Printable View
If the cooling is adequate for 1.4v then I don't see the problem, I've run 1.6v for months in the past and it never bothred my X48 chipset as long as I kept temps sane, same goes for P45s. Asus P5Q Deluxe for example has atrocious NB cooling and that board doesnt like more than 1.34V or so. I've got a Thermalright HR-05 sli/ifx on my DFI X48 board and even when running ~1.5v my temps are 37-38c idle, so NB voltage is dependent on temps being kept stable.
hello~ i am a newbie, just got a q9650 with rampage formula
but there is a problem, i have 4hdd where 3hdd have vista in it
so i found that if i need to boot correctly, i must disable the 2 hdd drive in the boot drive selection
otherwise i get nothing showed when i power on the computer
now the problem is, when i try to overclock, no matter whats the setting is, i can boot after the first time modification but not the second time
like sometime i can boot up with 467*9 and i want to set the voltage lower so i enter the bios again, when i set the voltage lower i cant boot up
so i thought maybe the voltage is not enough, then i clear cmos and re-do the above thing, on the second modification, i change nothing in the bios
but it just cant boot up, so whats the problem?
i even tried disabling the other sata drives except my main hdd, putting 1.42v for vcore, 1.45 for nb etc
so i thing the voltage should be enough, i just cant figure out why the bios only boot once
also is there some example of q9650 with a rampage formula?
i just know how to set vcore, nb v, pll and vtt, want to learn more
thanks
The saying goes something like "Only jump in the deep end of a pool if you aren't going to drown!". Best thing to do is start in shallow waters and slowly progress to the deepest of waters as you get better until you master the waters that once mastered you.
Same goes for overclocking if you are a newcomer, those of us who like to help out the new guys all started at the shallow end and gradually learnt to swim and now we are the kings of our pools :D
All boards, cpus and memory behave differently so you need to get a feel for your hardware and more importantly figure out what your baseline voltages and settings should be. What works for one person with similar hardware may not work at all for you, and this is a testament to how small manufacturing flaws give you big headaches!
Your baseline needs to have Vcore, Vnb, Vsb, CPU GTL, NB GTL, Vtt, CPU Clk Skew, NB Clk Skew, DRAM Freq, and Mem Divider.
400Mhz FSB is a perfect frequency to setup your baseline on, after you've of course made sure it all works at 333mhz :)
All voltages should work at either nearly default, or slightly raised at this FSB and should need small adjustments at most to give consistent error free results.
400FSB, I'd use 1066mhz or 1200mhz DDR2 freq (if mem clocks to 1200 else 1066), and if 2GB sticks, 5-5-5-15-6-3-55-3 (i think it is, tRTP = 3, tRFC = 55, tRRD = 3), and default mem voltage on sticks.
CPU GTL should be 0.67x, NB GTL 0.67x, Vtt = ~1.14-1.16v, Vcore = VID (Auto, ~1.24v), Vnb = ~1.30-1.35V, Common PL (perf level) = 9, AI Clock Twister = Auto, Vsb = ~1.10v, Vsb_pll = 1.55V, Vcpu_pll = 1.50v, CPU Clk Skew = Normal, NB Clk Skew = Normal.
From here you work your way up FSB in steps, I'd go at steps from 400 -> 422 -> 438-440 -> 452 -> 466. While setting up FSB frequency and voltage change CPU multiplier to 6 or 7x so Vcore is isolated and CPU will post/work stable at default or slightly higher voltage. Vnb should always be set higher than you are aiming for, a good 0.04-0.06v above Nominal required voltage will give you headroom for Performance Level, and lowers the chance of Vnb being too low or refusing to post / locking during post. After you've got FSB + DRAM freq/timings stable, you can play with Perf. Level, at 400FSB PL = ~7 is a good start point, 420 - 440 FSB PL = ~8 , 450 - 480FSB PL = ~9, and go lower once Vnb is stable and no errops in tests. PL = ~5 is general limit at 400, PL = 6 is general limit for 420-450, PL = 7, 450 - 500. Rest depends on Vnb, NB temps, DRAM stick quality, motherboard manufacturing flaws, etc.
