Problems with my dfi expert

b0x1d

New member
Hi guys, ive tried getting some sense on the dfi forums and my post dont seem to have generated much replies. Ive seen the effects of a cold bug before now but im unsure wether this cpu has it or not, here is the link to the thread i started http://www.dfi-street.com/forum/showthread.php?t=52649

If it is the cold bug does anyone have any suggestions as for what cpu to go for ? :/ price range (£100-£140 tops)

ps - im looking @ getting a single core, not dual.
 
I did know of the top of my head not long back... i know its a cabge im not sure of the week tho, when i take the prommy of next i'll check it and let you know. (hoping to get a venice 3800+ to test in here soon :D)
 
When you say htt you mean the htt multiplier yes? I have also tried lower drive strengths ;/ anything past 250fsb is iffy to use, i get random reboots when its not under load but yet i could play css for 3-4 hours then it will be fine?. Like i posted on the dfi forums this cpu used to run 300fsb x9 1:1 rock solid, on air. I cant even post 300mhz ratiod now(windows).

Ps nice one for the link mate, i found that a great help on my future buys also..
 
Your chiilly1 Extreme is too cold for this CPU - really you need a CPU with a higher or unlocked multiplier than what you have now to unleash the power.

I am selling an FX60 and an X2 4600 - the X2 does 3.2ghz under phase, the FX60 does 3.4ghz :) X2 CPU's are hard to find that do not have a cold bug, and this unit is faster than a x2 4800 ;) The FX60 will do 212x16 all day long and will push up to 215x16 foir benching (cold bugged but unlocked multiplier still allows you to rech your goal of max overclock).

HTT = FSB, LDT multiplier should be on x4 for 250mhz x3 for 300mhz

Drive strength should be 5&2 on the Expert, also bang CPU VID volts up to 1.525 with 116% special VID CTRL and LDT Voltage at 1.4v - your chipset volts should be 1.65v
 
b0x1d said:
When you say htt you mean the htt multiplier yes? I have also tried lower drive strengths ;/ anything past 250fsb is iffy to use, i get random reboots when its not under load but yet i could play css for 3-4 hours then it will be fine?. Like i posted on the dfi forums this cpu used to run 300fsb x9 1:1 rock solid, on air. I cant even post 300mhz ratiod now(windows).

Ps nice one for the link mate, i found that a great help on my future buys also..

When the CPU's loaded, the temps increase, probably to a point at which the CPU is a little happier.

Theres probably a few options, firstly as mav said get a new CPU. Secondly, again as mav said whack up the CPU volts. This will increase CPU temps and should allow you to clock a bit further. Thirdly replace your cooler. Fourthy, manually make "bad contact" between the CPU and the evap. Best bet for this is stick a 2p in between the heatspreader and the evap itself. Adding a little thermal goop if you want. This should increase temps a fair bit, and might well help reduce the bug.
 
Hi again guys, thanks for the advice and offers ;D Mav, how much you looking for the x2 4600? The advice above has been a great help, i upped the vcore yesterday morning to try and avoid the cold bugged point only to see that it will now do 273x11 13h prime stable ;o. Highland3r i'll try the 2p method also when i get time again to remount it. (im off to try the advice above ;))
 
Hi mate,

Glad I could help.

The pitfall of the 2p trick is the extra distance in creates between the mach head and the cpu mount - more seal string maybe required to prevent condensation. Also the uneven surface could effect the load temps quite a bit also - a lapped 2p would be better.

The x2 4600 is not cold bugged and will easily push up to 3.2ghz rock solid stable - these are not easy to come by (without cold bug and guaranteed overclocking under phase) and I am selling this for £375 delivered cost just under £400 new and is just over 7weeks old - if I can find it I'll supply receipt of copy invoice from online store.

PM if interested.

Cheers

Mav
 
Another option is a few layers of tin foil.... Grab a strip about the same width as the IHS and fold it over few times. Play with the number of layers to get the CPU running at a temp that reduces the bug enough to get a decent clock.

The coin "trick" will muck up the contact enough, load temps are affected more than idle, if you lap the coin to get a better contact temps wont drop as much.

It really depends on the extent to which your CPU's bugged, temps etc and the capacity of your unit.
 
Without any of the above method's ive got the cpu running sweet @ 3.1ghz :D (1.744vcore 282x11) once again thanks for the advice guys, fiddled with dram strength's and im using 2t so the mem controller must be a little more @ ease :D

ps - i really couldnt be fcuked to remount the prommy again lol

pps - wp on the 44k in 03 mav :)
 
Back
Top