Restarts intermittently. Pls help!

rejikumarrv

Hi all

My system restarts for no specific reason.

Specifications
OS: WinXP Home with SP2
CPU: Intel Celeron(TM) 1000 Mhz MAX: 1500 Mhz
BIOS Date: 10/17/01
BIOS Type: Award Modular BIOS v6.00PG
BIOS ID: 10/17/2001-i810-NS363-6A69MV3FC
OEM Sign-On: BIOS-N-2M (2K011017)
Chipset: Intel Whitney 82810 rev 3
Superio: NSC 363 rev 12 found at port 2Eh
BIOS ROM In Socket: Yes
BIOS ROM Size: 256K
Memory Installed: 128 MB
Memory Maximum: 1536 MB
Memory Slot 01: 128 MB
Memory Slot 02: 0 MB
Memory Slot 03: 0 MB

Observations
Restarts
when opening/using files of types .psd, .pdf, .vob, .dat etc.
when using opera, firefox etc.
when a folder containing large no of thumbnails scrolled
sometimes during windows/linux setup
sometimes even as I click the start button
Shutting down happened in Linux also.
Couldn't relate to any specific event from my side
CMOS checksum error shown twice in the last month
After the shut down procedure, the HDD light remains ON, sometimes
it even keeps running the motor
Sisoft Sandra says SMBIOS/DMI information inaccurate
The clock shown in windows pauses every 5 seconds, ie it takes a little
more time to proceed to next count.But uniform in BIOS setup display
Obviously the time lags.

Actions taken [but failed]
Replaced CMOS battery, votage and contacts verified
Loaded BIOS defaults
Reinstalled the OS and programs[tried xp pro, xp home, linux, win98]
Ran stress test on CPU, but no errors shown
Applied thermal paste and provided additional cooling for CPU
Swapped RAM, SMPS, HDD, IDE cables
Ran with the minimum config of MBD+CPU+RAM, HDD, Keyboard

Have been trying different options for the past one month whenever I had time. Now taking it as a challenge[ofcourse with your help only]. My suspects are motherboard & BIOS. What do you think?

profdlp

Did the problem appear right out of the blue, or did something get changed hardware-wise around the time it began?

Is there a newer BIOS available? Do you have a spare PSU you could try? What Brand and Model# is the MB?

You may wish to try running Memtest-86 as well. I know you mentioned changing out the RAM, but if every stick you try fails then I might suspect a memory controller failure.

What is the overall cooling inside the case like? Is it possible that the chipset is getting too hot?

edcentric

memtest is my bet. I have a machine that was giving reboots. Turns out to be flakey memory. When programs opened and saved to memory it would barf. Now I need to test and see if it the memory stick or hte controler.

rejikumarrv

Hi All, thanks for responding

Did the problem appear right out of the blue, or did something get changed hardware-wise around the time it began?

Restart on thumbnail folder scroll was there from the time I bought the system.
Current restart issue started nearly 4 or 5 months ago. I am not sure,
as the MTBF(mean time bw failures) was long initially. By now it has become unbearable.
The HW changes(that I remember) in last 6 months are
1. My SMPS gone bad as a lizard got trapped bw smps circuit board and body. It worked for a few days (with occasional spark noises) and then stopped. Only then I noticed the issue and got it replaced with new SMPS.
2. I had been using two 80 wire IDE cables in Primary and Secondary IDE. I replaced them with new 80 & 40 wire cables as they failed
3. Changed the CMOS battery after CMOS checksum error and the Time lag issue.

Is there a newer BIOS available?Do you have a spare PSU you could try? What Brand and Model# is the MB?
Unfortunately there is no free BIOS update I could find. The MBD company site www.vtech.com doesn't seem to have any updates there. Distributor site www.vesta-im.com doesn't seem to be UP for the past one month. The award bios affiliate www.esupport.com provides paid updates, but I need to make sure whether BIOS is the culprit before opting that.
The system is nearly 5 years old. I could not find a spare board or processor.
Details of the board are given at the end of this message

You may wish to try running Memtest-86 as well. I know you mentioned changing out the RAM, but if every stick you try fails then I might suspect a memory controller failure.
This one helped me to close in. The test is done. There are 10 (0 to 9) tests. It fails in test no 8. ie the system restarts just the same way as it does in the OS. The test is named as "Modulo 20, Ones and Zeroes" with a pattern 00000000-0. It doesn't go more than one percent. But surprisingly, the test doesnt' seem to be affected by which area of RAM it is testing. It only fails when I give a memory range of width greater than 33MB(eg: 40-74MB, 0-34MB). If I Check 0-10M,10-20M....100-110M etc it runs fine. I tried two RAM sticks, with same result. One more thing, it is a 100MHz FSB board. The sticks are pc133 and pc100 respectively.
What is the overall cooling inside the case like? Is it possible that the chipset is getting too hot?
I don't think that is a big issue. I had the board outside the case and running. Even provided table fan + ceiling fan blowing at it. Also it is raining here heavily at this time. I can keep the finger on the chipset IC as the system is running, without any discomfort.

