Is it normal to get a little fluctuation on the Vcore voltage?

[Twigsoffury]

I'm trying to get a little bit higher on the OC and i'm at 1.401v / 3.771Ghz but i get extremely random BSOD and hard resets with no pattern, it'll play BF3 perfectly for 3 hours straight, play another game and exit to the desktop and be fine. other times it'll just BSOD right off the bat.

I've noticed my Vcore fluctuates between 1.395 ~ 1.408v or so despite the fact that i have it manually set to 1.401v in the bios menu. (looking at HWmonitor and CPUZ)

VR's are about 98~101F the bricks are 90~93F and the capacitors aren't even warm to the touch. (using my LCD thermometer to probe the VR's and **** if your wondering how i know the temps lol)


But anyways is this normal to get this sort of flux in the vcore voltage, or is my power supply showing its age? (2005 Enermax 600W Noisetaker II 1.0a E80+S certified)

[Twigsoffury]

no help ?


: (

[diluzio91]

Whats your mobo brand? There should be a setting, but i can't remember the name of it, i think it's v droop? if you tighten that it wont flux.

[Twigsoffury]

Whats your mobo brand? There should be a setting, but i can't remember the name of it, i think it's v droop? if you tighten that it wont flux.


MSi-7599 (770-C45) think its a AMI bios

Matter o' fact i just flashed the bios to the newest one yesterday : )

i'll poke around, any other "names" that goes by?

[Konrad]

I'd be very surprised if the mobo VRM could accurately regulate voltages within <5 mV.

Also bear in mind that the hardware monitor (voltages, fan tachs, and thermal sensors) is not 100% accurate either ... in most cases such functions are handled by a chip which sits somewhere kinda-near-the-processor towards the top of the mobo, sometimes these functions are integrated into the programmable southbridge/SuperIO chip. What this means is that the chip isn't always reporting the voltages and temps which are present around the critical components so much as around its immediate proximity on the mobo.

I wouldn't be especially concerned with such little fluctuations, especially since the margins for measurement error and component tolerance exceed them. You might want to observe if the voltages deviate whenever your HDD spins up/down or some other power-hungry device suddenly changes the load on the mobo power bus or even on the main PSU. A sad truth is that, contrary to what is advertised, and what the user explicitly configures, the BIOS firmware may be programmed to "automatically" adjust processor voltages as the mobo manufacturer sees fit.

Having said all that, you could have a lemon mobo or proc. If the problem is really a problem and it persists or worsens then try these components out in different machines to isolate which is the cause. For all you know you've got a bad capacitor or something.

[Twigsoffury]

I'd be very surprised if the mobo VRM could accurately regulate voltages within <5 mV.

Also bear in mind that the hardware monitor (voltages, fan tachs, and thermal sensors) is not 100% accurate either ... in most cases such functions are handled by a chip which sits somewhere kinda-near-the-processor towards the top of the mobo, sometimes these functions are integrated into the programmable southbridge/SuperIO chip. What this means is that the chip isn't always reporting the voltages and temps which are present around the critical components so much as around its immediate proximity on the mobo.

I wouldn't be especially concerned with such little fluctuations, especially since the margins for measurement error and component tolerance exceed them. You might want to observe if the voltages deviate whenever your HDD spins up/down or some other power-hungry device suddenly changes the load on the mobo power bus or even on the main PSU. A sad truth is that, contrary to what is advertised, and what the user explicitly configures, the BIOS firmware may be programmed to "automatically" adjust processor voltages as the mobo manufacturer sees fit.

Having said all that, you could have a lemon mobo or proc. If the problem is really a problem and it persists or worsens then try these components out in different machines to isolate which is the cause. For all you know you've got a bad capacitor or something.


well i didn't see anything about voltage dropping crap in the bios.


but i did up the voltage to 1.410 and it seems to be doing well, but the next few days will tell me how stable it is i guess.


I think i just drew the short stick in the X4 955 processor bin of random quality chance lottery.

[diluzio91]

You might have. Both of the 955's i've had my hands on overclocked well above 4.0 without an issue. one got downed to 3.7 because of the mobo making weird noises, but thats it.

[Twigsoffury]

You might have. Both of the 955's i've had my hands on overclocked well above 4.0 without an issue. one got downed to 3.7 because of the mobo making weird noises, but thats it.

this is a pretty budget mainboard as well.


I mean it didn't even come with VR cooling sinks and the front side bus seems limited to a hilarious 233mhz or so.

[diluzio91]

lol. with a 955 don't bother with a fsb overclock. just jack the multi up...

[Twigsoffury]

lol. with a 955 don't bother with a fsb overclock. just jack the multi up...

"Faster" FSB = Faster computer don't it?



I want everything overclocked, including the clock.

[diluzio91]

"Faster" FSB = Faster computer don't it?



I want everything overclocked, including the clock.

Not nessecarily... the PCI lanes for example only work properly when set to the correct frequency. the biggest benefit of overclocking your fsb in my experience is the boost to ram speed, assuming you can get it stable.

[Outlaw]

Not nessecarily... the PCI lanes for example only work properly when set to the correct frequency. the biggest benefit of overclocking your fsb in my experience is the boost to ram speed, assuming you can get it stable.

Or if the multiplier is already maxed.

[blaze15301]

With the 955 be your ram can not exceed 1600mhz. I have found it is normal for the v core to fluctuate. The best way I can overclock my ship is threw the fsb with a bit of a multiplier I can't exceed 17.5 or it wont stay stable and the board gives me weird errors. I have had my ship up to 4.5 with a v core of 1.6. it stay like that for maybe 15 mins. Then crashed.

[Konrad]

It's not the BIOS firmware automatically assigning whatever vCore the mobo maker feels (based on Intel specs) the processor should get?

[diluzio91]

When my roommate gets back i'll post his OC specs. he's been as high as 4.2 stable, we just heard some funky noises from his computer when it was up that high.

[Twigsoffury]

With the 955 be your ram can not exceed 1600mhz. I have found it is normal for the v core to fluctuate. The best way I can overclock my ship is threw the fsb with a bit of a multiplier I can't exceed 17.5 or it wont stay stable and the board gives me weird errors. I have had my ship up to 4.5 with a v core of 1.6. it stay like that for maybe 15 mins. Then crashed.

sounds exactly like mine.

i mean I "can" pass DDR3-1600 but i get strange a@# errors. Right now its DDR3-1500 @ CL8-8-8-8-24-2T


Not nessecarily... the PCI lanes for example only work properly when set to the correct frequency. the biggest benefit of overclocking your fsb in my experience is the boost to ram speed, assuming you can get it stable.

It's manually set to 100mhz. I don't think i've ever once messed with the
PCi-E or AGP bus ever in my life when overclocking.


[edit] FYI this computer has been running non stop since the last mention... no crashes so i guess that .004 volt made the difference.

http://i41.tinypic.com/6ekygk.png


3dmark 2006 score is 20,993

[blaze15301]

I found out that the 955 cant go past 1600 or you will get really weird errors or it just wont work.

[Konrad]

You can always reflash the firmware on the RAM sticks themselves, change SPD configuration and such. Of course, simply changing a few hardware identifiers is not going to be a substitute for higher quality RAM. That sorta stuff is a bit too crazy for me.

[Twigsoffury]

That sorta stuff is a bit too crazy for me.

ya the same here.

[edit]

its working great so far with the voltage raised.