How much modding is really needed?

purrkur · Post by **purrkur** » Wed Dec 29, 2004 5:10 pm

I know I will probably get bitched at because I am going to question the need for modding but here goes

I am currently running a version 1.0 BP6 running two 533MHz Celerons at 576MHz. It is rock stable and it never complains about anything. It has absolutely no modding whatsoever, even the good old EC10 is standard issue. This board will even boot and run for some time at 600MHz before randomly giving in when a heavy task is thrown at it.

I bought a BP6 board version 1.1 with faulty caps. I bought replacement caps and started working on it immediately (so I have no idea if it works or not because I never tried starting it). So far all the caps that needed replacing are replaced. I have found that the replacement caps are a lot larger than the ones that are standard issue, making the replacements look somewhat silly on the board. But they are there just the same.

None of the caps around the ZIF sockets are damaged. My replacement caps are a larger than the ones on the board allready. I was thinking of replacing all of them but now I am having doubts because getting the old ones out is pretty difficult, because my equipment isn't the best to do this type of work. I am not that scared that I will damage the board but it is pretty time consuming work as well. Oh, and I have changed the EC10 on this board.

I intend to run two P3's on this board that I have bought (if it works). They are 866MHz /133MHz fsb chips (yes, I got the NEO adapters).

I read some articles and checked out a bit of specification for CPUs and found the following:

Celeron Mendochino core is built on 25 mn technology. The top processor tops out at 533MHz (what I have), uses 2.0 volts and has a power dissipation of 28.3W I am overclocking mine so I guess it is dissipating about 30 watts. So far I have no problems with my setup using this processor.

The P3's that I bought are Coppermines that were found in speeds from 533MHz to 1133MHz. They use a voltage 1.65 to 1.75 volts and the power dissipation ranges from 14W to 29W. These are built on an 18 nm technology.

If the top processor in that range tops out at 1133MHz and dissipates about the same as the top Mendochino Celeron then the question is, how much current does it draw? I haven't found figures for that. Yes, it has a higher transistor count but it is also built on a smaller transistor size and uses a lower voltage so I am not convinced that transistor count is a factor.

Given that my two processors are 866MHz then they should theoretically fall into the 22W area. Given that I will run the processors at 100MHz reduces the speed down to 650MHz or so, which should reduce the power dissipation from the CPU's. 18W or 20W?? I don't know.

So before I go into the time consuming job of modding the Q6 regulator, my questions are: Does the Q6 really need to be replaced?? I am simply not convinced. Given the statistics above, I am inclined to say that it is not really needed. Is it also worth removing the caps around the ZIF sockets (that all look fine) and replace with larger caps that might eventually be a problem when attaching a large heatsink? Has anybody at all tried running P3's (running at 6-800MHz) on this board without modding Q6 and succeeded?

Sorry for the long rant but I haven't found this issue being brought up on this forum at all so I thought it was time to do so. I don't really want to spend more effort on this mod than needed, and if it isn't needed then I don't want to do it. I got better things to do with my time

If somebody brings me evidence that shows that I am way off then I will do it. I have all the components allready so it is just a matter of finding the time to do it.

So what do you fellow BP6:ers think??

Dave Rave · Post by **Dave Rave** » Thu Dec 30, 2004 10:38 am

the ec10 mod is necessary in as afar as the board gets flaky.
checking the mbm hi/lo page, my board ran fine with a range of 1.34 to 1.67v (thereabouts)
it ran fine, until it didn't....
not to say that it was the ec10 that caused probs, might have been something else....
but if something is supposed to rev out to 6,000rpm and it only goes 5,500rpm, it's an indication there is a prob.

(i haven't done other mods, apart from replacing all the 1500 caps on two boards, and they both don't work properly anymore, yet, sigh)

hyperspace · Post by **hyperspace** » Thu Dec 30, 2004 12:02 pm

I searched the WEB for the original PIII modifications that were done by BBA and JJC on the BP6. I did find the original thread (link) on this, http://www.mrcs.net/board/mrcubb/ubb/Fo ... 00030.html , but it is dead.

Here's one link that has comments from BBA about the modifications (from this link, http://www.xtremepccentral.com/forums/s ... -3236.html ).

Code: Select all

BBA
I will argue against that point of view for a great number of people at BP6.com

The Abit BP6 was the original dual socket 370 motherboard, many had attempted to run dual coppermine P3's on it and failed...they said it was not possible. That is exactly why I made it work, and I was the first to do so, by directly modifying the motherboard cpu socket configuration. This eventually lead to other improvements in the BP6 such as the improved filtering by replacing voltage regulation filter capacitors with larger ones and discovering the voltage regulator IC being insufficient for the task of running stable even with a celeron on it, not to mention the P3. Changing that one vr chip was pretty easy to a skilled solderer and cost only a few dollars to do, much less than buying a new motherboard.

Maybe Derek can give me access to our original BP6 forum so I can try to find some specific information about modifying the BP6 for PIII operation (in regards to the Q6 component).

Personally, I really think it all depends on the "mix" of components on your mainbaord. Your mainboard's overall quality or component tolerances may allow PIII operation without changing Q6. I don't think ABIT tested or QA'd the BP6 running at 100 Mhz or more. So, they never saw the instabilties we have seen as BP6ers tried to get more out of this mainboard. The components ABIT used to manufacture the BP6 were fine for the processors that were available at the time.

davd_bob · Post by **davd_bob** » Fri Dec 31, 2004 12:03 am

hyperspace wrote: Maybe Derek can give me access to our original BP6 forum so I can try to find some specific information about modifying the BP6 for PIII operation (in regards to the Q6 component).

I hope he is "daring" enought to do it and to the "devil" with the risks involved.

hyperspace wrote: Personally, I really think it all depends on the "mix" of components on your mainbaord. Your mainboard's overall quality or component tolerances may allow PIII operation without changing Q6. I don't think ABIT tested or QA'd the BP6 running at 100 Mhz or more. So, they never saw the instabilties we have seen as BP6ers tried to get more out of this mainboard. The components ABIT used to manufacture the BP6 were fine for the processors that were available at the time.

My system gets 104FSB fine but at 108 only one cpu shows up. At 110 it hangs on reading the HD(at least I HOPE its just reading)

hyperspace · Post by **hyperspace** » Fri Dec 31, 2004 12:12 pm

davd_bob wrote:My system gets 104FSB fine but at 108 only one cpu shows up. At 110 it hangs on reading the HD(at least I HOPE its just reading)

Note from manual:

CPU bus speed above 66Mhz/100Mhz supported but not guaranteed due to the PCI and chipset specs.