| View previous topic :: View next topic |
| Author |
Message |
dskeeles
Joined: 03 Feb 2005
Posts: 26
|
 Posted: Sat Mar 01, 2008 9:15 am Post subject: Self-build: Cry for help |
 |
|
Hi Folks,
I've been quiet for a few weeks since posting about getting a new motherboard - the reason being that I then managed to find one, but since haven't been able to get it working.
OK - this is probably slightly OT, but I've been acquiring all the parts based on discussions here, so I hope someone can help me through to the end! I've tried every single test and combination I can think of, but nothing shows any single point of failure, and as a whole it still doesn't work!
Specs
Basically, I'm keeping my Silverstone LC11 case, but upgrading to a Gigabyte GA-G33M-S2H mobo with E6750 CPU and 2Gb Crucial DDR2-667 RAM - this is from an AOpen i855GMEm-LFS, so it's complete new internals.
Symtoms
When I boot from CD, for example, the boot sequence starts, and then hangs: with a Ubuntu Live CD I get to "Decompressing Kernel" after the start-up screen. With Windows, it just manages to get into "Setup is inspecting your computer's hardware configuration", and then blacks out and either restarts, or says txtsetup.sif is corrupt or missing.
However, if I boot something small (eg. Damn Small Linux, Norton Ghost Startup CD, BootItNG), then it starts up all the way into that OS.
IDE Channel
First, I assumed it was the IDE channel; so I changed the cable, dis/connected the IDE HDD, DVD; changed Master/Slave/CS - no difference. Also made sure AHCI was disabled, and that's OK too.
Then I tried another boot device: an external USB CD-ROM, which did get to a different stage on the Windows XP installer, but then bluescreened on "Starting Windows NT".
I've also tried booting from a USB stick formatted using HP's bootable USB-as-floppy utility, manually running WinNT.exe, and with that it went all the way into a Windows install - but tried to install it on the USB stick!
I also tested the IDE channel by doing a test copy of an IDE HDD partition using BootItNG off CD, which worked and verified with no errors - implying the IDE channel is reliable!
Memory
Then both Gigabyte support and a friend suggested the memory, so I ran MemTest86 on it. It passed with no errors. The specs are as suggested by Crucial for the -DS2R.
Power
I also suspected a PSU shortage (it's a 240W PSU), so I ran a power meter at the socket: it shows 80W consumption at worst.
Mobo
Finally - I replaced the motherboard with a GA-G33H-DS2R, and that gives exactly the same errors - so it can't be a motherboard fault!
Other
All other devices: SATA drives, front audio, USB/Firewire headers, PCI cards - are all disconnected. All that's in are power, fans, the boot device, and a USB keyboard.
CPU
I guess the last thing to test is the CPU itself - assuming that there's some spurious instruction that the smaller bootable CDs and memtest didn't use, but Windows/Ubuntu do...!
In Conclusion
As you can see - I've tried everything I can think of! If anyone has any suggestions, it would be a great help!
Damian |
|
| Back to top |
|
 |
|
|
dskeeles
Joined: 03 Feb 2005
Posts: 26
|
| Posted: Sat Mar 01, 2008 10:08 am Post subject: Re: Self-build: Cry for help |
|
|
| dskeeles wrote: |
CPU
I guess the last thing to test is the CPU itself - assuming that there's some spurious instruction that the smaller bootable CDs and memtest didn't use, but Windows/Ubuntu do...! |
I've just run the various CPU stress tests from the Ultimate Boot CD, and they all also pass with flying colours (well, 1 million iterations)....
Still, I'd appreciate any pointers! I'll post my otherwise-unused GA-G33M-DS2R (still a decent motherboard) to whoever can come up with the solution that leads to a fix!
Damian |
|
| Back to top |
|
|
psiu
Joined: 23 Aug 2005
Posts: 1201
Location: Plymouth, MI
|
| Posted: Sat Mar 01, 2008 10:51 am Post subject: |
|
|
In order:
Update BIOS on motherboard?
Doublecheck all BIOS settings?
Different IDE cables? check
Just 1 stick o'ram?
Loosen up RAM timings?
Different HD?
HD has jumper in MA or CS? check
Try a different PSU?
Try different video?
Harsh language?
