Chipset Voltage too low?

[ccrespo]

Here are my comp specs:

Athlon Barton 2500+ (running at 1.1ghz stock)
Cooled with a Coolermater...vornado? very loud keeps it at 30degrees or so.
Epox 8rda+ nforce2 mobo
(2x512) 1gb Corsair XMS3200c ddr ram
Sapphire 9800 pro 128mb (256bit)
200gb Seagate Barracuda HD
8x NEC nd2500a DVD-/+r burner (having issues with this too!)
5 small case fans running
450w JustPC powersupply (just bought)
Windows XP Pro.

I've been running this system for more than a year and it's been doing good (only overlcoked to its normal 2500+ and occasionally brought it to 2800+), besides the occasional OS hiccups, it's been running smoothly. A few weeks ago the computer began to freeze up a lot then suddenly I got IRQL errors with physical memory dumps. My 9800 pro got damaged (screen was all different colours and pixelated). newegg RMA'd it, and I got it replaced.
Before I got it back, I noticed that the Chipset Voltage (vdd) was flucuating a lot from 1.20-1.60v, turning red everytime it was below 1.5v.
I just reformatted and have a clean install of everything. However, if I try to clock the cpu up to 2500+ (1.8ghz) it'l give me IRQL_NOT_LESS_OR_EQUAL blue screen errors all the time. I replaced my aging 350w ps with a 450w and it still didnt change. I can't even mess with the memory timings without it freezing up. I tried disconnecting all the case fans but that did nothing. Maybe it's the dvd burner? (it freezes up my computer at times and makes it really really slow when reading dvds or cds).

UPDATE:
Also at times, like just happened when playing a game, i hear a BUZZ noise through the speakers, as if it wasn't grounded and you can actually hear the current going through the speakers, what gives??!

any help or comments on what could be wrong?

thanks in advance

 

[montag451]

Can you try a new PSU -branded?

And while you are at it, run memtest over night

 

[ccrespo]

I left memtest running when I left for work in the morning, its been going on for around 8 hours now, i'll check when I get home.

Sorry, what do you mean by PSU -branded? as in a name-brand PSU? or am I being dumb? dont know too much about the tech details of these things. And I only have access to my old 350w PSU that came with my case, and it does the same exact thing, same voltage. I'd have to go buy another to test it

 

[montag451]

Sorry - I would say a brand of psu that is well-known for reliability = thermaltake/fortron/antec/TAGAN [yep]

 

[Fern]

Yeah, I think he means a good brand psu.

In addition to metest, I'd put motherboard monitor 5 on there. Choose the "sys log" feature and have it record to a text.doc. Then later go back and have a look at the text.doc and see if your voltages are over the place. The swing in the chipset voltage is not good.

Watch your AGP voltage too. Don't wanna a fry another card.

 

[ccrespo]

Well I just got home, memtest ran all these hours and had 1310% coverage, not one error found, so i guess the memory is fine.

Im not sure where I can find a name brand PSU, maybe i'll try to return this one i just bought and get credit for another one, i dunno (i dont think its the PSU though, but what do I know..?)

I'm getting motherboard monitor 5 now

 

[ccrespo]

UPDATE:
as you can tell, Core 0 is flucuatiing like crazy and actually never goes about 1.5 (where i think it should be), and +5.00 is never quite at 5, rather at 4.8 or so...

I have NO IDEA what any of this means!


