• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Got The FSB To 504MHz!! (XP1600 overclocked to 2.27GHz Stable!!!

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
that's pretty amazing. however, I'm still skeptical. Many other people have that board, good memory, cooling, and vid cards, and have troubles going over 215 FSB. I can't believe that is a stable overclock. Most likely HD corruption would occur...blah blah blah.

OK, i'm really just jealous that I can't run at 250 FSB. :)
 
Interesting to see this. I can't tell if the memory was ran synchronous. Not impossible, but improbable...so you won't benefit any running the FSB at this spec- unless your doing it for the "Gee Whiz" factor.

-PC
 
NeoMoses said:
that's pretty amazing. however, I'm still skeptical. Many other people have that board, good memory, cooling, and vid cards, and have troubles going over 215 FSB. I can't believe that is a stable overclock. Most likely HD corruption would occur...blah blah blah.

OK, i'm really just jealous that I can't run at 250 FSB. :)

The PLL has been modified from what I can see...This way, the other bus(s) like the PCI will run in spec, so the HDD won't see corruption.

-PC
 
well jee golly whiz this guy proc is fast then mine @ 1/4 the cost........no wait...... that cooling cost 500 bux .... what dumbass hahahah seriously though...
 
Looking at this post, I thought it was actually oc-master that was doing the o/c.
 
NeoMoses said:
that's pretty amazing. however, I'm still skeptical. Many other people have that board, good memory, cooling, and vid cards, and have troubles going over 215 FSB. I can't believe that is a stable overclock. Most likely HD corruption would occur...blah blah blah.

OK, i'm really just jealous that I can't run at 250 FSB. :)

If its a segate drive it could handle it. Thats a 50mhz PCI bus that he is running. My PCI bus is at 49mhz. If my MB could could higher so could my other components.
 
PCphreak said:


The PLL has been modified from what I can see...This way, the other bus(s) like the PCI will run in spec, so the HDD won't see corruption.

-PC

Is there any way an average guy can do this mod? I wouldn't mind having a locked pci/agp bus now instead of waiting for an nForce 2 :D


Thanks

Tipycol
 
my name is obi wan knobi and i come from an alternate univers known as uranus... i come to suck ur blood
 
youthemandan said:
Just wait till i do my LN expiriment with AGOIA 0213. I can't wait to see what can get out of it.

Yea but would your pci/agp buses be in spec like the one in the review?
 
have to say this site looks like a photoshopped benchmark collage although the numbers all workout perfectly
 
Tipycol said:
Is there any way an average guy can do this mod? I wouldn't mind having a locked pci/agp bus now instead of waiting for an nForce 2 :D
Thanks
Tipycol

Sure. But when you say average, I'm not sure just how 'average your talkin':D!?
To modify the mechanics of the PLL's operation requires good de-soldering skills for starters. I'll try to give you a simplified overview as to how the PLL works, and what needs to be modified.
The basic function of a PLL or also called "clock synthesizer" is to produce an electrical pulse as an output, which is used in keeping the entire system's busses in time or in rhythm together and each at their respective frequency. The PLL has to reference it's output frequencies from somewhere also. The PLL's reference freq. come from the Clock Crystal, which is at a pre-defined frequency. It's at a pre-defined frequency because of physics. The Clock Crystal is just that: Quartz Crystal. Whenever a certain amount of voltage is applied to Quartz, it vibrates at a certain rate or frequency. Some watches and clocks keep their time this way, because it's very accurate.

So let's do a quick review:
The clock crystal produces a frequency for reference. The PLL uses this reference so it knows what to base it's output frequencies on. The PLL will only divide or multiply the clock crystal's frequency according to what it's output(s) need, and according to the way the PLL's been programmed for.
They only have a certain amount of programmed divisors or multipliers.
______________________________________________
EXAMPLE: (The PCI output has been programmed in the PLL to stay at 33 MHz)

If FSB = 100 MHz
then PCI = /3 (or 1/3 divisor to keep it at 33.33 MHz)

If FSB = 133 MHz
then PCI = /4 (or 1/4 divisor to keep it at 33 MHz)
______________________________________________

There is sometimes two PLL's onboard:
1) One of them does RAM, and
2) the other operates the FSB, AGP, & PCI.
We'll focus on #2.

Ok, now you may already be following what I'll soon be getting at..so I'll ask this question:


(Q.) What would happen if we replaced the clock crystal with another of a higher frequency value?

(A.) You essentially 'lie' to the PLL, so it's actual output frequencies will be higher than it would have with the original clock crystal.


This can be both good and bad.

Good, because your squeezing more of an overclock from your FSB.

Bad because some busses other than the FSB, will be too overdriven. Especially the time(hour becomes less than 60min), the PCI bus becomes out of spec, as well as the AGP. This leads to instability...


(Q.) So, how do you stop the other busses from becoming overdriven?

(A.) You intercept their input lines where they leave the original PLL, and wire them into your own PLL circuit you made on another circuit board complete with the original Clock Crystal you removed earlier.


So now you have the capabilty of having a set frequency for the PCI & AGP bus regardless of the FSB setting.

I do realize I haven't given a thorough overview (for the sake of time), and that in parts it may seem sketchy & hard to follow, but I'll try to update this with better explanations and examples....or pictures if I can find a host....

-PC
 
Last edited:
Yup even if it is genuine it cannot claim to be stable as there is no evidence of full loading - would have liked to have seen a 3DMark as well.
 
Last edited:
Back