Kitz Forum
Computer Software => General software => Topic started by: renluop on February 26, 2012, 07:56:51 PM
-
Recently a number of unplanned restarts have occurred on my computer; black screen and immediate reboot without BSOD, or router disconnection
What do you think the cause may be? Might it be somehow connected to my other new thread on uPnP?
Thanks!
-
What do you think the cause may be?
It could possibly be due to the power supply starting to fail. :-\
Might it be somehow connected to my other new thread on uPnP?
Highly unlikely. :no:
-
Does the operating system have some sort of system logging facility? Perhaps that could reveal the cause of the reboot? Is it preceded by a serious crash?
-
There is an Event Viewer that covers system events amongst others. What sort of entry should I be looking for in its logs?
No serious crashes; just a click, black screen, reboot.
Re PSU: computer was new from Dell 04/2010.
-
Apart from the power supply, the other common cause of abrupt reboots is defective memory. It would be a good idea to run a memory test, but I don't know if Windows provides such a capability.
-
BTW there is a substation a couple of doors up, but as there is no corresponding dip in house lights etc I have discounted that.
-
There is an Event Viewer that covers system events amongst others. What sort of entry should I be looking for in its logs?
No serious crashes; just a click, black screen, reboot.
Re PSU: computer was new from Dell 04/2010.
Most PSUs are not made very well these days. They do just fail for no obvious reason. Do you have a spare one you could swap in? Does the system reboot under any particular conditions? A heavy power load, perhaps? Do you have a tool that displays CPU load? There was/is a tool built-in to Windows for this. Not sure if it's any good.
Are there any speaker beeps before the reboot? Stabbing in the dark, but could the CPU be overheating. That can cause the system to shut-down. You usually get warning beeps first though.
As Eric has recommended.. maybe check the RAM.. There's a free tool to do this called memtest86. [1] Be warned, the most extensive memtest86 RAM tests can take days!
People often use a prime number-finding tool called Prime95 aka MPrime for 'torture testing' a PC.. [2] It purportedly tests the PSU, the CPU, the RAM and the Northbridge.
There are several Live CDs which include a range of stress-testing tools. You just burn the CD image and boot straight from it. The advantage in using a Live CD is it eliminates any problems in the Windows operating system.
cheers, a
[1] http://www.memtest.org/
[2] http://en.wikipedia.org/wiki/Prime95#Use_for_stress_testing
[3] http://www.overclockers.com/forums/showthread.php?t=486495
-
Ran test in Win7 and report came back "no fault found".
Also ran Microsoft's Fixit for power problems and nothing was found. I have a suspicion that the tool may be inadequate other than for the simple matters. Is there any other simple and two left feet way to test the PSU.
Having now consumed my tea I have used SisoftSandra and this has come up
Warning W2519 - One of the mainboard fans has failed or is spinning too slowly. This should appear only if the mainboard has a temperature-monitoring sensor installed and the fans have tachometers. If the fan is speed controlled, this may be normal.
Fix: If the fan is not speed controller, power down the system immediately. Open the case and check that all the fans are working and check whether components are unusually hot inside. Make sure the components have good all-round ventilation.
-
Is there any other simple and two left feet way to test the PSU.
Not that I'm aware of, at least not without test equipment. The PSU fault, if there is one, might only occur at higher loads. You would have to simulate those conditions before the fault manifested. The ATX PSUs also have multiple rails providing various voltages to the motherboard, drives, etc. Each of those rails would have to be stress tested individually on a test bench. It's not worth the bother. PC PSUs are basically disposable items.
Before you consider replacing the PSU, you might want to check for compatibility: some Dell motherboards/cases have used non-standard PSU connectors or sizes.
Having now consumed my tea I have used SisoftSandra and this has come up -- Warning W2519 - One of the mainboard fans has failed or is spinning too slowly. This should appear only if the mainboard has a temperature-monitoring sensor installed and the fans have tachometers. If the fan is speed controlled, this may be normal.
