Result of oc celeron300a@580 MHz.

Result of overclock celeron300a@580 MHz.

I have one celeron300a, slot1 type, OEM from Costa Rica.This is the old celeron just don't like the lastest PPGA celeron.That claim to be overclockable like 558 MHz and 580 MHz.or 600 MHz.This chip is good one it can run at 504 MHz.with retail heatsink(I have retail heatsink) It can't run at 527 Mhz. with conventional air-cool heatsink so I tried to use peltier and sandwich it.With 40*40 mm peltier in the front and another 30*30 mm. peltier in the rear,I can put this chip up to 558 MHz at the temperature about 3-5 celcius but couldn't do more than this.

After playing with water cooler, I think that it is very effective way to get hot temperature out of the chip,especially when using with peltier.So I decided to make one water cooler for my celery.Making it one isn't a hard task, if you have some idea and tools.My cooling is made of one water block and a 94 watts peltier in the front and AAVID heatsink in the rear to make it like a sandwich cooler. I didn't think that using a small heatsink may help in cooling but I used it for reduce the stress that will happen when I attached water block with PCB.I didn't want to bend it. The water pump I used is for fish tank.It was 120 gallon per hour.Then I installed in my Bx6-2 mobo.I really love this board due to it's ability to read internal temperature diode. That I think it is more accurate than using probe.I set the FSB to 112 MHz.at 2.2 volts and boot the window. Then I waited for the peltier water cooler to do it's work.I saw the temp drop to 0 celcius and to -8 celcius that was the maximum temp that I could get.

cooling

At the lowest temp that I could get,I saw the ice from at the cool plate.So I changed the FSB to 124 MHz. At 558 Mhz,the system ran fine no sign of instability.Then I adjust the FSB again to 129 MHz,it failed. I restarted again and got to the soft menu II and increased the voltage to 2.4 volts and rebooted . By changing voltage,I could get to 580 MHz.At 133 MHz the system was stable.Then I adjusted FSB to 133 MHz.,the system was unstable although I disable l2 . The temperature should lower than this point to run this system stable at 600 MHz.I think may be -20 celcius.To lower the temperature to this point isn't easy.Now I think of phase change system,which can lower the temperature to that point.Condensation was the problem.If you like to play with them don't afraid it too much.There are many ways to prevent it. As you see in other sites,they close the system by using cellfoam.I will too.I will spent time to build the close system and will post you the results soon.

Below is the benchmark using Sandra Sisoft at 580 MHz. and Cpuid at 600 MHz.

My system

Cpu : celeron 300A ,sl2wm OEM,Costa Rica
Mobo : Abit bx6-2
Ram :Micron pc 133 ,128 Mb
Vga :TNT2 m64 ,32 Mb
Harddrive :Samsung 2.1 Gb 5400 rpm
Cdrom :Creative 24x
Power : two 250 watts