Finally, the clock pause in Windows Xp (I mentioned in the first message) seems to be a bug of Xp. Many users report the same. Your systems too may have the same problem. They say that the time remains accurate (I don't know how. In my system time lags by hours at the end of the day).

Mainboard
Manufacturer : Vtech
MP Support : No
Model : i810-NS363
MBD Layout : IM-VMBO108/9-P-01-02

System Memory Controller
Location : Mainboard
Error Correction Capability : None
Number of Memory Slots : 3
Maximum Installable Memory : 1.5GB
Bank0/1 - A0 : DIMM 128MB/64
Bank2/3 - A1 : Empty
Bank4/5 - A2 : Empty

Chipset 1
Model : Intel Corporation 82810 810 Chipset Memory Controller Hub
Bus(es) : ISA PCI USB i2c/SMBus
FSB Speed : 1x 100MHz (100MHz data rate)
Max FSB Speed / Max Memory Speed : 1x 100MHz / 1x 100MHz
Width : 64-bit
IO Queue Depth : 4 request(s)

Logical/Chipset 1 Memory Banks
Bank 0 : 64MB SDRAM 2-2-2-0
Bank 1 : 64MB SDRAM 2-2-2-0
Shared Memory : 1MB
Width : 64-bit
Refresh Rate : 15.60µs
Power Save Mode : No
Fixed Hole Present : No

Memory Module 1
Manufacturer : Apacer Technology
Type : 128MB SDRAM
Technology : 8x(8Mx16)
Speed : PC133U 3-3-3-5
Version : 1.2
Date of Manufacture : Saturday, 25 December, 1999
Set Timing @ 133MHz : 3-3-3-5

PCI Bus(es) on Hub 1
Version : 2.10
Number of Bridges : 1
PCI Bus 0 : PCI (33MHz)
PCI Bus 1 : PCI (33MHz)
Multiplier : 1/3x

LPC Hub Controller 1
Model : Intel Corporation 82801AA 8xx Chipset LPC Interface Bridge
ACPI Power Management Enabled : Yes
Delayed Transaction Enabled : Yes

System SMBus Controller 1
Model : Intel 801xx/63xx SMBus
Version : 0.02
Advanced TCO Mode Enabled : No
Slave Device Enabled : No
PEC Support : No
Speed : 100kHz


I would also like to know what this "Modulo 20 test" is. If there is a memory controller problem is there any way around to keep using the board.

profdlp

rejikumarrv wrote:

...You may wish to try running Memtest-86 as well. I know you mentioned changing out the RAM, but if every stick you try fails then I might suspect a memory controller failure.
This one helped me to close in. The test is done. There are 10 (0 to 9) tests. It fails in test no 8. ie the system restarts just the same way as it does in the OS. The test is named as "Modulo 20, Ones and Zeroes" with a pattern 00000000-0. It doesn't go more than one percent. But surprisingly, the test doesnt' seem to be affected by which area of RAM it is testing. It only fails when I give a memory range of width greater than 33MB(eg: 40-74MB, 0-34MB). If I Check 0-10M,10-20M....100-110M etc it runs fine. I tried two RAM sticks, with same result. One more thing, it is a 100MHz FSB board. The sticks are pc133 and pc100 respectively.

This makes me think that the memory controller is kaput. You might see what options you have in the BIOS to adjust the RAM timing, etc, but if every stick you try fails and nothing you attempt setting-wise helps then I'd be inclined to believe that the controller itself is the culprit.

I would try another stick of RAM, too, if at all possible. All of that stuff is getting up there in years and you may just have two bad sticks. The fact that each fails in the same test under the same conditions leads me to believe otherwise, though.

One thing that can cause this problem is if your Northbridge is overheating. I know you've said that cooling is pretty good, but if none of the changes to the BIOS memory settings work (if they are available at all) then this might be worth checking anyway.

Good luck.