Threats involving a screwdriver, soldering iron, and cola?
edit: yay for reading...I added check to what you already covered...
edit2: and...that sucks. I feel your pain...especially when it turns out to be something little...all auxiliary power plugs are plugged in on the motherboard right? |
|
| Back to top |
|
|
mike42
Joined: 18 Feb 2008
Posts: 3
Location: UK
|
| Posted: Sat Mar 01, 2008 11:52 am Post subject: |
|
|
Check voltages on the PSU.
(It happened to me - if the PSU's 5V line sags, the first thing to complain is the HDD.)
BIOS settings for IDE channels - detect and set manually.
(If this is the problem, then you would only see errors before the OS loads and starts
using its own drivers and IDE settings.) |
|
| Back to top |
|
|
ntavlas
Joined: 16 Jul 2007
Posts: 614
Location: Greece
|
| Posted: Sat Mar 01, 2008 2:11 pm Post subject: |
|
|
If your hard disk and dvd are on the same ide channel the jumpers should be: master on hdd, slave on dvd.
Another possibility is that the 5v line sags indeed, but it`s hard to imagine this happening in a modern system.
_________________
My PCs: Workstation : HTPC in a CD player case : Custom Nettop |
|
| Back to top |
|
|
dskeeles
Joined: 03 Feb 2005
Posts: 26
|
| Posted: Sat Mar 01, 2008 4:03 pm Post subject: |
|
|
Sorry for the delay - thought I'd try some more playing with BIOS settings and un/plugging bits before I replied.....
Update BIOS on motherboard?
= Yep, currently R3
Doublecheck all BIOS settings?
= I have several times... there could always be something, but it's a fairly vanilla system and should also run with defaults or Fail Safe
Just 1 stick o'ram?
= I do - 1x2Gb. But it'll work, it just runs in single channel mode, right?
Loosen up RAM timings?
= Already 5-5-5-15, but I tried 7-7-7-20 - no change
Different HD?
= Yep - I've tried booting off CD with it disconnected - I have 2 SATA drives, but without OS; no change
HD has jumper in MA or CS? check
= (Also see next post)
Try a different PSU?
= Don't have one 
Try different video?
= On-board!
Harsh language?
= Already exhausted 
Threats involving a screwdriver, soldering iron, and cola?
= All at the same time?
all auxiliary power plugs are plugged in on the motherboard right?
= Erm.. the main ATX cable, and the processor cable? Yep!
I can't make sense of it either. It's either something blindingly obvious, or something deep in the space-time continuum.... |
|
| Back to top |
|
|
dskeeles
Joined: 03 Feb 2005
Posts: 26
|
| Posted: Sat Mar 01, 2008 4:12 pm Post subject: |
|
|
| mike42 wrote: | Check voltages on the PSU.
(It happened to me - if the PSU's 5V line sags, the first thing to complain is the HDD.) |
I'll get the multimeter out..
| mike42 wrote: |
BIOS settings for IDE channels - detect and set manually.
(If this is the problem, then you would only see errors before the OS loads and starts
using its own drivers and IDE settings.) |
Only IDE BIOS settings are for the IDE Access Mode (Auto/Large); normally Auto, but I tried Large, with no improvement I guess since once I boot into Ubuntu or Windows, it might override with its own OS drivers, but again, it should be fairly standard. |
|
| Back to top |
|
|
dskeeles
Joined: 03 Feb 2005
Posts: 26
|
| Posted: Sat Mar 01, 2008 4:14 pm Post subject: |
|
|
| ntavlas wrote: | If your hard disk and dvd are on the same ide channel the jumpers should be: master on hdd, slave on dvd.
Another possibility is that the 5v line sags indeed, but it`s hard to imagine this happening in a modern system. |
That's a good point; I have admittedly always had the Optical drive on master and HDD on slave because I wanted it to take priority over the HDD (when playing DVDs). I swapped them round anyway, to discount this, with no change. |
|
| Back to top |
|
|
line
*Lifetime Patron*
Joined: 28 Sep 2005
Posts: 333
Location: Israel
|
| Posted: Sat Mar 01, 2008 4:41 pm Post subject: |
|
|
| Quote: | | I do - 1x2Gb. But it'll work, it just runs in single channel mode, right? |
Yes.
| Quote: | Try a different PSU?