Fix: If the fan is not speed controller, power down the system immediately. Open the case and check that all the fans are working and check whether components are unusually hot inside. Make sure the components have good all-round ventilation.
Hmm. not sure that warning is very trustworthy. The fan controllers on a motherboard are usually generic devices. The controllers can support multiple fans than might not actually be present on the motherboard. It doesn't bode well that the SiSoftSandra tool cannot distinguish between a failed fan (or a missing fan) and a fan that has simply slowed.
It's not much use for Windows, but the de-facto Linux software for fan control and monitoring is called lmsensors. http://lm-sensors.org/ It will be a standard tool included on those stress-testing Live CDs linked above. Worth a look. This is the sort of thing it reports:
w83627ehf-isa-0290
Adapter: ISA adapter
VCore: +1.33 V (min = +0.00 V, max = +1.74 V)
in1: +11.88 V (min = +10.08 V, max = +13.46 V)
AVCC: +3.30 V (min = +4.08 V, max = +2.93 V) ALARM
3VCC: +3.31 V (min = +4.08 V, max = +4.08 V) ALARM
in4: +1.64 V (min = +2.04 V, max = +2.04 V) ALARM
in5: +1.65 V (min = +1.52 V, max = +1.53 V) ALARM
in6: +0.23 V (min = +6.53 V, max = +6.50 V) ALARM
VSB: +3.33 V (min = +4.08 V, max = +4.08 V) ALARM
VBAT: +0.00 V (min = +4.08 V, max = +4.06 V) ALARM
in9: +0.07 V (min = +2.04 V, max = +1.53 V) ALARM
Case Fan: 0 RPM (min = 55 RPM, div = 128) ALARM
CPU Fan: 1864 RPM (min = 0 RPM, div = 4)
Aux Fan: 0 RPM (min = 0 RPM, div = 128)
fan5: 0 RPM (min = 0 RPM, div = 8)
Sys Temp: +36°C (high = -1°C, hyst = -1°C) ALARM
CPU Temp: +30.0°C (high = +80.0°C, hyst = +75.0°C)
AUX Temp: +48.0°C (high = +80.0°C, hyst = +75.0°C)
(as an aside)
Silent PCs are interesting. These have sexed-up passive cooling systems that have to be closely monitored, especially during the building stage. There are quite a few tricks to keeping everything cool. For example, the CPU core voltage (VCore) can often be dropped below the voltage specified by the chip maker. The specified voltage is only a guide. A lower core voltage can make the CPU run cooler, etc.. However, drop the voltage too low and the CPU becomes unstable. Also one of the symptoms of a failing PSU..
cheers, a
-
Silent PCs well that's left me speechless and more ;D
Got this from SSS just now.
Temperature Sensors
Board 55deg; CPU 37 deg
Cooling Devices
Cgassis Fan496 rpm; CPU Fan 848 rpm
SSS has ! against chassis fan
-
Silent PCs well that's left me speechless and more ;D
Got this from SSS just now.
Temperature Sensors
Board 55deg; CPU 37 deg
Cooling Devices
Cgassis Fan496 rpm; CPU Fan 848 rpm
SSS has ! against chassis fan
What model is the CPU? 37deg is not very hot. 95deg is not uncommon.
The chassis fan is helpful but it's non essential. Often the GPU and the northbridge are actively cooled with fans of their own. The northbridge fans are usually very small (40mm) and poor quality. They fail fairly regularly, and cause stability problems. When they start to fail they can make a horrible screeching noise. At that stage, a sudden silence is not good news!
Are you comfortable opening the PC to visually check things?
cheers, a
-
I will at a pinch open up things, but it's certainly not in the centre of my comfort zone. Mental wobblies and ageing creeks! ;D
Intel(R) Core(TM)2 Duo CPU E7500 @2.93Hz (CPU:0) is what Dell's hardware checker is saying.