidump files, no BSOD. In Startup and Recovery settings option Automatically restart turned off.
There is a list of mine researches and what have i done so far:
1. Restoring system to factory defaults(F11)
1.1 I did not installed any windows updates or drivers updates first. Just for testing. Leaving its working with no apps running(empty desktop) - shuts down and restarts in approximately 24hours.
1.2 Installed windows updates, all drivers updates from hp site. Almost the exact same picture, exept i not always just leave it working with no apps running, i could work for example with internet browser, remote desctop connection, some office apps, but no heavy load for sure.
1.3 Installed all drivers updates from intel(there are newer versions than on hp site): Graphics driver, WiFi driver. No changes.
2. Updated BIOS.
3. Running all diagnostic tests in BIOS(F2). All components test separately and system extensive test. All passed.
Now its more interesting...
4. I replaced original memory module Samsung 4GB DDR3-1600MHz PC12800(m471b5273ch0-ck0) with Kingston 8GB DDR3-1333MHz PC10600(KVR1333D3S9/8G). I describe events in sequence, so its not new memory module was source of problem. Problem was before. I did it for 2 reasons: 1. I had suspicions on bad memory module or just frequency incompatibility(this was not confirmed, but now i know its not memory issue, though ) 2. Just wanted to upgrade memory...
To replace the memory, I had to disassemble the laptop almost completely. It took me almost 6 hours using this document http://bizsupport2.austin.hp.com/bc/docs/support/S
5. There is absolutely no dust in fan system. I had the opportunity to see this on step 4.
6. After memory replacement i launched extensive system test from BIOS, and left it work exactly for 106hours(4 days 10 hours 41 min 37 sec, 103 Completed iterations). How i mentioned before it have never stay on more than 48 hours in windows without self shut down...
7. Boot in windows, disabled network adapters and run Prime95 blend. It has worked slightly more than 96 hours. I'm just taking 96 hours, as it is twice more than he normally works, its just a guess. During stress testing temp on cores did not exceed 66 degrees(with one exeption - on that at the end). I used SpeedFan with event logging. So i thik it is safe to say it is not overheating issue, CPU issue or memory issue...
8. I run video stress testing, again for 4 days(96 hours), so laptop has working now more than a week without shutdowns. Used FurMark 1.11(fluffy donut ).
9. Just in approximately 15 minutes after i stopped FurMark its shuts down with no apps running, no netwoks adapters enabled, on empty desktop. I just planned what tests to do next...
10. I reinstalled windows 7 from completely different image, not using system restoring function, just from mine flash drive. I installed Windows 7x64 Ultimate SP1, originaly it was Win 7x64 Home. In windows Event log first system start was 02.10.2013 0:24:03, first kernel-power error(shutdown) was already 02.10.2013 21:37:04. At this time, i've just turned on some streaming video, turned sound off and left it working. No any stress tests. Btw its looks like i've completely destroyed original system restore function, but its a different story now..
So now i have exhausted all my guesses. its look like he shuts down only when there is no heavy load, just on internet browsing, rdp woking etc, yotube video or just empty descktop. Stress testing for a week passed just fine exept end of step 9. When i run some games its never shuts down. Mine last suspicion is and it is mine main question to the forum: in SpeedFan Temp2 sensor shows always 75 degrees, does not depend on load, its 75 from system starts and 75 during all stress tests(i will attach image). Perhaps broken sensor? Could it be a cause? But again it was not shuting down during stress tests. Not looks like overheating issue..
Do not ask me why I did not resort to the service center. Was thinking to solve the problem myself. Now its most likely i broke warranty terms, but..hell with it
Any advises with shutdown issue will be much appreciated. Sorry for bad english again, hope its at least understandable...
10-07-2013 11:55 AM
Hi dmr4lx. First off let me say that your English is very good, and you seem very technical as well.
Here is the document you have went through already I suspect. (read through it to make sure you have done all of the steps).
So you did mention overheating and that last screen shot looks pretty high. Here is a general purpose document on heat just to make sure you have checked everything: http://h10025.www1.hp.com/ewfrf/wc/document?docnam
If everything else checks out, I suspect this could be some combination of a power or battery issue. This document is specific to battery not holding a charge, but go through the steps to calibrate your battery and ensure you have the proper power adapter.
10-08-2013 07:32 PM
Thanks for reply. I will carefully study all links and will surely report about results. And any info about "temp2" sensor in SpeedFan(Which always shows 75 degrees)?
10-09-2013 09:32 AM
Yes dmr4lx, I suspect it may be a misreading if it never changes under any circumstance. It may be an available port for a probe that was not even connected, and perhaps that is a default read.