![]() |
![]() |
![]() |
|||||
![]() |
![]() |
![]() |
![]() |
![]() |
|||
| Welcome
to Tech Support Forum home to more then 136,000 problems solved. Issues
have included: Spyware, Malware, Virus Issues, Windows, Microsoft,
Linux, Networking, Security, Hardware, and Gaming Getting your
problem solved is as easy as: 1. Registering for a free account 2. Asking your question 3. Receiving an answer Registered members: * See fewer ads. * And much more..
|
| Want to know how to post a question? click here | Having problems with spyware and pop-ups? First Steps |
|
|||||||
![]() |
|
|
Thread Tools |
|
|
#1 (permalink) |
|
Registered User
Join Date: May 2006
Posts: 8
OS: xp pro
|
OC'd system stability problem
Hi all,
ive got a problem with my system stability. Just decided to upgrade my dying 4 year old pc in order to play newer games better. i purchased two sticks of 512mb corsair pc3200xl cas 2225 to increase my ram to 2Gb. Ok heres my problem, now i have 4 sets of 512mb pc3200xl running my system it will quickly Fail Memtest when i OC the FSB to 240, but will pass if i only use 1gb (2 sticks @ 240) or use a FSB < 220(with 4 sticks). i originally thought it may be faulty memory, but it looks like it could be power problem, as this memory should overclock further . Has anybody else had a similar problem and knows how to fix it? im running p4c800e duluxe p4 3.2e s478 Corsair cmx3200xl cas 3-4-4-8 @2.75v (512x4) Antec truepower trio 650w His radeon x1950pro 2 seagate drives in raid 0 |
|
|
|
|
|
#2 (permalink) |
|
Mentor Hardware Team
Join Date: Aug 2005
Location: Florida
Posts: 2,619
OS: xp
|
Re: OC'd system stability problem
I would say you have a stick of ram, or two that isnt happy with
each other voltage wise, or may just not be compatible, or bad. It happens. You would want to be running in dual mode, so compatiblity is a true key. You would have a better chance at using 2 matched sticks to achieve this versus 4 sticks, Ive had fair luck using 4 matched sticks. Bout 50/50 for me...
__________________
|
|
|
|
|
|
#3 (permalink) |
|
TECH HARDWARE TEAM
|
Re: OC'd system stability problem
I've also had about a 50% success rate running 4 identical sticks of memory. I was able to calm down a couple of the systems by raising the northbridge voltage, but some boards just don't do a great job with northbridge cooling, so procede cautiously. When someone asks, I rarely recommend running four sticks because of this. Mike
__________________
![]() Speak kindly to me, beloved master. Revel in my unconditional love, and give me every minute that you can spare, for my time with you is short. --Your faithful dog |
|
|
|
|
|
#4 (permalink) |
|
Registered User
Join Date: May 2006
Posts: 8
OS: xp pro
|
Re: OC'd system stability problem
Cheers guys
looks like ive got 3 options. Which do you think is best? 1). replace the two sticks with a matched pair (as one pair already is matched). will i still have stabilty problems? 2) sell it all up and buy a matched (just two sticks) pair of 2gb DDR 500/550. will i still have stabilty problems if i go higher than 240 FSB? 3) or increase northbridge voltage. I could only find a setting in my bios called AGP VDDQ voltage which has the options 1.50, 1.60, 1.70, and 1.80v. which i think is tied to the northbridge. well i opened up my case, aimed a desk fan at my board and raised the AGP vddq voltage to 1.70v. i managed to run memtest for 4 hours with no errors. woo hoo! Do you think i should buy fan for the northbridge cooler or invest in a new heatsink altogether? and then try 1.80v or is this way too much!? |
|
|
|
|
|
#5 (permalink) |
|
TECH HARDWARE TEAM
|
Re: OC'd system stability problem
The AGP VDDQ is the video card slot voltage and I would not raise that if all possible. If it were my system and I was having instability, I would back off the OC and test my components to be known good at stock speed. Then slowly raise the OC and test along the way. The only voltages you should really need to raise are the Vcore(cpu) and the Vdimm(memory). You're best bet is to sell the 4 sticks and buy 2 1gb sticks and procede from there. The issue with all memory slots full could never come into play. While you are running the computer touch the northbridge heatsink, if it is getting real hot put a fan on it. I use old 486 cpu fans for this. You need to remember that every piece of computer equipment has it's own OC'ing stability limit, when you reach that limit, which it sounds like you have. Only voltage or different parts will raise that limit again. Do a google search for Corsair cmx3200xl and Vdimm voltage, there is a lot to be learned from other peoples trial and error. Mike