thanks so much for helping meQuote:
Originally Posted by mikeyakame
but the problem is not how deep or how hard i overclock
is like what i said, i have a setting which can boot up at first
and then when i enter the bios for second time, even i changed nothing, it doesnt boot up
even i just raise i very little step in fsb and putting a lot of voltage too
i am wondering if its my hdd problem or what
since i have 3 hdd with vista on it and dont know if it has problem on the boot priority
but i have tried to disabled all other sata drives and left 1 hdd only, but the problem continue
now i knew whats the problem
i tried clear cmos and get post on first time, i enter bios and CHANGE NOTHING , and select save and reboot
then i got no post, the lcd poster shows "EN SETUP", it shows "EN SETUP" when i enter bios and it never shut down even the computer reboot
i tried 0902 bios and 0803 bios, tried putting only 1 stick of ram in 4 different slots, tried plugging only 1 sata hdd
so is my mobo broken?
thanks
Yoh Mikeyakame,
you mentioned a point about the PL where I am still wondering even being able "to swim in deeper waters and avoiding the contact to seewasps". :cool:
There is a relation between the PL and the fsb you want to run, thatīs for sure.
The AnandTech site makes it really clear: http://www.anandtech.com/mb/showdoc.aspx?i=3208&p=1
But the point I am still wondering about is if I can take the AnandTech. article just for my P45 chipset. Right now I am running a fsb of 470 MHz and a PL of 8. If I lower the PL to 7 my system wouldnīt start.
If I had a X48 Chipset the formular says I could because: [tRD-(tCL/N)]>[x/N]
N= FSB-RAM-Divider => 7-(5/4)= 5.75 is greater than 2.75/(5/4)=2,2
2.75 is for fsb> 466 MHz.
Is it possible to run a PL of 8 together with a fsb of 470 by increasing the vnb?
My vnb right now is 1.2950v (=1,26v real), CPU VTT is 1.23v.
Do you think I should just force the PL of 7 increasing the vnb?
Well considering I was running PL7 with 2gb sticks and dram clk skews at cha d50ps, chb norm with ~1.55v Vnb at 481fsb until i got sick of fixing it up everytime bios broke the clock skews but it was stable and no problems at all! the problem was the rf bios screwing up mchbar registers.Quote:
Originally Posted by Fritz_the_germ
Maybe the flash went wrong.:shrug:Quote:
Originally Posted by derek0309hk
Disconnect and remove everything that you don't need to post your board. Use one stick of RAM.
Remove the power cable from your PSU and remove the battery from the board. Push the Clear CMOS button for 1 minute and leave the board without power for at least a half an hour.
Power up your board and hit F2 to load setup defaults. Restart your computer and enter BIOS Setup. Flash your board and shut it down.
Do a Clear CMOS again after flashing your board and Load Setup Defaults again afterwards.
All you can do now is hope that the problem is gone.:D
I think that he means that it doesn't work on his DFI board. If I'm not mistaken this is a problem with the P45 Chipset and the Maximus II has that problem too.:shrug:Quote:
Originally Posted by mikeyakame
You have got the point. The question is if I can use this calculation for my p45 chip. I think these values are for the X48 only.Quote:
Originally Posted by A-Grey
However is it possible to run a lower PL (7 instead of 8) by increasing the nb voltage and maybe the fsb vtt?:confused: (please donīt tell me "just try"- I am sick to set the jumper back and forward adjusting all my bios options again and again each time I try and the system wonīt f****** start).
thanks for your suggestion but no luckQuote:
Originally Posted by A-Grey
i removed the psu cable, press the clear cmos button for 1 min, take out the battery, remove all drives, left only 1 stick of ram and 1 display card, flash to bios 0902 in ezflash, no luck
the problem seems to be when i save the bios and restart, the power to the board is not cut completely, so the motherboard stay on the status that is entered the bios but the power require it to restart sth like that, i dont know very much
but it is very frustrating that i cant do any overclock or i search some method on the net is that cut the power directly but not pressing the cmos button, doing it 1-3 times the computer can boot up with saved setting, but i dont want to do it cause i am afarid it hurts my parts
so i may send it back for fixing it, but i dont want to be with out computer for days:(
Here's your answer to your question:;) http://www.overclock.net/intel-mothe...rding-p45.htmlQuote:
Originally Posted by Fritz_the_germ
Thank you Alien-Grey,
this makes a P45 interesting, doesnīt it ?
P.S. I am still checking around for the Q9650 with the special batch.
