Forum Home
Press F1
Thread ID: 85280	2007-12-05 08:29:00	Upgraded computer, system spontaneously hangs	D. McG (3023)	Press F1

Post ID	Timestamp	Content	User
618149	2007-12-05 08:29:00	I have just performed a major upgrade of my PC (motherboard, CPU, RAM and graphics card) and have noticed that every now and again it freezes. The strange thing is that pressing the Reset button doesn't properly reboot the machine, instead the screen stays black (no signal) and the HDD light stays on without doing anything. Instead, I am forced to power down and restart. My old setup (that worked perfectly): Soltek NV400-L64 mini-ATX motherboard AMD Athlon XP 2600+ (Barton core) 512MB DDR-400 RAM (2x256MB, dual channel) Leadtek GeForce 6600GT 8xAGP video card with 128MB RAM (required additional power via Molex socket), actively cooled Zalman ZM-460APS 460W PSU (Active PFC) Seagate 7200RPM 250GB HDD (IDE) 16x DVD-ROM drive DVD R/RW (Dual layer) 3.5" floppy drive Sound Blaster Live Value PCI soundcard Compro PVR TV tuner My new setup: Asus M2N-SLI Deluxe motherboard AMD Athlon64 X2 4800+ (65W) 2GB DDR2-800 RAM (2x1GB, dual channel) Asus GeForce 8600GT PCIe-16 video card with 512MB RAM, actively cooled (same hardware as before: ) Zalman ZM-460APS 460W PSU (Active PFC) Seagate 7200RPM 250GB HDD (IDE) - with Manhattan IDE to SATA adapter, therefore effectively becoming a SATA drive 16x DVD-ROM drive DVD R/RW (Dual layer) 3.5" floppy drive Sound Blaster Live Value PCI soundcard Compro PVR TV tuner Also connected via USB (and drawing power) are a Gigabyte WLAN adapter and Wacom Bamboo tablet. So what would be causing the computer to hang - inadequate power supply? Faulty motherboard? Faulty CPU? The PC most often hangs when booting Linux (Mepis) from a LiveCD, especially when KDE is loading the desktop; and when playing games such as Far Cry and Crysis in Windows XP. There was one exception - I was browsing the Internet and had music playing from Windows Media Player when it froze. It seems unlikely to be the HDD, as Linux was still hanging before I connected the HDD. It's also unlikely to be the RAM, having run memtest-x86 on the live CD. Temperatures seem quite normal - according to Aida32, Motherboard reads 40°C, Aux is 33°C, and the HDD is 37°C (strangely missing the CPU reading). I have also applied the latest BIOS update (ASUS M2N-SLI DELUXE ACPI BIOS Revision 1302, 29 Oct 2007), but this doesn't seem to have achieved a thing. Normally I'd be inclined to believe it was the PSU not meeting the power demands of the hardware, but surely this shouldn't happen when loading KDE from a Linux live CD? And surely rebooting via the reset button should not hang with a blank screen (requiring a complete power-down)? Any help much appreciated. -D. McG	D. McG (3023)
618150	2007-12-05 08:38:00	Did you reinstall XP when you changed the mobo or do a repair?? Whats event viewer say?? About the time it froze?? Did you disable the onboard sound?? It sounds like the reset wires arent connected to the right headers maybe?? If reset doesnt work. Look in device manager, is anything conflicting with something else?? Anything here show a X or !	Speedy Gonzales (78)
618151	2007-12-05 09:08:00	Hi Speedy Gonzales, I kept XP as it was (but uninstalled the graphics drivers first, considering I was upgrading the video card). XP handled the re-detection of the hardware fine, and installed the motherboard drivers off the CD when it asked for them. The event viewer doesn't show any warnings or errors since the upgrade was performed (02/12). Onboard sound is disabled - yes I had thought of that. ;) The Reset button is connected okay, just tested it and it worked fine. The only thing in Device Manager with an exclamation mark is the scanner - just because I haven't yet needed to install the drivers since migrating to XP a year ago. This was there before Sunday's upgrade. I also disabled the nForce networking controllers #3 and #4 to remove some of the clutter from the system tray (since I'm using wireless LAN). However, this doesn't explain why I'd be getting lock-ups when loading KDE on a Linux LiveCD (even with the "noapic" boot option set). Thanks for the speedy response!	D. McG (3023)
618152	2007-12-05 09:31:00	See if putting this Seagate 7200RPM 250GB HDD (IDE) - with Manhattan IDE to SATA adapter, therefore effectively becoming a SATA drive Back on a IDE conection by itself fixes something? Thats if there are any spare ide connections. It could be this hdd whose light is staying on?	Speedy Gonzales (78)
618153	2007-12-06 10:41:00	Thanks for your suggestion, though it's funny you should say that - the motherboard only has one IDE channel, occupied by the two optical drives. I've seen that it is a bad idea to put an IDE hard drive on the same channel as an ATAPI optical drive due to the differing communication protocols. So that's why I have the IDE to SATA converter. Also, disconnecting the optical drives and connecting the HDD to the IDE channel is out of the question as I need to be able to load the Linux live CDs to be able to test the stability - remember the hanging isn't just an issue under Windows. And I also had this issue when the HDD was disconnected entirely - leaving just the optical drives to test booting Linux (let alone KDE). Anyway, for an update: Tonight I installed the NVidia NTune software to perform stress tests. The application crashed on many occasions within the first three minutes of the 10 minute test. GPU temperature hovered around 50°C, while Aida32 reported some funny temperatures (with the CPU reading randomly disappearing for variable lengths of time): Temperatures: Motherboard 40 °C (104 °F) CPU 29 °C (84 °F) Aux 43 °C (109 °F) Cooling Fans: CPU 3041 RPM Chassis 4592 RPM Voltage Values: CPU Core 