Old Review
__________________
![]() Speak kindly to me, beloved master. Revel in my unconditional love, and give me every minute that you can spare, for my time with you is short. --Your faithful dog |
|
|
|
|
|
#6 (permalink) |
|
Moderator, Hardware Team
|
Re: OC'd system stability problem
I agree with the fellas above / I have never hit a systems max overclock using 4 sticks of ram
alos if you are using a prescott CPU you had better watch your temps carefully those run hot already! in your shoes I would look for a matched 2gb pair (2 x one gig each stick) like the crucial ballistix DDR500 you will need to bump up the vdimm to 2.75 and probably the cpu voltage a touch too dont go crazy / if you burn up that board you wont like the replacement costs ! even used your board sells for more than it was new! you will need a good aftermarket cpu cooler and maybe even a better northbridge cooler what cpu cooler are you running now ???? 240 on the cpu clock is a hot clock for sure! you will really need to watch temps / have A+ air flow inside the case >>>> like two high RPM 120mm case fans >>>>> like 3000 rpm's >>>>> those will make your case sound like an electrical transformer; but ???????????? if will keep your system cool !
__________________
![]() IF YOUR HARD DRIVE DIED RIGHT NOW; WOULD YOU CRY ??? ASK US FOR HELP DEVELOPING A DATA BACK-UP METHOD Last edited by linderman : 02-21-2008 at 05:48 PM. |
|
|
|
|
|
#7 (permalink) |
|
Registered User
Join Date: May 2006
Posts: 8
OS: xp pro
|
Re: OC'd system stability problem
Think i might start looking for a nice new set of ddr500
Thanks for your help guys Im using a coolermaster hyper 6, it does a good job keeps my cpu cool at about 32 degrees idle, and 55 degrees under load. Anything higher than 57degrees triggers a reboot though!!!!, i thought prescotts could get hotter than that. ive also got two 80mm intake fans and two 80mm fans on the rear. i was also thinking about lapping my cpu to lower temps further, but havent got round to doing it yet. I also set the ram to 320 (5/4) and the max FSB i could reach was 247 so i think maybe is the most im gonna get to, but with some nice new ddr500 i should at least be able to do it with 1:1 ratio ![]() |
|
|
|
|
|
#8 (permalink) |
|
Moderator, Hardware Team
|
Re: OC'd system stability problem
check the temps in the bios too ???? it shouldnt restart until it hits around 70C
you may well & should be able to handle MAX work load temps of 60C with a prescott (while gaming)
__________________
![]() IF YOUR HARD DRIVE DIED RIGHT NOW; WOULD YOU CRY ??? ASK US FOR HELP DEVELOPING A DATA BACK-UP METHOD |
|
|
|
|
|
#9 (permalink) |
|
Mentor Hardware Team
Join Date: Aug 2005
Location: Florida
Posts: 2,619
OS: xp
|
Re: OC'd system stability problem
I ran a 2.8ghz prescott cpu with a p4c800-deluxe, basically the same
board you have. Used a zalman for the hsf, did a good job. At idle my temps would be around 108 degrees fahrenheit, but gaming it would stay at 165 degrees fahrenheit, for a while I was real paranoid about that temp, but as mentioned those cpus got hot. I was over clocked to a hair over 3.00ghz, ran fine for a few years of heavy use, loved the board, I ended up selling that system, and its still running today afaik...those early prescotts did get toasty....
__________________
|
|
|
|
|
|
#10 (permalink) |
|
Registered User
Join Date: May 2006
Posts: 8
OS: xp pro
|
Re: OC'd system stability problem
Ok think i may have found the problem
Ive read around and is seems that the asus probe temps may differ from the bios temps. i looked at my bios temperature section and found it was set to do a thermal active temp of 60 degrees, which may be why it cuts out at a low temp. ive now set it to 70, but i havnt tested it out yet |
|
|
|
![]() |
| Thread Tools | |
|
|