Huge Problem with Overclocking, NEED HELP!

[Dynamix3D]

Okay I am having a problem right now with overclocking on my MSI K8N Neo2 Platinum. I cannot seem to get passed 230Mhz FSB Stable on my system.

My system specs are as follows:
AMD 64 3000+ Winchester
Zalman 7000B-Cu Heatsink/Fan
MSI K8N Neo2 Platinum
CM Real Power 450W
PQI Turbo PC 3200 Dual Channel 2 x 512MB DDR (2-2-2-5; Samsung TCCD)
WD Raptor 74GB
SB Live Audigy 2
BFG 6800GT

Okay so at first I was having problems getting passed 209Mhz FSB, then I searched the forums and people were telling me to lower the HT to 4, I did that, and it helped. I got all the way to 230Mhz FSB stable. Now as I tried to get passed 230Mhz, I changed the HT to 3, and it still crashes. I can make it all the way to Windows, but once programs started loading it freezes, and I have to reset the system manually. I have tried almost every memory timing you can think of, from the 2-2-2-5, all the way to 3-4-4-10. And everything is unstable and continues to lockup. Now I have seen many other people with this same memory getting speeds up to 270-280Mhz, I am having a hard time understanding why I cannot make it that far. My guess is the memory has to be defective possibly? I mean I woulda assume its the memory causing the problems since the system locks up once Windows starts loading my startup programs into memory.

Any suggestions?

 

[ts3433]

Run Memtest on the RAM; if it's OK at stock, run a RAM divider. You may also have to bump up your VCore a bit.

 

[canadageek]

hmmm...my first AT post....
what's you're voltage set at?

 

[Dynamix3D]

I've upped the voltages to 1.55, and then I even got the 1.41 modded BIOS where you can go higher, and it still is unstable at all higher voltages. I've tried up to 1.75v

 

[canadageek]

dunno...maybe you just got a wimpy overclocker?

 

[Dynamix3D]

Originally posted by: canadageek
dunno...maybe you just got a wimpy overclocker?

Do you think its the memory? Or the motherboard? Or the CPU? I am wondering if I should RMA the memory and get it replaced. I don't wanna refund it cause Newegg charges 15% restocking fee for refunds. Since everyone else has gotten 270-280 I was wondering if maybe I just got a bad batch...

 

[JBDan]

My guess is the MEM. Up the mem voltage to 2.7 and test then try 2.8 then test again. Make sure your RAM is rated correctly for these voltages b4 testing though. The increased voltage to my mem made all the diff in the world for me. You should leave your cpu voltage at stock as many reviews have gotten this chip to 2.5Ghz at stock CPU voltages. Your mem is rated at 3200 so DDR 440 or 450 would be considered a VERY GOOD oc. Run 2.5-3-3-10 timings and up the mem volts like I said and test memtest86. Prime95 should be tested as well. On my sys (3500+ Win) and A8V deluxe I run 225fsb x 11 (2.475)@ 1:1 ratio so mem is @ 450MHz > HTT at 800 or 4x with mem voltage at 2.8 (corsairxl is rated @ 2.75 volts) and cpu voltage is at stock (1.45)

 

[Loki726]

This could take along time to do, but will tell you what part of your system is causing you problems.

Do stability testing at stock speed (run prime95 or something like it for a few hours) to make sure everything is ok before you start.


Start with the fsb. Lower your cpu multiplier so the processor speed is never greater than the default speed (1.8ghz). Loosen your memory timings to something like 3-3-3-13. Use the memory divider to keep the memory speed as close to stock as possible. Make sure your agp and pci locks are on. Now raise the fsb until you find the highest speed your board is stable at.Also, make sure your HT is around 1000mhz, decrease the multiplier to keep it around this.

Next do your memory, keep the fsb and cpu speed at default, and use the dividers to find out how high your memory can go, raising the voltage each time you hit a wall. Use memtest for stability here. You can stop if it goes higher than your max fsb.

Finally do your cpu.

This might take an extremely long time if you do alot of stability testing, but it will tell you what is keeping you from overclocking very high.

 

[Dynamix3D]

Originally posted by: Loki726
This could take along time to do, but will tell you what part of your system is causing you problems.

Do stability testing at stock speed (run prime95 or something like it for a few hours) to make sure everything is ok before you start.


Start with the fsb. Lower your cpu multiplier so the processor speed is never greater than the default speed (1.8ghz). Loosen your memory timings to something like 3-3-3-13. Use the memory divider to keep the memory speed as close to stock as possible. Make sure your agp and pci locks are on. Now raise the fsb until you find the highest speed your board is stable at.Also, make sure your HT is around 1000mhz, decrease the multiplier to keep it around this.

Next do your memory, keep the fsb and cpu speed at default, and use the dividers to find out how high your memory can go, raising the voltage each time you hit a wall. Use memtest for stability here. You can stop if it goes higher than your max fsb.

Finally do your cpu.

This might take an extremely long time if you do alot of stability testing, but it will tell you what is keeping you from overclocking very high.

This is very good advice, and yes although it might take very long, I think its the only option I have right now, because I have no idea for sure what exactly is causing the bottleneck.

 

[ts3433]

Originally posted by: Loki726
Do stability testing at stock speed (run prime95 or something like it for a few hours) to make sure everything is ok before you start.

I had forgotten about that one batch of Winchesters that failed Prime even at stock speed, so this would be a worthwhile check. If you happen to have that week of production, see if you can RMA it to get a later batch.

 

[Dynamix3D]

Hey everyone thanks for the help, but I finally figured out the problem! Someone from ard|OCP helped me out, and told me that I needed to switch my SATA drive to slot 3 or 4 to lock it. After I did that, I was overclocking like a dream. Rigth now I am running stable at 270x9 = 2,430Mhz... 2.5-4-4-10 1T, stock CPU voltage, and 2.80v memory. Thanks again everyone I appreaciate it.