1.17 V +2.5 V 3.15 V +5 V 4.73 V +12 V 11.52 V +5 V Standby 4.73 V VBAT Battery 2.96 V and even weirder when the stress test wasn't running: Temperatures: Motherboard 40 °C (104 °F) CPU 2 °C (36 °F) Aux 39 °C (102 °F) Cooling Fans: CPU 3054 RPM Chassis 4720 RPM Voltage Values: CPU Core 1.20 V +2.5 V 3.15 V +5 V 4.73 V +12 V 11.58 V +5 V Standby 4.73 V VBAT Battery 2.96 V CPU running at two degrees??? That would be an impressive heatsink, considering ambient temperatures are about 21°C. I also had some strange problems with my WLAN disconnecting, prompting me to restart the configuration/connection utility on two occasions tonight. Tonight I've had one periodic freeze while doing this monitoring while playing an MP3 (causing sound segments to repeat), but this only lasted about five seconds and didn't require a reboot (er, power-down and restart). So far, writing this response has been okay! So, back to my original question: Is this instability in Linux and Windows likely to be caused by a faulty motherboard, faulty CPU, or inadequate power supply? I'm starting to wonder if it is indeed the motherboard, given the range of some of the comments here: www.newegg.com (www.newegg.com) Here's some more info on the motherboard: www.asus.com One customer review on NewEgg also recommended having at least a 600W PSU, considering that it is a SLI board. Would this still hold true even though I'm using only one video card?	D. McG (3023)
618154	2007-12-06 21:46:00	Hi Speedy Gonzales, I kept XP as it was (but uninstalled the graphics drivers first, considering I was upgrading the video card). XP handled the re-detection of the hardware fine, and installed the motherboard drivers off the CD when it asked for them. The event viewer doesn't show any warnings or errors since the upgrade was performed (02/12). Onboard sound is disabled - yes I had thought of that. ;) The Reset button is connected okay, just tested it and it worked fine. The only thing in Device Manager with an exclamation mark is the scanner - just because I haven't yet needed to install the drivers since migrating to XP a year ago. This was there before Sunday's upgrade. I also disabled the nForce networking controllers #3 and #4 to remove some of the clutter from the system tray (since I'm using wireless LAN). However, this doesn't explain why I'd be getting lock-ups when loading KDE on a Linux LiveCD (even with the "noapic" boot option set). Thanks for the speedy response! OKAY, you used the same XP build as the old mobo etc....but did you prepare the build...you need to change the ide controllers to generic etc and all the bus drivers to generic before swapping mobo...if you havent, i would be looking to start again with a new build....or it will never get rid of the stutter\freeze. Oh, and freezing pc usually come down to RAM, or software drivers. So I would try the other 2 slots for the RAM, then a XP rebuild. I'm surprised XP didnt ask to be re-activated!?	SolMiester (139)
618155	2007-12-06 21:50:00	Thanks for your suggestion, though it's funny you should say that - the motherboard only has one IDE channel, occupied by the two optical drives. I've seen that it is a bad idea to put an IDE hard drive on the same channel as an ATAPI optical drive due to the differing communication protocols. So that's why I have the IDE to SATA converter. No mate, the reason for separating optical drives (usually writer) and IDE HDD is because a writer should always be a MASTER, not a slave or you get gaps in data transfer to the writer while recording which could make the cd a table coaster, while a HDD usually wants to be a MASTER too, also because of data access. Nothing to do with communication protocals, they are the same, they are on the same bus interface.	SolMiester (139)
618156	2007-12-06 23:21:00	No mate, the reason for separating optical drives (usually writer) and IDE HDD is because a writer should always be a MASTER, not a slave or you get gaps in data transfer to the writer while recording which could make the cd a table coaster, while a HDD usually wants to be a MASTER too, also because of data access . Nothing to do with communication protocals, they are the same, they are on the same bus interface . nothing to do with master/slave . (with exception of a few buggy firmware which are temperamental with what goes where) . you get gaps in data transfer because it only uses one device at a time on the ide, ie it reads HD, stores it in ram, then writes it to cd drive . so the data flow is very stop start which is where problems occur . tho not really an issue these days except being slow . usually the biggest problem is the data speed modes are different for cd and hd . some pc's will make the hd go at the cd speed which makes the whole pc go at a crawl .	tweak'e (69)
618157	2007-12-07 00:07:00	nothing to do with master/slave . (with exception of a few buggy firmware which are temperamental with what goes where) . you get gaps in data transfer because it only uses one device at a time on the ide, ie it reads HD, stores it in ram, then writes it to cd drive . so the data flow is very stop start which is where problems occur . tho not really an issue these days except being slow . usually the biggest problem is the data speed modes are different for cd and hd . some pc's will make the hd go at the cd speed which makes the whole pc go at a crawl . Which is exactly what I said . . . . . . Master\Slave separates the channels, older cutting devices needed master, however the main concern was continuous data to the writer . . . . . . which is what I said . . . . .	SolMiester (139)
618158	2007-12-07 01:44:00	nothing since p1/2 era needs to be on master. ditto for to constant stream to writer. not really a problem these days.	tweak'e (69)
1 2