K8T Neo: Problem at power-on...

Question

Hello!
Somehow I experience a strange startup-problem ... and I could really need some advice to resolve this issue as I'm out of ideas.
Last weekend I completely disassembled my rig for the annual "spring-time maintenance". I cleaned out dust, I cleaned the inside and outside of the case and I replaced my DVD-ROM and CD-RW drive with a LG DVD+/-R(W)/RAM drive. After everything was put together again I powered up my system - and it started to act strange.
Symptom: At power-on the system hangs at "Early Chipset Init" (D-Bracket 1-4: GRRR) and keeps on self-resetting. Several seconds (10-30) later the system starts to POST as expected and runs smoothly until ... power down.
During the night the system is completely disconnected from the power as it drains a lot of juice in stand-by (~12W for the rig + ~18W for several other external components like scanner, audio, ...).
What I've done so far to track down the error:
- Reseated power connectors. Nope - problem persists.
- Pulled all unnecessary components (just MoBo, CPU, RAM, GFX-card, Keyb). Nope - same prob.
- Pulled all memory sticks, left only one in the slot (and tried every other one). Nope - same prob.
- Cleared CMOS - Nope - same prob.
- Ran memtest86+. Showed no errors after 10+ passes.
Now I'm out of ideas ... except for the power-on problem the system runs just fine as before. No crashes, no malfunctions, no nothing.
Technical Data:
MSI K8T Neo FSR (BIOS 2.0)
AMD Athlon 64 3000+ (Clawhammer Core)
Zalman CNPS7000B CPU-Cooler
3x 512MB Kingston CL2.5 PC400 DDR-RAM (@DDR333 - 3 sticks = DDR333 only!)
NVIDIA GeForce 6600GT AGP (Power Connector is plugged in)
Highpoint HPT370A IDE Controller (PCI 3)
Hauppauge WinTV PCI FM (PCI 5)
Seagate 80GB SATA - Operating system
Hitachi 160GB - OnBoard Pri IDE, Master
LG 4167B - OnBoard Pri IDE, Slave
Maxtor 200GB - HPT370A Pri IDE, Master
Seagate 400GB - HPT370A Pri IDE, Slave
Maxtor 80GB - HPT370A, Sec IDE, Master
Tagan TG480-U01 480W PSU (3.3V@28A, 5V@48A, 12V@28A)
As I have a Wattmeter at my avail: After power-on and boot-up the system "sucks" ~315W (rig + monitor). If I put load on the system (CPU at full clock, GeForce displaying 3D) it drains ~390W.
As mentioned above I never had any issues with the system.
Does anyone have a suggestion how to resolve the rather annoying power-on problem ?!?
Thanks,
Jochen

Answer 1

Quote from: Tiresmoke on 06-April-06, 20:09:53
Did you checl and replace the CMos Battery? You may need to do a complete Bios Clear and start fresh there as well.
Also as the Optical drives changed did you try removing them one at a time and see if the system acts noirmally then?
Hello, Tiresmoke!
. I unplugged the PSU from the wall socket and removed the battery while clearing the CMOS. The CMOS Battery is OK - ~3.34V (reading from a Voltmeter - matches with the display at PC Health Status).
. I already took out the LG DVD-RAM drive and put the old optical drives back in. Even in its old "original" configuration the aforementioned problem persists.
. I already reseated the CPU and renewed the thermal compound.
However, I even get the "Early chipset" startup when the system is bareboned ... CPU + 1 memstick (doesn't matter which one out of the three I plug in) + Keyb + GFX. D-Bracket shows "Early Chipset Init", the boards goes through several reset-loops and fires up fine.
BTW: Maybe this will give some clue ...
. When I do a warm reset (reboot triggered by OS) the board restarts fine - Reset, POST, Boot.
. When I shutdown and disconnect power from the PSU, lets say for about 5 minutes, then reconnect it and power-on the "Early chipset init" shows up.
Are there any chances that the PSU is actually causing the problem although this particular problem never showed up before the "spring-cleaning"?
Thanks,
Jochen