Is my 1.6A slowly dying after 8 months???

[Regalk]

I just rebooted and cannot boot into Windows anymore. "Win/system config corruption - use r to repair Windows". Try at default 1600 and then at 1700 boots into windows fine. Increased FSB to 133 and again corruption - cannot load windows. Is this woody losing its pwer perhaps after 8 months at 2144??
After several things I tried I believe the only 2 issues left are the cheap powersupply could be failing or CPU going bad?? Still its boots into windows even at 109 FSB (1733). Strange. I even reloaded Windows using the repair feature (at 2144) and on first reboot again windows corruption.
What do you guys think. Voltage is at 1.525 where it has alwys been.

 

[Duvie]

I have had one die (or diminish it oc tremendously) and I did not get error like that....You may want to look at ram or a stick of ram...may also want to look at the harddrive itself...

Too early to tell...May want to swap some parts if you have access to extra parts!! I did that to isolate the cpu as my cause of problems....

If 1.525v was all you ever used I would be highly surprised that would have damaged it!! I have ran at 1.71v actual for 6+ months now with no issues....


What mobo?? What is actual volatge of vcore as seen my mbm??? Some mobos greatly undervolt or overvolt...Does your mobo have or are you using any overvolt jumper??? I am just trying to figure out if the 1.525v is actually what the volatge is or it may be a bit higher...


Have you tried to reduce all the memory timings to the least conservative, then set cpu back to default, then reformat, then reinstall winxp in a clean install....

 

[THUGSROOK]

try using memtest86 (it also tests cpu cache)
that way youre eliminating HD and OS as a factor

HTH

 

[Regalk]

Great I have 2 heavy weights online.
Thugs Knows my board - P4S333 (we went thru this back in Feb/Mar). I tested the RAM thoroughly and relaxed the timings to the slowest possible. I have now ruled out the RAM. The Mobo does not overvolt (even though I flashed it to the P4S533 1009C) Bios.
I have ruled out Video card and NIC. ALso ruled out HD since I did a reinstall of WinXP using 2133 but as I said after the first reboot WIN corruption. So I reboot at say 1730 - absolutely no more corruption - loads Windows fine.
I am now leaning towards Power supply now (too generic and may be dying when pushed too far).
I am thinking of also redoing the CPU/Heatsink installation even though I know this is not the problem since CPU temp is only 34C at idle.

 

[Duvie]

I am familiar with that board as well as it was the one I had that with its overvolting (which mine overvolted .03v over bios setting) and overvolt jumper set destroyed my first 1.8a chip....I thought I had it at 1.7v set in the bios but in fact the voltage was 1.83v....

 

[Regalk]

But its not a (over) voltage problem - I know since I spent the last 2 hours with it at 1.5, 1.525, 1.55 1.575v. Same problem at each. I am convinced that its either the cpu or PS - I would appreciate it if someone can come up with a third factor.

 

[Duvie]

I am not saying it is....I am just wondering if the voltage was or is higher then you think (ie 1.5,1.525,1.55, etc) and it may have been damaged....

I have the board ad I know the overvolt jumper when jumpered on board will not show up in the bios in hardware monitor or settings....


What did the blue screen at bootup exactly say...Did it just say "Win/system config corruption - use r to repair Windows"??? Mine would list register type codes and had a whole bunch of writing on the BSOD....

 

[Regalk]

I can't remember using the overvolt jumper - it is probably at default setting. It is not a BSOD at all. Just a regular black screen with the about 5 - 6 lines saying that there has been a system error or something like that in c:\windows\system etc etc. and to use the repair feature at the first screen to do a repair etc.
I then backed down to 1600 and viola it booted right into windows. I inched it up to 1728 and again no problems. I went straight to 133 and boom crashed again. All the while I left it at 1.5 then at 133 I moved it up to 1.525. So even if it overvolted it would probably be around 1.54/1.55 which is fairly low enough anyway that is why I am ruling that out.
Before ai started all this I had the compter on for about an hour doing nothing (running at 136 x 16!). As soon as I rebooted all hell broke loose.

 

[THUGSROOK]

mighta killed the motherboard...

youll need to swap out parts til ya find the prob.

 

[orion7144]

Originally posted by: Regalk
I can't remember using the overvolt jumper - it is probably at default setting. It is not a BSOD at all. Just a regular black screen with the about 5 - 6 lines saying that there has been a system error or something like that in c:\windows\system etc etc. and to use the repair feature at the first screen to do a repair etc.
I then backed down to 1600 and viola it booted right into windows. I inched it up to 1728 and again no problems. I went straight to 133 and boom crashed again. All the while I left it at 1.5 then at 133 I moved it up to 1.525. So even if it overvolted it would probably be around 1.54/1.55 which is fairly low enough anyway that is why I am ruling that out.
Before ai started all this I had the compter on for about an hour doing nothing (running at 136 x 16!). As soon as I rebooted all hell broke loose.

