flexy
07-27-2005, 05:53 PM
hi,
so i decided to post a problem description of what's going on here. I know that there are some VERY competent modders/hardware people around and maybe one or the other can hint me in the right direction.
Be advised that we can almot certainly exclude ALL software/bios settings etc. related solutions - believe me, i tried 'em ALL. I am on a dfi LP nf4-D. Please see my sig for details.
** March: I was with the hundreds of other people who got their dfi NF4 LP boards when they came out here in the us. (Or was it Feb ? )
I used to find my overclock at 289x9 with 2.5 4 3 7 timings on the 2 sticks OCZ TCCD (week 37) 512 MB sticks. Halfway stable. Stable means: OCCT/Prime95 etc. Bios 310p. VMem was at 2.9 and i was a happy camper.
** Months later:
I notice more and more instability at settings i used to run fine.
All of a sudden i got BSODs of all kinds and the ONLY way to get rid of stability and BSODs is to increase VDimm voltage.
Was running those TCCD at 3.1 289 2.5 4 3 7 to get em *somewhat* stable at 289 as i used to. Didnt last long. I think right now there is NO way to get the sticks stable at 289 *at all* !!
** switched to 10x260 2.5 3 3 7 3.0V. VERY NICE settings, almost as good as 289 4 3 7 and i can go at 3.0 without being too close to 3.3 rail. Run at 3.0 and dont see the big fluctuations in VDIMM as at 3.1. (AS said 3,1 is reallt darn close at 3,3 especially since i think my 3.3 rail is a BIT underpowered. All other rails are great or some minor percentages above specs, but 3.3 is always a bit low, eg. reads 3.26 in mbm while the other rails all read higher than spec. Anyway, i can live with 3.0 and 260x10
** few days ago:
INSTABILITY at 3.0V 260x10 2.5 3 3 6, BSODs, SH*T !!! Imagine those chips MIGHT run 300 or so 2.5 4 3 7 at 2.85, now they cant even do 260 at 3.0.
Need to inc VDIMM at 3.1 to get rid of BSODS and now i am halfway stable.
=====> following our known "nf4 vdimm mod/vtt" issue thread i was measuring my VTT too, also posted some postings in this thread.
I can see VTT dropping even on DMM, i can see my machine getting instable in OCCT on the DMM :) Eg..run....look at DMM....VTT is ok....VTT all of a sudden DROPS...and BINGO! Occt/Prime failed.
So....some bright people in the above thread already suggested that my increasing of my VDIMM needed for TCCD might be that this is a way to compensate for the VTT drops.
Anyway....i really need to know what component is LIKELY "defective", "weird", whatsoever.
Please note that i also STRONGLY fear that this will continue, eg. more and more instability over time....need more and more VDIMM (i CANT, 3.1 is the max i will give this TCCD and its already on the very edge !!) and i am just waiting til it is instable at 260 mhz.
-- motherboard ???
- voltage regulator ????
-- OCZ memory sticks ???
-- Memory controller on CPU ???
One person kind of brought me on to something....because i remembered i once did something stupid and pushed like 1.9 or so volts through my winchester. For maybe 5 seconds or so. (Wrong bios settings, you guessed)
I do not know however whether my probs started AFTER that or if i had the issues (constant degrading overclockability and instability) before already !
Maybe this powersurge "shot" the memory controller on the cpu ?
Note that i used to run this system for MONTHS at 9x289 2.5 4 3 7 vmem 2.9V ....and now its going downhill.
Next thing i MIGHT probably try a VDIMM mod since people say it might help for the DIMMs not needing that much voltage (as i read on bleedinendge, tony said that)...and i am also VERY interested whats going on with the NF4 VTT mod thing.
thanks !
so i decided to post a problem description of what's going on here. I know that there are some VERY competent modders/hardware people around and maybe one or the other can hint me in the right direction.
Be advised that we can almot certainly exclude ALL software/bios settings etc. related solutions - believe me, i tried 'em ALL. I am on a dfi LP nf4-D. Please see my sig for details.
** March: I was with the hundreds of other people who got their dfi NF4 LP boards when they came out here in the us. (Or was it Feb ? )
I used to find my overclock at 289x9 with 2.5 4 3 7 timings on the 2 sticks OCZ TCCD (week 37) 512 MB sticks. Halfway stable. Stable means: OCCT/Prime95 etc. Bios 310p. VMem was at 2.9 and i was a happy camper.
** Months later:
I notice more and more instability at settings i used to run fine.
All of a sudden i got BSODs of all kinds and the ONLY way to get rid of stability and BSODs is to increase VDimm voltage.
Was running those TCCD at 3.1 289 2.5 4 3 7 to get em *somewhat* stable at 289 as i used to. Didnt last long. I think right now there is NO way to get the sticks stable at 289 *at all* !!
** switched to 10x260 2.5 3 3 7 3.0V. VERY NICE settings, almost as good as 289 4 3 7 and i can go at 3.0 without being too close to 3.3 rail. Run at 3.0 and dont see the big fluctuations in VDIMM as at 3.1. (AS said 3,1 is reallt darn close at 3,3 especially since i think my 3.3 rail is a BIT underpowered. All other rails are great or some minor percentages above specs, but 3.3 is always a bit low, eg. reads 3.26 in mbm while the other rails all read higher than spec. Anyway, i can live with 3.0 and 260x10
** few days ago:
INSTABILITY at 3.0V 260x10 2.5 3 3 6, BSODs, SH*T !!! Imagine those chips MIGHT run 300 or so 2.5 4 3 7 at 2.85, now they cant even do 260 at 3.0.
Need to inc VDIMM at 3.1 to get rid of BSODS and now i am halfway stable.
=====> following our known "nf4 vdimm mod/vtt" issue thread i was measuring my VTT too, also posted some postings in this thread.
I can see VTT dropping even on DMM, i can see my machine getting instable in OCCT on the DMM :) Eg..run....look at DMM....VTT is ok....VTT all of a sudden DROPS...and BINGO! Occt/Prime failed.
So....some bright people in the above thread already suggested that my increasing of my VDIMM needed for TCCD might be that this is a way to compensate for the VTT drops.
Anyway....i really need to know what component is LIKELY "defective", "weird", whatsoever.
Please note that i also STRONGLY fear that this will continue, eg. more and more instability over time....need more and more VDIMM (i CANT, 3.1 is the max i will give this TCCD and its already on the very edge !!) and i am just waiting til it is instable at 260 mhz.
-- motherboard ???
- voltage regulator ????
-- OCZ memory sticks ???
-- Memory controller on CPU ???
One person kind of brought me on to something....because i remembered i once did something stupid and pushed like 1.9 or so volts through my winchester. For maybe 5 seconds or so. (Wrong bios settings, you guessed)
I do not know however whether my probs started AFTER that or if i had the issues (constant degrading overclockability and instability) before already !
Maybe this powersurge "shot" the memory controller on the cpu ?
Note that i used to run this system for MONTHS at 9x289 2.5 4 3 7 vmem 2.9V ....and now its going downhill.
Next thing i MIGHT probably try a VDIMM mod since people say it might help for the DIMMs not needing that much voltage (as i read on bleedinendge, tony said that)...and i am also VERY interested whats going on with the NF4 VTT mod thing.
thanks !