Friend is having an OC issue, and I can't figure it out

[Avalon]

I just recently got a mobile barton to my friend, and he hooked up his rig. It was a 35w 2400+, an NF7-S mobo, a 512mb PC4000 stick of Adata, an SLK800, Radeon 9700 pro. I think that's the important stuff. He was able to get up to 235FSB on his board, but wanted to hit 250, so he got himself one of those beta BIOS's...forget which one, but he said it was a real popular one, and it seemed to do the trick. He can now do 250FSB. Here's the issue:

He can do 250x9 or 240x10, and it requires 1.9v to prime stable. Mid 50s in temp under load. Then, when he sets it to 250x10, he still only needs 1.9v. I can understand that such a high FSB would require higher voltage, but why does it only run stable with 1.9v when he runs the system at a measly 250x9? That's only 2.25ghz. It makes no sense why that needs 1.9v and a 2.5ghz setting with the same FSB doesn't need any voltage boost at all. It seems like there is a problem holding him back from being able to run stable with a lower voltage at a lower speed setting. Anyone have any idea what may help him?

*Edit*
My friend just tried the setting of 12.5x200, and he said he still needed 1.9v vcore to pass prime. Wtf? I tested this chip personally on my own rig and it did 216x11.5 (similar clockspeed) at 1.775v stable. Is there a problem with this beta BIOS he's using? That's what all this is leading me to think.

 

[Corporate Thug]

maybe a chips thats not that great of an OCer on low vcore?

 

[Shimmishim]

Originally posted by: shady06
maybe a chips thats not that great of an OCer on low vcore?

kinda like how the bh-5's don't show their potential w/o high voltage

btw, what beta bios was it?

 

[Avalon]

That's not the case at all. If you've seen any of my posts in this particular forums, they are often filled with detailed results of my overclocking. All of my chips I had experience with were excellent overclockers with low voltage. This chip was no exception. I asked him what bios he was trying, but he signed off, and didn't appear again until just a while ago this evening. He's using a new bios now, one called modbios3 he says. He says this one seems to scale correctly with CPU voltage, and he's currently doing 10x250 with 1.85v, prime stable, which is a good deal better. I think, factoring in the very high FSB, that this should be the correct voltage needed to attain this speed. Here are the details I recorded on the chip when I had it in my possession.

*Sixth Chip* 35w AQYHA 0413
XP-M 2400+ - 166FSB x 11 @ 1.4v = 1840mhz 38C max
XP-M 2400+ - 200FSB x 10 @ 1.4v = 2000mhz 39C max
XP-M 2400+ - 200FSB x 10.5 @ 1.45v = 2100mhz 41C max
XP-M 2400+ - 216FSB x 10 @ 1.5v = 2160mhz 42C max
XP-M 2400+ - 216FSB x 10.5 @ 1.575v = 2280mhz 43C max
XP-M 2400+ - 216FSB x 11 @ 1.7v = 2390mhz 48C max
XP-M 2400+ - 216FSB x 11.5 @ 1.775v = 2500mhz 51C max
XP-M 2400+ - 216FSB x 12 @ 1.925v = 2600mhz max

I guess it was just the beta bios he was using. Now with the new one, he's having a strange new problem. Like I said earlier, 10x250 works stable at 1.85v for him, but if he ups it to 1.875 just for giggles, sirens go off. Yet, 1.9v works. Why would one particular voltage setting not work correctly? This is some very weird stuff. Maybe he should have gotten a DFI so he wouldn't have to play with modded bios's :\

 

[Shimmishim]

um... so what bios was he using before?

 

[AWhackWhiteBoy]

"Maybe he should have gotten a DFI so he wouldn't have to play with modded bios's"

stock DFI bios, 236x11@1.825V

 

[Avalon]

He was online for one second to tell me that he just got the modbios3 and that it was working at the settings, and told me about the weird voltage problem. By the time I got to his message to ask him what he was using prior, he was signed off again. He's probably trying to reach 2.6ghz now. Maybe I should call him up and tell him it's going to be really damn hard on air cooling with a 250FSB.

 

[rogue1979]

Maybe it is a power supply issue. I have noticed when overclocking is pushing the power supply to the limit, it requires more cpu core voltage to remain stable.