I RMA'd a XP1700+ 2 months ago for similar problems, except mine would boot into windows then out of nowhere data corruption. Put my kids CPU in and work fine. Funny thing is AMD sent me a letter with the processor saying they were out of Retail 1700+'s so here is an 1800+ with a retail HSF.

 

[orion7144]

Forgot to mention that my 1700+ would run fine at 100mhz FSB, no corruption.

 

[MadTom]

Regalk this maybe will sound stupid (as I'm aware that you probably checked it) - how is your AGP/PCI ratio, e.g. are they working at 66/33 MHz. My main idea would be that the RAM is crappy, but as you checked it...

Hope you will find the solution to the problem.

 

[Regalk]

PCI/AGP all in spec at 133. Only slightly out at 136 - this board does not lock the AGP/PCI.
Remember this cpu ran very well at 136FSB for 8 months.
This is actually my friends computer I built for him. I built a similar one for myself around the same time and my PCU temp is 43C (10 C higher than his) using the Volcano7+ (I suspect that contact is not as solid as the Intel HS). Anyway even though mine runs hotter it is still going great at 2133 plus I use mine more intensively than his by miles.
SO I rule out RAM
I rule out core volts and temps since these are fine
I rule out PCI/AGP since these were at 34/68
System is relatively clean - no dust build up so thats not it
Video card (MSI GF3 Ti200) and it runs well when CPU is at 1600
What do I have left = CPU/Heatsink and Power Supply and this is why I am leaning towards PSupply or slowly dying CPU. Since the CPU can O/C to 1728 I am now ruling out CPU.
I will be swappig out the CPU/HS in 2 weeks to see if it could also be the Motherboard for some reason.

 

[Duvie]

Don't rule out cpu yet....With sudden northwood death syndrome, a name being dropped all over the net that is a bit deceiving as most do not die but actually start having diminished returns on ocing....

My 1.8 did 2.4 w/ 1.68v for 2 months then I raised it to 2.52 w/ 1.83v where it did it for 1 week then I had to back it down to 2.4 again where it lasted for 1 week then I had to go down to 2.25 w/ 1.58v where it stayed for 2 months until I removed it....

This is what happens on a damaged cpu is diminished returns. You did 2133 for a long time now you can only do 1733...sounds like diminished returns to me...

 

[Regalk]

Originally posted by: Duvie
Don't rule out cpu yet....With sudden northwood death syndrome, a name being dropped all over the net that is a bit deceiving as most do not die but actually start having diminished returns on ocing....

My 1.8 did 2.4 w/ 1.68v for 2 months then I raised it to 2.52 w/ 1.83v where it did it for 1 week then I had to back it down to 2.4 again where it lasted for 1 week then I had to go down to 2.25 w/ 1.58v where it stayed for 2 months until I removed it....

This is what happens on a damaged cpu is diminished returns. You did 2133 for a long time now you can only do 1733...sounds like diminished returns to me...

I was hoping you would not say that - yeah I know of that diminishing death syndrome for Woodys - but I am hanging in on this one. The only difference between yours and mine is you pushed the vcore too high whereas I do not.

 

[Duvie]

True...Both Thugs and I damaged one with that board and we have always wondered if there was a voltage issue with that board....

I hope you are right, the PS would be cheaper to fix.....

 

[Regalk]

Currently running a P4 1.8A on the same board at 2502 at 1.5V (dual boot Linux/WinXP) and been running stable since 4 weeks now. S0 even though it overvolts to 1.52 it hasn't adversely affected anything. Linux is known for bringing out any weaknesses and it is stable as a rock in Linux. Temps are 28C at idle. Well who knows - I will be running a series of tests in 2 weeks so I will find out one way or another if it is SNDS.
If I lose this battle this will only be my second in over 5 years (The MSI KT333 was my first).

 

[mrman3k]

Sounds like SNDS, ouch!

 

[Duvie]

Just make sure that jumper for overvolt is not enabled...it is just up and to the left of the cpu....I like you can't imagine damaging it with 1.525v....Even if the overvolt was on it likley would have only been 1.625v and that in itself would still not seem like enough.

Sometimes with mobos it is not the voltage it runs steady at as much as the possibility it spikes voltage and for a brief second or so could send a damaging about of vcore to the chip....I saw a lot of fluctuation in vcore when running my chip. Later on with a different mobo I did not see that same vcore fluctuation....

 

[Regalk]

Hey MD are you reading this. Duvie is pretty well known around here and he has a valid point concerning voltage spikes. Check out that jumper he spoke about top of board to the left (Read the manual first to understand about how it works).