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

ncch-dl - what mods used to hit 200 mhz fsb

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

mbentley

Gloriously Lead, Overclockix Chief Architect
Joined
Sep 26, 2002
Location
Indianapolis, IN
okay, so i am curious what all modifications you guys did on your ncch-dl's. right now i have swapped out the northbridge heatsink with an active hsf cooling solution. i was curious how many of you modded anything involving the vrm's. also, what else did you guys do? right now i am having issues getting my 1.6's to be 100% prime stable @ 200 mhz fsb. i had some of the same issues with my 2.4's so i am thinking that i could be a few issues with the chips and the board... any ideas please?

*edit: right now i am trying to lower the multiplier and continue running 200 mhz fsb to see if the clock speed is just too much. so far i have run prime stable for a good bit at 12*200 = 2400. haven't tried anything else yet*
 
Last edited:
this board shouldn't need any mods to run 200mhz, it was designed to run at 200mhz (noconas)...

i would do what you are doing (12x200)---with 4x prime95 overnight...then keep increasing the multiplier...

then go back to the highest stable multiplier and start increasing the fsb...and repeating the prime95 testing...

also, clockgen and tweak 865 are great programs for adjusting fsb, ram timings in windows for testing
 
could the fact that i have 2 gigs of memory (4 x 512) have any effect possibly? i don't see how as i run memtest 86+ fine without errors for hours... i just want to eliminate every possibility...

this is the ram i use with all 4 sticks:

http://www.newegg.com/app/ViewProductDesc.asp?description=20-145-520&depa=1

it says 2.75v and i have it cranked up to 2.8v in the bios...
 
mbentley said:
could the fact that i have 2 gigs of memory (4 x 512) have any effect possibly? i don't see how as i run memtest 86+ fine without errors for hours... i just want to eliminate every possibility...

this is the ram i use with all 4 sticks:

http://www.newegg.com/app/ViewProductDesc.asp?description=20-145-520&depa=1

it says 2.75v and i have it cranked up to 2.8v in the bios...


maybe that ram needs more voltage...you have the bios vdimm set to 2.8v, but what voltage is actually being given to the ram as read in the bios?

my bios lists only 2.73v when i have it set to 2.8v...if yours is also 2.73v, maybe that's not enough for that ram..
 
should i try to run the ram at ddr333 speeds to see if the instability goes away (should that fix it?)? might i need to loosen the timings as well? i'd just like to find out what the weak link is in my system...

*edit: and yeah, i am getting 2.72/2.73 for my ram voltage.*

any clue what line on the psu that the ncch-dl gets the ram voltage?
 
Last edited:
I don't know how much the nb voltage affects stability, but any active cooling is probably better than stock.

Definately try different RAM timings and maybe even lower speeds like you were suggesting and report back. I've also had problems when using clockgen to change fsb so if possible try to do it form BIOS alone.
 
Well, coming from the 875p 1cpu, I know it does work quite well for higher fsb areas, but increasing sticks I'm not sure. Would bet it would, as 4 sticks does put quite a stress on the memory controller. However, I doubt it's enough to not allow 200fsb...
 
i have tried everything and i can't get this thing prime stable at 200 mhz fsb at any multi... :bang head

i tried lowing mem timings and speed, no change. i swapped it out with 2 sticks of 256 ecc, no good... i am thinking i just got some crappy chips or something...
 
mbentley said:
i have tried everything and i can't get this thing prime stable at 200 mhz fsb at any multi... :bang head


Yeah I'm in the same boat.

Then you see others report "Just pop in the cpus, set it to 200x16 and go!" and it just makes you wonder what you're doing wrong.

Maybe these boards just aren't as oc'able as the PC-DL's are??
 
Did you try the active NB cooling?

The mem controller is in the NB, it needs cooling with that much RAM.

-Sam
 
Well, I hate to ask this, but have you tried one cpu at a time, to get a feel for how much the nb is being stressed? FSB is handled by the memory controller, or in this case the NB. Using one cpu generally will burden the NB less, and might give you a hint if the board just plain old wont hit 200fsb.

Otherwise, I am running thin on ideas, but I'm sure there is a solution yet to be thought of.
 
no i haven't tried running a single cpu... i will try that when i am cleaning off my thermal paste and reapplying... i am getting pretty desperate...
 
quick question which xeons are you using the 1.6's or 2.4's and how did you set up the wire mod???
 
Rpkole said:
quick question which xeons are you using the 1.6's or 2.4's and how did you set up the wire mod???


i am using my 1.6 d1's wire modded by u wire with B2-B3, C2-C3, and D2-D3. voltage shows up around 1.56 in bios
 
after playing with the settings for a good couple of days, i think i finally have a stable 200 mhz fsb using a multi of 14. 2.8 isn't bad for a 1.6. i won't complain about 75% overclock anyday...
 
:clap: 2.8 is still a very good OC plus your FSB is OC 100% so look at it that way :clap:
 
mbentley said:
should i try to run the ram at ddr333 speeds to see if the instability goes away (should that fix it?)? might i need to loosen the timings as well? i'd just like to find out what the weak link is in my system...

*edit: and yeah, i am getting 2.72/2.73 for my ram voltage.*

any clue what line on the psu that the ncch-dl gets the ram voltage?


i believe it's the 3.3v line on the psu supplies the VDIMM...

but increasings that line past 3.3v will not increase the VDIMM...

with the patriot xbl tccd ram, i can't use loose timings at all...the machine won't boot with looser than 2,2,2,6 timings, and also won't boot at 2,2,2,5 timings...
 
Back