 

[JBDan]

Originally posted by: Dynamix3D
Hey everyone thanks for the help, but I finally figured out the problem! Someone from ard|OCP helped me out, and told me that I needed to switch my SATA drive to slot 3 or 4 to lock it. After I did that, I was overclocking like a dream. Rigth now I am running stable at 270x9 = 2,430Mhz... 2.5-4-4-10 1T, stock CPU voltage, and 2.80v memory. Thanks again everyone I appreaciate it.

Lock your SATA? How?

 

[Dynamix3D]

Originally posted by: JBDan

Originally posted by: Dynamix3D
Hey everyone thanks for the help, but I finally figured out the problem! Someone from ard|OCP helped me out, and told me that I needed to switch my SATA drive to slot 3 or 4 to lock it. After I did that, I was overclocking like a dream. Rigth now I am running stable at 270x9 = 2,430Mhz... 2.5-4-4-10 1T, stock CPU voltage, and 2.80v memory. Thanks again everyone I appreaciate it.

Lock your SATA? How?

LOL I have no idea how that works, but I did wha he said, and it worked. Moving the hard drive from SATA 1 to SATA 3 allowed me to overclock past 230Mhz FSB.

 

[JBDan]

Originally posted by: Dynamix3D

Originally posted by: JBDan

Originally posted by: Dynamix3D
Hey everyone thanks for the help, but I finally figured out the problem! Someone from ard|OCP helped me out, and told me that I needed to switch my SATA drive to slot 3 or 4 to lock it. After I did that, I was overclocking like a dream. Rigth now I am running stable at 270x9 = 2,430Mhz... 2.5-4-4-10 1T, stock CPU voltage, and 2.80v memory. Thanks again everyone I appreaciate it.

Lock your SATA? How?

LOL I have no idea how that works, but I did wha he said, and it worked. Moving the hard drive from SATA 1 to SATA 3 allowed me to overclock past 230Mhz FSB.

Wow ok maybe I need to try this LOL

 

[Dynamix3D]

yea try it and let me know if it works for u...

 

[DrCool]

you might want to check out MSI's official forums.. this is a well documented limitation.
SATA ports 1 and 2 are not clock locked, while SATA 3 and 4 are.. so as you up the FSB ports 1 and 2 are affected, while 3 and 4 are not.

http://forum.msi.com.tw/board....5cd18bc4c7273c997ad6fa

 

[JBDan]

Originally posted by: DrCool
you might want to check out MSI's official forums.. this is a well documented limitation.
SATA ports 1 and 2 are not clock locked, while SATA 3 and 4 are.. so as you up the FSB ports 1 and 2 are affected, while 3 and 4 are not.

http://forum.msi.com.tw/board....5cd18bc4c7273c997ad6fa

Does anyone out there know if the Asus A8V Deluxe's SATA ports are affected like this?

 

[Duvie]

Try this for all the tips for this board!!!


I am glad some things are being solved...However most of the problems ppl have is they are going about ocing all wrong. The key is to limit all the other components and isolate the cpu...You want to find its limits first. When you try to oc all components at the same time issues are hard to diagnos.

With this board

HTT=3x Take this out of the equation

Mem = 133 Take this out of the equation (wont make this an issue with pc3200 until 300fsb)

cas 2.5,4,4,10 and maybe even 2t...to take memory out of the equation

Vdimm = 2.7 this board undervolts a bit....

vagp = 1.55min again same as above

vcore 1.5v with no % over vid yet....again, again the board undervolts

Use highest multiplier available....


Then start climbing...issues that arise now will likely be more cpu related either needing more vcore, limit of fsb on the mobo, limit of cpu period, etc....

Once the limit is found for the vcore you feel comfortable with...then and only then...start messing with the ram. You may find that top speed cannot be had once with highest speed and best cas settings. It will take your testing to find that happy medium if it exist.

I wont go over my stability testing as I am clibing in fsb, cause I have gone over it countless times, but you should do something more then sandra test!!!


Edit: this has worked very successfully for me, and many have seen it work for me for the last 2-3 years....It is fast, reliable, and you dont see many problem post from me....I know the problem when it arises cause it can only be 1 or 2 things....

 

[JBDan]

Originally posted by: Duvie
Try this for all the tips for this board!!!


I am glad some things are being solved...However most of the problems ppl have is they are going about ocing all wrong. The key is to limit all the other components and isolate the cpu...You want to find its limits first. When you try to oc all components at the same time issues are hard to diagnos.

With this board

HTT=3x Take this out of the equation

Mem = 133 Take this out of the equation (wont make this an issue with pc3200 until 300fsb)

cas 2.5,4,4,10 and maybe even 2t...to take memory out of the equation

Vdimm = 2.7 this board undervolts a bit....

vagp = 1.55min again same as above

vcore 1.5v with no % over vid yet....again, again the board undervolts

Use highest multiplier available....


Then start climbing...issues that arise now will likely be more cpu related either needing more vcore, limit of fsb on the mobo, limit of cpu period, etc....

Once the limit is found for the vcore you feel comfortable with...then and only then...start messing with the ram. You may find that top speed cannot be had once with highest speed and best cas settings. It will take your testing to find that happy medium if it exist.

I wont go over my stability testing as I am clibing in fsb, cause I have gone over it countless times, but you should do something more then sandra test!!!


Edit: this has worked very successfully for me, and many have seen it work for me for the last 2-3 years....It is fast, reliable, and you dont see many problem post from me....I know the problem when it arises cause it can only be 1 or 2 things....

TY very much for that