Can't use high vcores

[ben rogers]

Hi there

Ive just got a DFI Ultra D motherboard but it seems my 3500+ Venice doesn't want to know about vcore >1.7v. If I set 1.73v (speedfan) it reboots instantly when I run SuperPI at at speed that is stable with 1.55v so its definitely the vcore that is causing this? has anyone came accross this before and know how to rectify it? I know the chip can benchmark at 3050MHz but now im getting worse superpi times than my old nforce3 due to it not liking high vcores.

I also wouldn't mind knowing when to set the jumper over to make the vdimm come off the 5v rail instead of the 3.3v rail as I am using 2.9v atm maybe this is causing a problem?

Thanks for any help.

[v0dka]

wtf..

Not meaning to be rude but what in godsname are you doing with 1.75v?

You are only producing excessive heat for very marginal gains, that's probably also the reason it's unstable. Or are you using phase?

1.65v is as high as you want to go for A64, above that it's mostly no use for OC, gives off excecssive heat, power consumption is high and last but not least more volts reduce lifespan fast.

[Lastviking]

Hi there

Ive just got a DFI Ultra D motherboard but it seems my 3500+ Venice doesn't want to know about vcore >1.7v. If I set 1.73v (speedfan) it reboots instantly when I run SuperPI at at speed that is stable with 1.55v so its definitely the vcore that is causing this? has anyone came accross this before and know how to rectify it? I know the chip can benchmark at 3050MHz but now im getting worse superpi times than my old nforce3 due to it not liking high vcores.

I also wouldn't mind knowing when to set the jumper over to make the vdimm come off the 5v rail instead of the 3.3v rail as I am using 2.9v atm maybe this is causing a problem?

Thanks for any help.

1.7v is about max if you use air....thats why its rebooting, the cpu dont like it

[ben rogers]

well using that voltage helped me get the world record on air for a 3500+ Venice by about .2 of a MHz.

the temps are only 32 idle so I thought what the heck I might as well see what it can do but it makes no difference. I was planning on trying 1.8v tbh.

no im on air cooling - a £3.50 akasa heatsink.

regards,
ghgh

[LIKMARK]

Fry That Thing!!

[Aerou]

I hope either dont need that cpu and want to really damage it or seeking whatever extremely unthoughtful purpose for it. 1.55V 24/7 is the max for long term usage, above that you are exponentially shortening the cpu's lifespan. Although I guess you know that, unless your shooting for the moon and doing some WR, it isnt wise

[ben rogers]

yeah , I don't use this kind of vcore 24/7, just for benchmarking the most I use 24/7 is 1.55 and even that gets hot. Added another 7MHz onto previous after this is doesnt want to know really. Tis the highest 3500+ Venice on air on ripping.org mind.

http://valid.x86-secret.com/show_oc.php?id=236806

[PhilDoc]

Hi there

Ive just got a DFI Ultra D motherboard but it seems my 3500+ Venice doesn't want to know about vcore >1.7v. If I set 1.73v (speedfan) it reboots instantly when I run SuperPI at at speed that is stable with 1.55v so its definitely the vcore that is causing this? has anyone came accross this before and know how to rectify it? I know the chip can benchmark at 3050MHz but now im getting worse superpi times than my old nforce3 due to it not liking high vcores.

I also wouldn't mind knowing when to set the jumper over to make the vdimm come off the 5v rail instead of the 3.3v rail as I am using 2.9v atm maybe this is causing a problem?

Thanks for any help.


Just as everyone says its your cpu not liking the high vcore. The tend to lose overclock if you get beyond their sweet spot.

You do realize that the DFI Ultra vcore is going to be higher than what you show in speedfan. Do a search on N4F in the mb vmod section. Your true vcore is about .04-.05v above what the programs states and it tend to gain another .01v under load. So you were probably giving it more like 1.78-1.79v.

Not to freak you out even further, but the voltage I get from pcwizard is .08v off in vcore. You could have potentially been running 1.82v on air. You're lucky you didn't fry your cpu. You should use the settings you have in the bios to calculate vcore.

I'm hoping that speed fan is closer to the correct vcore than other programs and I'm glad your cpu is still running.

[ben rogers]

thanks very much for the replies especially the last one.

I never realised how much the board is liable to overvolt as everyboard i've had on skt939 has undervolted....I wont be putting that votlage through it again - thats for sure - and its not like it made any difference.

it seems like I can acheieve what I can on 1.8v with 1.55v :S

the chip just isnt that great.

[PhilDoc]

thanks very much for the replies especially the last one.

I never realised how much the board is liable to overvolt as everyboard i've had on skt939 has undervolted....I wont be putting that votlage through it again - thats for sure - and its not like it made any difference.

it seems like I can acheieve what I can on 1.8v with 1.55v :S

the chip just isnt that great.


The board doesn't overvolt that much on vcore, just .01v at load. For the most part it gives you what you ask for in the bios. The chip that is used to monitor vcore or the programs that reports it, give too low of a value.

I did download speedfan and on my board it .5v low on vcore at idle. So you would have probably been at 1.79v.

