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

confused and frustrated

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
Oh yea it won't like three sticks.. 1Gb is ok but 2gb is better for gaming. v1.2 is BH-6 which I doubt will OC at all, doesn't respond to voltage either!! It's same like my CH-6 which is v2.2 or v3.2 I forget but .2 is xH-6 and .1 is xH-5. xH-5 does respond to voltage but xH-6 won't budge at all ..

To make sure to use dual channel it's either slots 1 and 2 or slots 3 and 4, seems 3 and 4 give more stable results. :)
 
Last edited:
well i can definately say ... it aint budging ... but to add insult to injury i cant even push my htt above 225 ... was getting better results off the mushkin (was only running 2 sticks for 1gb)

also fails prime95 at 1:1 (htt: 200x9, ddr200)

they do say that it should get 2.7+v so looks like im gonna have to break down and do the v-dimm mod just to get voltage to it as according to corsair the mb tops out at 2.65v ...

gggrrrrrrr .....
 
Last edited:
Stock for mine is 2.75v but it did prime 12 hours at DDR400 [200mhz] no problem without any mods. Guess if it's failing then it will only need a small boost, 2.78v would be perferct.
 
RAA said:
Stock for mine is 2.75v but it did prime 12 hours at DDR400 [200mhz] no problem without any mods. Guess if it's failing then it will only need a small boost, 2.78v would be perferct.

Guys,

It seems voltage varies slightly from MB to MB. I checked mine some time ago with OC's guidance and at idle it was 2.74v but under load it was ~ 2.66v. Probably close to yours. The drop in voltage under load can make a big difference.

These Patriots are rated from PC3200 to PC4200 depending on the voltage the MB puts out. Patriot states that with an small overclock of DDR433 they jump right away from their stock 2.6v to needing 2.8v and I don't have that much. It will fail Prime95 very quickly and usually fails and Superpi 32mb about halfway through.

A vdimm mod is the only way to go and doing the vcore at the same time seems to alleviate 95% of overclocking issues. Otherwise you're guessing to what the problem may be when in fact it is usually your vdimm and depending on your CPU, your vcore too.

Been down this road with 2 sets of Crucial Ballistix (2.8v). They would run great for a while then eventually failed. These Patriots run great and cool at stock and if I had the vdimm I would bet they would run just as good on any rated overclock.
 
RollingThunder said:
Guys,

It seems voltage varies slightly from MB to MB. I checked mine some time ago with OC's guidance and at idle it was 2.74v but under load it was ~ 2.66v. Probably close to yours. The drop in voltage under load can make a big difference.

These Patriots are rated from PC3200 to PC4200 depending on the voltage the MB puts out. Patriot states that with an small overclock of DDR433 they jump right away from their stock 2.6v to needing 2.8v and I don't have that much. It will fail Prime95 very quickly and usually fails and Superpi 32mb about halfway through.

A vdimm mod is the only way to go and doing the vcore at the same time seems to alleviate 95% of overclocking issues. Otherwise you're guessing to what the problem may be when in fact it is usually your vdimm and depending on your CPU, your vcore too.

Been down this road with 2 sets of Crucial Ballistix (2.8v). They would run great for a while then eventually failed. These Patriots run great and cool at stock and if I had the vdimm I would bet they would run just as good on any rated overclock.
RT, those PDP TCCD "815" love voltage unlike younger TCCD's (805,807).
I found 2.88 unstable over 225-230mhz. I needed not less than 2.95v to get them do DDR500 and now I am at 3.05v. One problem I came to find out is that Asrock board has weak vdimm (vdt) and PDP drew a lot of amp during heavy load (p95, spi). I soldered another capacitor to one of the memory cap below the vdimm slots, but still see a wide fluctuation. The voltage jumps from 3.05 to 2.88, 2.95, 2.90, 3.0, 2.85, etc.
So I think 2.70v (asrock's high setting) will probably drop the mem down to 2.5v during prime and spi. on those occasions, is when the tests fail.
I think any memory module is safe w/ this board's vdimm mod up to 3.05-3.07v and may need this much vdimm anyway, to coup w/ the the wide fluctuation.

On the Happier note, I think I have found an easier vdimm mod (soldering) and better location. I am trying to refine it and testing it. If all goes well, I will post it in voltmod thread.
I hope you find courage to do it since it is much easier, and your PDP is struggling hungry for voltage :).
 
Jame Gumb said:
OC how do you tell if they are the 815,805,807...???

Mine say
Code:
 Samsung     431
K4H560838F   - TCCD

H56  8  AQF159P2
look on the corner of pcb and you see a # like "B6U815"
BTW, here is screenshot of benchmark. The latest version of cpuz shows vcore incorrectly. It should be 1.61v instead of 1.30v.
 
