RAM timings Aren't stable (Weird) OCZ PLatinum

DoOb · Post by **DoOb** » 08 Oct 2006, 08:04

Here is my setup

AMD ATHLON 64 3200+ @ 2.62Ghz @ 1.55v
WINFAST NFORCE 4 6100K8MA-RS 939PIN
AOPEN 19" 75HZ CRT
2 x 512MB OCZ DDR466 PLATINUM EDITION 2.5-2-2-5 @ 2.85v
LEGEND 7800GTX 256MB
LOGITECH X-620 6.1

Ever since i have put the DIY waterooling on my Chip my pc isnt stable at the above timings. But is stable at 3 3-3-6 (Ran it last night for 7hrs of prime95) and i think 3 3-3-5

What is the problem. I am even running my stock cpu fan on the ram and chipset and still not stable. So heat isnt the problem.

What could it be. This ain't right. My pc had 14hrs of Prime95 and it was rock solid at the above timings with my stock cooler

Can anyone tell me what the problem is.

DoOb's

*fixed title - jee*

DAE_JA_VOO · Post by **DAE_JA_VOO** » 08 Oct 2006, 08:36

Well, this is simple:

Before cooling = stable
After cooling = unstable

Obviously the cooling is a factor here. What temps are you getting on the CPU?

DoOb · Post by **DoOb** » 08 Oct 2006, 09:06

Hey Dae thanks for the reply

heat isnt the problem. I have a fan blowing on the chipset and the ram modules. And those timings i have in my sig were stable with stock cooler.

Is it possible that the block i have (K7) is the problem it cant take the amount of heat on the surface area of the chip. But it is idling at 30 degress.

Chip is at 1.55v and ram is at 2.85
Memory bus is 166mhz
Hyper transport is at 4X
Ram is nw at 2.5 2-2-5 timings

Cool let me know DAE

Post by **Iceblade** » 08 Oct 2006, 11:12

ram @ 2.85, WHY

Colin_69 · Post by **Colin_69** » 08 Oct 2006, 12:58

I have the same problem with my RAM now Doobs, thats why I pushed the FSB and raised the timings. Its because our cpu's are overclocked so high. If you have a higher clocked chip like 4000+ or whatever, the RAM will be stable at the normal timings.

I just dont know why its only happend now though. It took me a while to figure out my problem because my pc seemed stable.

DoOb · Post by **DoOb** » 08 Oct 2006, 20:03

Exactly Colin

WiK1d · Post by **WiK1d** » 08 Oct 2006, 20:12

Just drop the devider on the RAM to 150 or something..

Post by **maxxis** » 08 Oct 2006, 20:54

How on earth can a cpu cooler affect ram timings?

Redo your overclock. That should show you where you went wrong.

DoOb · Post by **DoOb** » 08 Oct 2006, 21:02

Hey wik1d can only drop the memory divider it in these settings:

100 133 166 200 233 250

Hman · Post by **Hman** » 09 Oct 2006, 08:27

Try 166. Maybe you have a small air bubble in your cpu block, wich would cause localized heat problems. Maybe the block is crap, try another, or go back to air cooling.

DoOb · Post by **DoOb** » 09 Oct 2006, 08:41

Ok i am back on air cooling and not even those timings are stable anymore. Temps are too high now aswell like 40 idle so i dont know what the heck is happened here. My north bridge , i took the heatsink off and i am seeing a ring of heat moisture so i am thinking its the board. i cant handle it anymore.

Hman divider was running on 166 all this time

Hman · Post by **Hman** » 09 Oct 2006, 08:49

You must have spilled water then. Take everything apart, and let it dry for a few hours. then try again.

DoOb · Post by **DoOb** » 09 Oct 2006, 11:50

I actually did funny enough but that was after this happened. And it wasnt much just a lil bit.

I dont know i cant even run timings of 3-3-3-8 on a oc of 2.6ghz so i am back to stock everything. I think the borad couldnt handle so it is messed up internally.

DoOb · Post by **DoOb** » 09 Oct 2006, 11:52

O well Thanks for everyones help you guys are stars.

Hman · Post by **Hman** » 09 Oct 2006, 12:46

Maybe it's your cpu, as the mem controller is onboard it.

Hex_Rated · Post by **Hex_Rated** » 09 Oct 2006, 13:06

You could try run your FSB at 233 (you must have it at 262 now to have your CPU at 2.62Ghz) and set your divider to 200. Then try use the RAM's rated timings so you will be running the RAM at DDR466 which is what it is made for. Running the CPU at 2.33Ghz will also determine if the CPU mem controller is not coping at 2.62Ghz.

You could also give a bit extra VMEM voltage, but don't go over 2.95V.

Hman · Post by **Hman** » 09 Oct 2006, 14:09

Or give the ram to me. I'll see if they're at fault.

Fcon_Vpro · Post by **Fcon_Vpro** » 09 Oct 2006, 18:27

Why such high volts for such a low overclock. I doubt you need more than 1.45V to get that CPU to that speed.
I think the issue here is simply heat.
Lower the volts on the cpu and see what happens.
Mr Bean had a good explanation about this a while ago... something about heat on the transistors in the cpu for even the smallest time period of milliseconds can cause a pc to become unstable... even if the overall temp is pretty low on the cpu.

I suppose the best way to look at it is to pretend u "idling" around a bon fire or basically you are standing a distance away from it. You can still feel the heat but not much. And you moving towards the fire is the same as loading the CPU. Even if you were to jump into the fire and jump straight out within a second, its still gonna hurt. And thats the same with your CPU and that high voltage. Sorry if the analogy is bad, but its the best way to explain it.

Fcon_Vpro · Post by **Fcon_Vpro** » 09 Oct 2006, 18:38

One last thing... change your HTT multiplier to 3x... you should try not to exceed 1000Mhz... either less than or exactly a 1000. Only with really high end boards like DFI and maybe ASUS can you get higher than 1000Mhz stable. 262x4>1000 where as 262x3<1000. Try the 3 multiplier.

DoOb · Post by **DoOb** » 11 Oct 2006, 07:57

Ok thanks will try that