= Don't have one |
I think it's worth the trouble trying a different one. Can you ask around to see if someone can lend you a good PSU for a test? |
|
| Back to top |
|
|
dskeeles
Joined: 03 Feb 2005
Posts: 26
|
| Posted: Sun Mar 02, 2008 12:42 am Post subject: |
|
|
| Quote: |
I think it's worth the trouble trying a different one. Can you ask around to see if someone can lend you a good PSU for a test? |
I've an old ATX case at my parents', so I'll bring that back; it'll also give me a spare Optical drive to try out. I did test this setup with the Optical drive in a powered external USB enclosure as well as the other external USB drive, so I'd have thought that would have negated a power dip causing drive failure; and the CPU stress tests should have done the same for the CPU
Thanks! |
|
| Back to top |
|
|
mike42
Joined: 18 Feb 2008
Posts: 3
Location: UK
|
| Posted: Sun Mar 02, 2008 11:20 am Post subject: |
|
|
You don't need to get the meter out to check PSU voltages - it should be in the BIOS.
And by IDE BIOS settings I meant speeds - the system with a bad PSU would run
correctly using PIO, but not DMA. It's not that the drivers are doing anything non-standard,
just that they may not be running the HDDs at the same speed as the BIOS does. |
|
| Back to top |
|
|
psiu
Joined: 23 Aug 2005
Posts: 1201
Location: Plymouth, MI
|
| Posted: Sun Mar 02, 2008 1:22 pm Post subject: |
|
|
| mike42 wrote: | You don't need to get the meter out to check PSU voltages - it should be in the BIOS.
And by IDE BIOS settings I meant speeds - the system with a bad PSU would run
correctly using PIO, but not DMA. It's not that the drivers are doing anything non-standard,
just that they may not be running the HDDs at the same speed as the BIOS does. |
Most accurate to check with meter--if there is something faulty on the board are you going to trust those readings? |
|
| Back to top |
|
|
dskeeles
Joined: 03 Feb 2005
Posts: 26
|
| Posted: Mon Mar 03, 2008 3:13 pm Post subject: |
|
|
[quote="dskeeles"] | Quote: | | I've an old ATX case at my parents', so I'll bring that back; it'll also give me a spare Optical drive to try out. I did test this setup with the Optical drive in a powered external USB enclosure as well as the other external USB drive, so I'd have thought that would have negated a power dip causing drive failure; |
Well... it gets stranger... but it's working!
I did try the other optical drive from my old PC (older Toshiba DVD-ROM vs my NEC 2510 DVD-RW), and... it went through the Windows Install process! I stopped it, checked for possible reasons (Master/Slave, etc), tried my other drive in case I'd accidentally corrected a cabling fault (but it still fails as before), then went back to the DVD-ROM and completed a successful Windows install.
I also tried the Ubuntu 7.04 LiveCD - and that still failed at the kernel even with the DVD-ROM.
So... really strange.... I wonder if perhaps it's a combination of reasons... ie.
- Perhaps the DVD-RW was faulty (but worked fine with my AOpen board)
- Perhaps the Gigabyte Mobo firmware, or the IDE Controller, had a disagreement with my NEC drive
- Perhaps the other (really old) USB CD-ROM also disagreed with the Gigabyte firmware
- Ubuntu fails anyway because perhaps it doesn't have drivers for the newer Gigabyte board (I think the installer is 12-18 months old)
......
I'm going to have to investigate more fully, as it now raises the question of what happens if I buy a newer DVD-RW or Blu-Ray drive; why should it work in future if my DVD-RW doesn't. I still need to take a look at voltages, so will check if that's one cause. For the moment, I need to get my Media Center back on its' feet so I can get my media/backups/etc. back online!
Thanks everyone for your help to date! I'll try and work out the root cause asap.
D |
|
| Back to top |
|
|
xan_user
*Lifetime Patron*
Joined: 21 May 2006
Posts: 1231
Location: Mendo/Sonoma (N.Cal)
|
| Posted: Mon Mar 03, 2008 9:52 pm Post subject: |
|
|
| maybe a piece of grit is between cpu and HS, causing overheat? what are temps in bios? |
|
| Back to top |
|
|
dskeeles
Joined: 03 Feb 2005
Posts: 26
|
| Posted: Tue Mar 04, 2008 10:02 am Post subject: |
|
|
| xan_user wrote: | | maybe a piece of grit is between cpu and HS, causing overheat? what are temps in bios? |
The temps are mid-30's (Centigrade), so it doesn't look like it should be a problem. Also, I've detached the heatsink 3-4 times while swopping mobos, so there should have been some variance in the behaviour at some point if it was this.... I will keep an eye on it though.
Thanks!!!
Damian |
|
| Back to top |
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
|
FAQ
Search
Memberlist
Usergroups
Register
Profile
Log in to check your private messages
Log in