Do you have an link for downloading LinX 0.6.2 ? The version 0.6.0 had 64 bit bug.
x48 > p45. Newest version of Linx you will probably find in its thread.Quote:
Originally Posted by Fritz_the_germ
I don't have a newer version. I'm still using an older version.Quote:
Originally Posted by Fritz_the_germ
Hi guys, my Rampage Formula died last night i think, i get this message after the checking devices: USB overpower detected, shutting down in 15 seconds, i tried to unplug everything, even to disamble the entire pc, piece by piece, asamble directly on the table only with the power cable, monitor and minumum stuff conected to the mobo and still i get the same error, reseted the bios couple of times, nothing helped, the only option right now is to send it to RMA, my question is if it will be better to change it with a P45 like Maximus II Formula or should i stay with Rampage Formula, i wasnt so satisfied with my old formula because i manage to get only 450 fps with quad and 1,55 on NB, i know P45 are better on quads.
DaLiu can you enter flash utility by pressing ALT+F2?
I didnt tried that, i will try it tonight, if i manage to enter in the bios what should i do?Quote:
Originally Posted by kuebk
If you can get in the BIOS try with disabling everything in the USB Configuration. You can try to flash the BIOS afterwards.Quote:
Originally Posted by DaLiu
I think you are right, i remember that the computer booted one time to load the windows, it freeze on the booting screen, i reseted and after that i got the error about USB overpower detected, so probably could be a bios error, the only question is if i manage to enter in the bios, probably if not i need to send the mobo to RMA, to bad i liked this mobo, i hope i can fix it. I update you probably tomorrow morning.Quote:
Originally Posted by A-Grey
Heh, my DFI DK X48-T2RSB Plus also has great NB cooling. Doesn't exceed 45*C full load at 1.78v. However I have benched at 1.55 vNB on both my P5Q-Deluxe and my Maximus II Formula, and although the temps are around 50 - 60*C under load, they behaved perfectly.Quote:
Originally Posted by mikeyakame
You can certainly run PL8 at 500 FSB on a P45, quite easily in fact. My M2F runs 500 FSB / 600 MHz ram (DDR2-1200) / PL8 / Clock Twister Moderate / static read and dram read training enabled @ 1.25 vNB totally stable 24/7. You are of course limited in your PL choice by the memory divider and CAS latency, and increasing the vNB will not allow a lower PL to be used when limited by these parameters.Quote:
Originally Posted by Fritz_the_germ
i will take my board for rma later and maybe seek for if i can change to m2f or not
now i have another problem
when i was using my old com, which has a realtek sound chip
i can open as much audio-output-programme as i want
but when i switch to this board which use the suprexfxII and switch to vista sp2 from sp1
i lost my sound when i open about 3~5 audio-output-programme, i need to stop and restart the windows service audiosrv
also i can enable microphone input and stereo mix at the same time with realtek chip
but now i cant
so it this the sound chip problem or because i switch to sp2
if i buy a creative sound card can i solve my problem?
cause i dont know what sound chip does creative use and i dont find a single sound card on the market which use realtek chip
thx a lot
Originally posted by Cryptik:
"You can certainly run PL8 at 500 FSB on a P45, quite easily in fact. My M2F runs 500 FSB / 600 MHz ram (DDR2-1200) / PL8 / Clock Twister Moderate / static read and dram read training enabled @ 1.25 vNB totally stable 24/7. You are of course limited in your PL choice by the memory divider and CAS latency, and increasing the vNB will not allow a lower PL to be used when limited by these parameters".
Another problem is that I donīt have a 3:4 divider to oc my mem. pretty easiliy. The next possible divider is 5:4. I donīt think that the CAS latency of 5-5-5-15 limits my system. If I want to run my mem. sticks on default I could adjust a fsb of 500 MHz (=600 MHz mem. speed).
I have got a fsb of 490 MHz prime95 stable firing 1,39v to the nb. Next week a will be on vacation so there is no way out to hit my goal.
@Alien-Grey,
if the P45 chipset ist better for quad, why donīt you get a P45 plus a Q9650?
Still trying to get your batch. ;)
I asked a friend of mine about our seeking. He told me the certain batch doesnīt mean the cpu will exactaly work like you suppose. He knows two guys who bought a cpu with the same batch and still the cpu behaviour is really different.
Maybe you are right. All CPU's overclock different. Maybe I should just order one and hope that I'm lucky this time.:yepp:Quote:
Originally Posted by Fritz_the_germ
EDIT: Wish me luck, I've just ordered my Q9650.:D
Alien-Grey,
I wish good luck for your new cpu. Some dealers didnīt even answer my mail.
If there is any response I will tell you by pm.
If you want to stress your new q9650 just use the new version of LinX: LinX 0.6.1
Quote: