New Build - vcore drops & possible psu problem?

C

ChillJF

n00b
Joined
Mar 24, 2003
Messages
5
Hello all, glad to be back. Its been 3 years + since my last nf7-s build which broke me into the ocing world. I just built a IP35-E with e2180 & 2g of ballistixs this weekend & blown away by the speed/gaming increases Im seeing, but still not content :), I now begin the journey to find my ocing limits. Im currently running 300x10 fsb, 3:5 divider, 1.405vcore, 2.2v ram & have been orthos + memtest stable & gamed on WoW for 6 hours+ no problems. After feeling comfortable with stabality, I try increasing the fsb 320x10 & lower my divider to ddr960. I post but crash b4 loading windows, so I up my vcore to 1.425. I now can load into windows but shortly crash upon running several apps. So I go back & bump vcore up to 1.445 now. Everything seems stable to attempt orthos now but I realize that cpu-z is reporting 1.384 vcore & speedfan is reporting 1.38 with a max of 1.39. I decide to give orthos a go & crash within 3 mins of running it. I am now stuck with a few questions b4 I continure my journey. Is cpu-z, speedfan & abit EQ incorrectly reporting my vcore in windows? Or is something else capping my vcore @ 1.384. I have C1E & speedstep disabled in bios, thermal control enabled, but was going to try disabling it (Im reading that most leave it enabled with oc) so I dont think that is the problem. It felt like when I was increasing the vcore @ 320x10 from 1.405-.445 in the bios it was doing something cause windows went from not loading, to somewhat stable b4 loading othros. I stopped my journey for now & was doing some google searching for solutions but couldnt find any. I read about vdroop but dont think that is my problem since Ive never seen cpu-z, speedfan, abit EQ report any higher vcore than 1.385 idle or load.

Only thing I can possible think of is maybe my psu is rails are too weak with running a 8800GT. Here is link to my psu http://www.ultraproducts.com/product...roductID=3 87. I got it Free AR like 6 months back & have had no problems with it. Im aware its on the border line of a cheaper psu, but I figure it would do until I got some more cash.

I did some more stabality testing with the system running 300x10, 1.405vcore which ran for 10 hours + yesterday & had no problems. After an hour of running Orthos small FFT & monitoring volts thru speedfan & cpu-z. Speedfan is locked at 1.38, while cpu-z keeps going between 1.376 & 1.384. I than goto load firefox in a window & happend to have my eye on cpu-z at the time & notice my vcore dropped for a sec to .040 & bam, computer shutsdown. I have cpu shutdown when fan fails in bios, but I did not hear the fan stop & dont think its related to that. It seemed like the power supply cutoff vcore juice for a second or was a major spike in power flow & computer shutdown because of it. After this I got kinda scared & dropped my oc settings back to default to check what vcore is being reported by speedfan & cpu-z & its 1.32 (bios is set to 1.325) so its accurate. Im running Orthos small fft & rthdribl now at the same time & notice vcore is 1.296 when stressing the video card. I will try & test what vcore drops too when running 300x10 1.405 & stressing the vcard.

I believe I might have causeed some damage to the psu yesterday when pushing the core with 1.445 & 2.2v ram. I was skeptical of the psu from the start since I got it Free AR & reviews werent all too high on it. I had no problems when running 300x10 1.405v yesterday for 10+ hours of stability testing & gaming. Today within an hour of orthos after last nights 320x10 attempt & problems, I have the drop in vcore & comp shutsdown. Even tho I think its the psu, I could easly be wrong but thats why Im here to ask the experts for help :)
 
Ok Ive done some more testing & determined that it has nothing to do with speedfan & cpu-z. I was monitoring vcore in bios while raising it & bios never reports it above 1.38. Interesting thing, I decided to try out out the different vcore settings to see what bios is reporting. Vcore 1.325, 1.345 are dead on according to bios, cpu-z & speedfan. Vcore 1.365 is being reported as 1.38v in bios (1st setting that isnt accurate, its actually overvolting, not undervolt) Vcore 1.385 & anything above (max I tried was 1.465v) report 1.38v in bios, cpu-z,speedfan.

I am running bios 15 which the board came flashed with. I was gonna try a new bios out to rule out a possible bad bios flash from factory, but dont wanna risk cause new problems. Has anyone ever heard of this vcore issue or should I contact abit to RMA the board as my issue is something more serious than I can fix. I purchased it last week from newegg.
__________
 
The problem is most likely to be that you have hit the max FSB the motherboard will do.
I built an IP35-E system before XMas which also had an FSB wall of 320MHz.

Test with even lower CPU multipliers and lower ram speeds settings.
If it still fails, its down to the motherboard.
You may be able to tweak it into submission.
All it gave me for my efforts was headaches and a corrupted hard drive so make a backup before you start tweaking!
 
You have to remember, vcore isn't the only limiting factor in overclocking. You may have to up the MCH (Northbridge) voltage in order for your motherboard to allow you to up the FSB from where you are currently. Also, Ultra power supplies aren't exactly known for their superior (read: Ultra) quality at any time so it might very well be a power supply issue. It's pretty weird that the vcore would drop down so low all of a sudden but weird things happen with lesser hardware all the time.
 
It's possible that you need to up voltages besides vcore, but it sounds like there is a problem getting your board over that 1.38V mark. Whether this is due to the PSU or board itself, I don't know. I would definitely flash to the latest BIOS and see if that helps. Might not be a bad idea to try a different power supply if you have one.
 
Thanx guys.. yea my main problem or concern was why my vcore is not reporting going above 1.38v in bios or windows. I do not have a spare psu around that can handle this setup (would eliminate psu being problem). I do have an old pci video card that Im goin to try & check if vcore goes above 1.38. If pci card card doesnt help than I try flashing the bios with a diff version. I saw beta 16 was the newest out & was gonna try that, or if someone has a suggested version to try. Im running bios 15 that came stock from newegg & have the double boot issue. I wonder if abit flashes with /wb ? :)Will post back with findings later
 
It seems weird that you can't get a higher vcore on that. I can understand software monitoring utilities in Windows reading a lower vcore when the processor is at full load. That's normal for any board and called vdroop. On my quads, I have to have the vcore in BIOS set at least 0.1 above what I want it to be while running an OS.

My Q6600@3.6 requires me to set the vcore in BIOS at 1.525. While in an operating system running all 4 cores at 100% it's stable as long as the vcore stays above 1.40. That's a pretty large vdroop and somewhat annoying. While the PSU could be a partial cause of vdroop, normally the culprit is the motherboard. Some motherboards don't have vdroop as bad if they have better power regulation on the board.

 
You must log in or register to reply here.
Back
Top