How can I tell the Duron bridge opening/closing worked?

[RmoR]

I cut and opened appropriate bridges on my Duron 700mhz according to John Carcich's home page schematics at http://members.nbci.com/candjac/index.htm to the settings for 900mhz. The mother board is MSI 6330 (KT7 Pro), so I had to perform the surgery to change the multiplier. I used simple tools: an exacto knife and graphite. Anyway, the startup post says that the freq is 700mhz after the modifications were made. I did not up the voltage. The system seems stable and temp remains low. However, Quake3 timedemo gives much higher fps now for same graphics settings (after: 800x600 32bit col/tex, high det/tex=58.4fps before: around 43fps w/ Diamond V770 TNT2 32mg). If the chip's multiplier is overclocked, then why is the freq not posted correctly? Or am I imagining things or got my fps values wrong? I.e. is it the case that the post HAS to be correct? I matched FID and BP_FID settings at 9X, and left the L1 bridges open in accordance with the suggestion on John's site to the effect that closing L1 bridges on this particular mobo allows the mobo to interfer with the bus for the other bridges. Any input from the crowd, please?

 

[MrDudeMan]

i have no experience with this board, but IMHO i think if it doesnt post, it isnt higher. if no one else knows about this, then id suggest looking over your surgery and checking the mulitipliers. also, the voltaged probably needs jacked up a bit because it will be needing more "umph"


just my .02...hope that helps

 

[kponds]

go here and check this out. It works for me (posts)



toms hardware

 

[RmoR]

I checked out that page from Tom's before too. It's got the same settings, albeit in a simpler format, as John's page. I did close and open as per specs which correspond to both Tom's and John's. Anyway, I've heard that MSI's posts are inconclusive, that in some cases, they simply don't post the correct frequency. Anyone know if that's true? Wcpuid even says 700. So does wcpuid use registered values or does measure actual speed? If it's still @ 700, then I guess I've got to open 'er up and try again. EEK! I hope I don't scratch the bridge posts right off!! Anyway, if anyone has any other suggestions or information, I would greatly appreciate it.

 

[johncar]

Rmor,
Was going to suggest wcpuid or wcpuclk as a check, but since you have already done so, that proves the mod has not taken effect as wcpuid measures actual speed. You can prove that by clicking the refresh option and see the last character, (.nn), change occasionally. We can send you wcpuclk which displays measured mhz in realtime in a small box on desktop...(saves refreshing wcpuid). Go to John Carcich's site and email him via the link.

Suggest you test your mods before you redo them. See page 2 at the Duron OC article...it suggests how to test for continuity using a multimeter...esp by NOT probing at the bridges where you can destroy the bridging material, but by probing from appropriate cpu pins to the ccorresponding 1K resistors connected to closed L3, L4 bridges for BP_FIDs...and from appropriate cpu pins to Gd pins for L6 bridges for FIDs. All info re pins connected to L3, L4, L6 bridges and Gd and Vcore pins is on the pages at the site.

You tried to change the bridges controlling the 2X "Adder" from LO to HI. So you only have to test 3 circuits...one on FID 2X L6, and two on
BP_FID 2X L4, (where you tried to open the closed one, and close the open one). After your continuity tests, suggest not to use problematic pencil closing material...use the rear window defogger repair material... and retest circuits before re-assembly.
Good Luck,
John C.

 

[DaddyG]

I have a sneaking suspicion that your MSI mobo is limiting your OC. Your CPU is identified as a Duron, and Durons are not available at 900. One other suggestion would be to make sure that you clear CMOS completely, some settings may be messing up the OC and the mobo defaulting to some kind of failsafe mode. Connecting all your L7s will kick up the default voltage to 1.85 which might help.

 

[RmoR]

Thank you all for the input. I will try again when time is not so important, like on the weekend, and I'll break down and buy new battery for my multimeter. I appreciate the suggestion for the conductive pen. I shall try that too. The only reason I used pencil was that a friend of mine recently did exactly that and it worked the first time. Naturally I assumed that it would be a breeze. Anyway, thanks.