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

Overclocking woes with Q8200

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

ArcturusVi

Member
Joined
Apr 24, 2009
Location
Yawn, Ohio
I have an M1 revision Q8200 overclocked to 3.5Ghz.

500x7 multiplier. Voltage at 1.3. Any less and Windows would eventually hang.

Memory is G.Skill 1066 running at factory timing, volts, but is at 1000 because of my board's ratio settings.

Mobo is a Gigabyte EP45-UD3P F9 revision BIOS

CPU HSF is Xiggy Dark Knight lapped with AS5.

I have my North and South Bridges set to auto everything, PCI-e clock locked at 100Mhz.

The problem I am having is Windows has been hanging up on me lately, and it seems to only occur when I am running CPU intensive programs like F@H, or just to see if it was stability, prime95. I left it on last night with just AIM and F@H open. Woke up this morning and it would not turn my monitor back on. No picture. Other times it would just lock up. No sound, no cursor, had to do a hard reset. =/

Thanks.
 
Post a pic. w/ EasyTune6 open showing all of your current voltages. Also post a direct link to your specific RAM, and include the quantity (2 x 2GB, 4 x 1GB, etc.) Also post the VID of your quad using Real Temp-> "Settings" page--> Max Core VID.
 
can you download Coretemp and run it, then post a SS? if im reading realtemp right the VID of the cpu is 1.1875v, coretemp is to just double check that voltage.
 
I have a Q8200 with the exact same family, ext family, model, ext model, stepping and revision and my max VID shows up at 1.2 how come we have the same chip and our VID differs?

coretemp used to show 1.2 as the VID but now when i look it shows 1.15, the minimum VID in realtemp?

im sorry if you see this as a thread jack if you do ill make a new thread, just seems weird that we have the same chip model numbers etc and yet different VID ?
 
Those 8200's are tough to clock. That being said I think you are going to need a pretty big boost in VTT/NB voltage to get that thing anything close to stable @ 500Mhz FSB. Might want to start back around the 400-450 range and see if you can work up from there. Thats how I always do it, I know some work from the absolute top down, but I prefer to go the other way.
 
Those 8200's are tough to clock. That being said I think you are going to need a pretty big boost in VTT/NB voltage to get that thing anything close to stable @ 500Mhz FSB. Might want to start back around the 400-450 range and see if you can work up from there. Thats how I always do it, I know some work from the absolute top down, but I prefer to go the other way.

VTT? It seems stable now. Hangups were more frequent when I had my 275 up a bit. Either way I want to get this stable and able to last as long as I can, so any advice is much appreciated. Thanks.
 
In order for EasyTune 6 to show all of the voltages, click the "Tuner" tab, then the "Voltage 1" tab, and the "Advance Mode" button.
 
In order for EasyTune 6 to show all of the voltages, click the "Tuner" tab, then the "Voltage 1" tab, and the "Advance Mode" button.

Thanks, I'll get some more prints then.

VTT

-taken from http://forums.extremeoverclocking.com/t282072.html

Basically the higher your FSB frequency, the more voltage you are going to need through it.

Makes sense. Thanks. More screens up in a bit.

Definitely a stability issue. Just have my GPU client of F@H running and it's fine. CPU client and it gets sketchy, even though I only have the basic.
 
Well right off the bat... lower the PLL voltage from 1.85V to the default of 1.5V (overvolts to 1.54V at idle, and 1.57V under load), and no more than 1.57V. Lower CPU Termination voltage (VTT) from 1.60V to 1.36V - 1.4V. Try a Vcore of 1.35V and up the vNB / vMCH from 1.40V to 1.45V.

Since you're increasing vNB and increasing VTT, you'll need to adjust both CPU Reference and MCH Reference voltages. For a 45nm quad, CPU Reference should be 0.6333% of whatever value you set CPU Termination to, and MCH Reference should be 0.691% of MCH Core.
 
3.50GHz is quite a bit for Q8200 for it to take and stable. My magic number is 3.40Ghz.
I have Q8200 clocked at 3.40GHz & Q6600 clocked at 3.52GHz, both of their Windows Index Experience are rated at the same score 7.4

BTW, sometimes you may need to load the CPU for Coretemp to read its VID properly.
 
Well right off the bat... lower the PLL voltage from 1.85V to the default of 1.5V (overvolts to 1.54V at idle, and 1.57V under load), and no more than 1.57V. Lower CPU Termination voltage (VTT) from 1.60V to 1.36V - 1.4V. Try a Vcore of 1.35V and up the vNB / vMCH from 1.40V to 1.45V.

Since you're increasing vNB and increasing VTT, you'll need to adjust both CPU Reference and MCH Reference voltages. For a 45nm quad, CPU Reference should be 0.6333% of whatever value you set CPU Termination to, and MCH Reference should be 0.691% of MCH Core.

Tweaked those, found a weak link. Appears to be with the NB. CPU reference and termination take and stay.

3.50GHz is quite a bit for Q8200 for it to take and stable. My magic number is 3.40Ghz.
I have Q8200 clocked at 3.40GHz & Q6600 clocked at 3.52GHz, both of their Windows Index Experience are rated at the same score 7.4

BTW, sometimes you may need to load the CPU for Coretemp to read its VID properly.

I may back off. It ran extremely stable and ran Prime for almost a full day stable. Only reason it didn't make it was I turned it off.
 
Back