Hmm didn't know it dropped so much. I tested this RAM on previous Dual Sata mobo and it passed 12+ hours +. Was 2.71 idle, didn't find out load but it passed. Maybe it didn't fluctuate so much, maybe my new one does.

I don't believe there's vdimm mod with VR, that would be great - are you making a VR mod OCH, looks like I might need to do it?
 
ochungry said:
look on the corner of pcb and you see a # like "B6U815"
BTW, here is screenshot of benchmark. The latest version of cpuz shows vcore incorrectly. It should be 1.61v instead of 1.30v.

OC,

That's an excellent Superpi 32mb time, especially for an AMD Athlon64 3000! I'm impressed. I only got down to 29m 30s with the Crucial Ballistix on this 3500+ stock before the board failed them.........
 
RollingThunder said:
OC,

That's an excellent Superpi 32mb time, especially for an AMD Athlon64 3000! I'm impressed. I only got down to 29m 30s with the Crucial Ballistix on this 3500+ stock before the board failed them.........
Consider me part lucky and part tweaking this board to death.
Goes to show you that you dont need super expensive motherboard to get best of your system. I honestly believe your PDP and hungry for voltage and cpu needs just a tad more. one of these days you have to break the ice and get down to it.

@RRA,
Yes I am using VR.
It's the same mod as i suggested to rebelranger.
All you have to do is solder a wire to bottom end of resistor #2 (the pic by TechpowerUp) and attach the wire to outer leg of a 2K(or 10K will work too) and attach ground wire to middle leg of VR.
This is the best vdimm mod IMO, because I can adjust vdimm to any value .01 increments up to 3.05-3.07v. it maxes out at 3.07v and no danger of overvolting. set bios to high and play from 2.75v to 3.05v.
depending on what speed the ram is, I adjust the vdimm.
 
well off to find a cheap soldering iron ... mine is burried somewhere and i am unable to find it ... will be doing the mosfet pad mod to see if that helps any ... i dont think i need to do the v-core as i was getting plenty of juice outta my cpu before (304x9 htt) its always been my ram that has given me the grief
 
okies ... did the the v-dimm mod ... that allows me to run memtest86 w/o error and prime95 at 1:1, 2-3-3-6 1t (200x9, 200/400) ... but i cant budge this thing for ANY kind of oc

gonna try setting the spd settings and see how she blows :(
 
nope ... nada ... best i can get is 225x9/133fsb ... effectively underclocking the ram ... my el cheapo mushkin performed better than the corsair does

extremely disappointed with the corsair at this point ... i thought this stuff was supposed to be good clocking ram ... instead i find myself in worse shape

any other suggestioins? besides, of course, going back to the mushkin
 
okies so update ... after i dropped my fsb divider to 100/200 i can now drive my cpu to 270x9 ... but thats just plain nuts!! ...

why i have to underclock my ram by sooo much to oc my cpu completely eludes me ...

and then to have to do a v-mod just to get sufficient v-dimm to even get the memory to respond kinda sorta correctly is unfathomable

thank god i'm not working on joe sixpack's computer ....
 
Last edited:
okies thread revisited ... running corsair twinx1024-3200c2 (matched pair) now

rma'd the previous pair ... now getting 300x9htt 1.45v, 100/200fsb @ 150mhz 2-2-2-5 1t, couldnt break 225x9htt w/ previous corsair

just cant get this corsair to run a faster divider on this board even w/ the rhcf v-dimm mod ... also wont boot at 200/400, defaults to 166/333

and cant get cpu stable above 300x9htt

but it is passing prime95 ... any suggestions?
 
Last edited:
If HTT can do 329 could try 329 x8.5 120/200 makes DDR400 2.79ghz. Also if temps are ok i'd run 1.55vcore.
 
cant do 1.55v but ill see if i can pull w/o

hhhmmm cant do 8.5 but ... got 340x8 (2.72ghz), 2-2-2-5 1t 170mhz (100/200fsb) which if im figuring all this right is ddr340 ... running prime again to see if it will hold it
 
Last edited:
Download CrystalCPUID, it can set 0.5 multipliers. A64Info can set 120/200 divider. I would think that a little more vcore might help but then maybe not, something to consider as a last resort.
 
get a not a valid win32 application when i try to run crystalcpuid ... looks like i have a rootkit on my system ... not a happy camper right now

definately a compromised registry ... dont have a back-up ... trying to see what i can do w/o wiping the system :(
 
Last edited:
Back