+---------------+---------------+----------+------------+------------+------------+------------+------------+------------+------------+------------+------------+------------+------------+------------+------------+
| | | | Case | CPU | Sensor 3 | Core 0 | Core 1 | +3.3 | +5.00 | +12.00 | -12.00 | -5.00 | Fan 1 | Fan 2 | Fan 3 |
+---------------+---------------+----------+------------+------------+------------+------------+------------+------------+------------+------------+------------+------------+------------+------------+------------+
| 3/23/2005 | 9:51:55 PM | 1101 MHz | 42° C | 42° C | 0° C | 1.33 V | 1.71 V | 2.80 V | 4.84 V | 12.28 V | -12.85 V | -5.14 V | 0 RPM | 0 RPM | 0 RPM |
| 3/23/2005 | 9:51:25 PM | 1101 MHz | 42° C | 42° C | 0° C | 1.46 V | 1.70 V | 2.83 V | 4.92 V | 12.22 V | -12.68 V | -5.19 V | 0 RPM | 0 RPM | 0 RPM |
| 3/23/2005 | 9:50:55 PM | 1101 MHz | 42° C | 42° C | 0° C | 1.49 V | 1.70 V | 2.80 V | 4.84 V | 12.16 V | -12.85 V | -5.14 V | 0 RPM | 0 RPM | 0 RPM |
| 3/23/2005 | 9:50:25 PM | 1101 MHz | 41° C | 42° C | 0° C | 1.33 V | 1.70 V | 2.83 V | 4.87 V | 12.22 V | -12.68 V | -5.14 V | 0 RPM | 0 RPM | 0 RPM |
| 3/23/2005 | 9:49:55 PM | 1101 MHz | 42° C | 42° C | 0° C | 1.34 V | 1.68 V | 2.85 V | 4.87 V | 12.28 V | -12.68 V | -5.14 V | 0 RPM | 0 RPM | 0 RPM |
| 3/23/2005 | 9:49:25 PM | 1101 MHz | 42° C | 42° C | 0° C | 1.47 V | 1.70 V | 2.82 V | 4.89 V | 12.22 V | -12.68 V | -5.19 V | 0 RPM | 0 RPM | 0 RPM |
| 3/23/2005 | 9:48:55 PM | 1101 MHz | 42° C | 42° C | 0° C | 1.49 V | 1.70 V | 2.82 V | 4.89 V | 12.28 V | -12.77 V | -5.14 V | 0 RPM | 0 RPM | 0 RPM |
| 3/23/2005 | 9:48:25 PM | 1101 MHz | 42° C | 42° C | 0° C | 1.49 V | 1.71 V | 2.85 V | 4.87 V | 12.22 V | -12.85 V | -5.14 V | 0 RPM | 0 RPM | 0 RPM |
| 3/23/2005 | 9:47:55 PM | 1101 MHz | 42° C | 42° C | 0° C | 1.42 V | 1.70 V | 2.80 V | 4.81 V | 12.40 V | -12.77 V | -5.14 V | 0 RPM | 0 RPM | 0 RPM |
| 3/23/2005 | 9:47:25 PM | 1101 MHz | 42° C | 43° C | 0° C | 1.41 V | 1.70 V | 2.82 V | 4.81 V | 12.40 V | -12.85 V | -5.24 V | 0 RPM | 0 RPM | 0 RPM |
| 3/23/2005 | 9:46:55 PM | 1101 MHz | 41° C | 42° C | 0° C | 1.46 V | 1.70 V | 2.82 V | 4.92 V | 12.28 V | -12.85 V | -5.19 V | 0 RPM | 0 RPM | 0 RPM |




any help would be appreciated!

 

[Fern]

You've set to record in 30 secs. Might wanna move to only one second intervals. Thats help catch when a voltgae fluctuation causes a crash etc.

I can see you cpu is prolly set at 100mhz x 11 (multi)

Loooks like vcore, voltage to cpu is @ 1.70v. 165v is spec, but a little extra doesn't hurt anything. Your temp looks fine too at about 42c

I guess the one flucuating too much (1.33v -1.49v) is chipset, which should be 1.50v. I have not seen this b4, but would guess it can be considered a possible source of the problem.

The one at 2.80v looks like vdimm, voltage to ram, and if so would be fine. However, it also could be the 3.3v rail. Hard to say because of the way the chart gets "distorted" when pasted. If the 2.8 is the 3.3v rail. It is outta spec. Shouldn't be lower than 3.13

The 5v and 12v rail seem to be within spec.

Here some "theory" on my part: IIRC, the mem controller for these systems is on the NB (on a 64 bit system the mem controller is on the cpu) if the voltage to the chipset is fluctuating too much it may cause instability via the mem controller.

Might be worth it to call Epox about it.

I would guess if the result (big fluctuations in chipset voltage) stayed the same with diff PSU's, the prob might be within the mobo itself?

BTW, is it unstable at 133 mhz too? Just curious

 

[ccrespo]

thanks for the reply, raising the power any bit, even to 133mhz will mess it up. I'll put it to 1 sec intervals when I get home, but it runs pretty stable at 100x11, I just want to use my barton where it should be, at 1.8ghz (plus, i really dont like the occasional buzzing noise coming from the speakers).

the 1 year warranty with epox is over, im gonna call em though to see if they can help me with this. Id rather not have to buy a new mobo/cpu, but ugh, it seems like that what'll have to happen