Dont know about the chip. Suicide runs usually get better with increased voltage. Things like pi, may not. Either way its still the quickest air 3500+ at your site. :up:

[ben rogers]

it might be too late - now it wont even benchmark at 3022MHz and before it went 30MHz higher :|

its definitely taken a hit having that voltage put through it - tried the 133 divider but it just errors on SuperPI, need a new chip me thinks.:mad:

[PhilDoc]

it might be too late - now it wont even benchmark at 3022MHz and before it went 30MHz higher :|

its definitely taken a hit having that voltage put through it - tried the 133 divider but it just errors on SuperPI, need a new chip me thinks.:mad:


Sorry to hear that. Before you give up, let the cpu have a rest. If you're going to run it, run it at stock and at stock vcore. Dont really know, but if it got too hot you may have dammaged the TIM, so I'd remount it.

Good luck.

[Gothic]

Another thing is to check the memories... swap them just in case....

[gOtVoltage]

Oooopsie . the Venice highly Oced dosnt like Vcore above 1.65v and thats pushing it ALOT. My Orleans F2 3200+ was only stable up to 2.8ghz 1.68v core @18c idle on water. this screen was taken right after my run on a 1M @22c ambiant roomtemp. The chips rated max temp was 70c! It would not go any further although my temps were good@46c load,the CORE couldnt handle any more :dammit: Vcore!

http://i168.photobucket.com/albums/u192/darrellsierra/3200orleans2.jpg

[PhilDoc]

Another thing is to check the memories... swap them just in case....


Good point. :up:

[gianni-gt]

my wr-3500+ venice could handle 1.92V and my 3800+ alsmost 2V :D

[PhilDoc]

my wr-3500+ venice could handle 1.92V and my 3800+ alsmost 2V :D

Yep, but you were using a chiller. The colder you get the more volts you can give em.

[gianni-gt]

i've used dice :D

[PhilDoc]

i've used dice :D

That'll work too :D . Nothing like freezin something to have fun. Congrats on your wrs. I've been following them some on the X Overclocking forum. :up:

[gianni-gt]

should i mount my boxed cooler on my 3500+ and cracking the air wr from ben? :rolleyes:

[ben rogers]

should i mount my boxed cooler on my 3500+ and cracking the air wr from ben? :rolleyes:

lol

thats a bit harsh took me ages to get that....but if you must. :ROTF:

im suprised your chip still runs seeing as how much vcore you put through it to acheive the WR!!! lucky guy.

as a note, mine stil seems to work alright - it must have needed a rest or something :S as before I went to bed last night it wasn't benchmarking much above 3.00GHz but now its benchmarking at 3055MHz which it did before but no more - with any vcore.

the most I can get with my shoddy cooling is 3280MHz I think - and even then it will take me a few days to get it not to reboot instantly.

[gianni-gt]

my wr chip is dead ;) but i have another one here with 3571mhz... but i think i'll leave it...

[PhilDoc]

lol

thats a bit harsh took me ages to get that....but if you must. :ROTF:

im suprised your chip still runs seeing as how much vcore you put through it to acheive the WR!!! lucky guy.

as a note, mine stil seems to work alright - it must have needed a rest or something :S as before I went to bed last night it wasn't benchmarking much above 3.00GHz but now its benchmarking at 3055MHz which it did before but no more - with any vcore.

the most I can get with my shoddy cooling is 3280MHz I think - and even then it will take me a few days to get it not to reboot instantly.

Terrific, glad to hear everythings working as usual. Good luck on all of your OCing :up:

[gianni-gt]

@ben: in this moment i've sold my 3500+, you're not in danger. :)

[ben rogers]

cool

i didnt know whether to make a new thread or if you guys could help me?

ive got an old soundblaster soundcard in the PCI slot and I have to put my graphics card in the second PCI-E slot as it blocks the chipset fan if its in the first one...and it wont recognise the soundcard - Disabled onboard sound I am thinking the IRQ is shared for the PCI slot and the second PCI-E slot....reason I can only use one PCI slot is that the 7800GTX takes too much room up.

thanks for any help
ben:up:

[PhilDoc]

cool

i didnt know whether to make a new thread or if you guys could help me?

ive got an old soundblaster soundcard in the PCI slot and I have to put my graphics card in the second PCI-E slot as it blocks the chipset fan if its in the first one...and it wont recognise the soundcard - Disabled onboard sound I am thinking the IRQ is shared for the PCI slot and the second PCI-E slot....reason I can only use one PCI slot is that the 7800GTX takes too much room up.

thanks for any help
ben:up:

You probably have it right. There's probably an IRQ conflict. Sometimes on sound cards you can change the IRQ with a jumper, but I'm not sure about yours. I have the same equipment. Thought about moving the video card for the same reasons, but will leave it alone now, lol. What I have is 1950pro and a live 24bit. The 1950pro does cover the chipset fan, but doesn't seem to be causing any problems.

[ben rogers]

thanks for the reply!

I reseated the GRAPHICS CARD and now the soundcard works absolutely fine!??!

anyone can explain this as im most intreegeed!

im glad its all working now however.