Help. The problem with the Crosshair V Formula

[Semer]

Hello. Help to deal with the problem.
There are: Motherboard Crosshair V Formula, CPU Phenom II x6 1055t 125W, Scythe Mugen2 rev. B, 4Gb SiliconPower 1333MGz CL9, Palit GTS250 E-green 512Mb, PSU Sirtec High Power HP-550-G14S.
I can not achieve stability when overclocking CPU, computer constantly reboots.
CPU 3.7GGz (265x14) / 1.33v, CPU / NB 2650MGz/1.2v, HT 2120MGz, LLC CPU and CPU / NB in ​​all positions - restart under load after 15-20min. Temperature of the CPU and MB on board sensors, CPU 56 * C, MB 35 * C. Overheating of the processor is not as at a temperature of CPU 60-62 * resets the multiplier to x4, without rebooting.
Yesterday put the dispersal 3,5 GGz/1.29v, LLC Regular, 1.5 hours watching a movie the computer hangs, the CPU temperature and MB were 40 * and 34 * respectively.
Before this was a motherboard ASUS M4A87TD, acceleration above 3,7 Ggz rested in the temperature of CPU, but there was no reboots.
I think that these stability problems when overclocking associated with overheating power CPU, in particular those that are located on the reverse side of the board. They are heated to 65 *- 75 * (measured the Chinese with a multimeter). To what temperature they can bask?
If you are not hard to see screenshots of the BIOS settings, I can what is not properly configured.
Thanks in advance. Sorry for my bad English.

Screenshots

Extreme Tweaker tab:





Tab DIGI + VRM:


Tab Dram Timing Control



Tab GPU.DIMM Post:

on this tab, the computer always freezes after 10-15 seconds, only helps the Reset button

Memory Configuration tab:

ECC Mode set to Disabled

Monitoring tab:

[Raja@ASUS]

VCore at 1.33V for 3.7GHz may be too low. With medium LLC load voltage will be around 1.28V.


There is no "overheating problem" with the board. You can place a fan over the VRM heatsink if you wish. The FETs on the back of the board can operate at temps up to 150C. 4GHz is possible fully stable with Thuban.

Also, set DRAM Command Rate to 2T. Set Read to Read Timing to 5. And set CPU/NB to more than 1.20V for 2650MHz or use around 2000Mhz for now.

Also which BIOS version is this?


Use this version here:

http://www.asusrog.com/forums/showth...beta-BIOS-0045

-Raja

[Semer]

Raja, thanks for your answer.
Oddly, on a budget motherboard ASUS M4A87TD CPU 1.33v and CPU / NB 1.2v was sufficient for the stability of OSST. And memory run at 8-8-8-24 1T 1333Mz. Oh well.
Now is the bios 0506. 0045 version already downloaded, the evening is set.

Raja, for the CPU PWM Mode [T. Probe] external thermal sensors are needed?

[Raja@ASUS]

Raja, thanks for your answer.
Oddly, on a budget motherboard ASUS M4A87TD CPU 1.33v and CPU / NB 1.2v was sufficient for the stability of OSST. And memory run at 8-8-8-24 1T 1333Mz. Oh well.
Now is the bios 0506. 0045 version already downloaded, the evening is set.

Raja, for the CPU PWM Mode [T. Probe] external thermal sensors are needed?

That is the problem, you need to take into account the LLC and real voltage level of both boards - then compensate accordingly.


You do not need external sensors for T.Probe, this is done by the VRM.

-Raja

[Semer]

Ok. In the evening I will try. Thank you.

[Semer]

Hello.
Yesterday, according to your advice, set up in the BIOS so: CPU-2808MHz (272x14) / 1.382v, CPU/NB-2720MHz/1.3v, HT-2176MHz, Memory-1446MHz/1.5v, the remaining voltage left by default, only ordered manually . CPU LLC-High, CPU / NB LLC-High, the memory settings - all on auto. Half an hour tested Prime95, CPU-maximum temperature 59 *. The radiator on the back of the board was heated to about 70 *. During testing it was stable. Immediately after stopping the test - BSOD 0x0000001A (Memory management).
Flashed bios 0045. Today I will continue to test the new BIOS.

did not know that the board would be so hot.

[Raja@ASUS]

Place a fan over the VRM heatsink if you are worried about temps. The whole power plane will be cooled.

Make sure you have memory set up properly for your overclock. In BIOS 0045 use the DRAM Stabilize function. If DRAM is still unstable, reduce CPU/NB frequency to under 2500MHz to debug the problem.

-Raja

[Raja@ASUS]

If heat was such a big problem, this would not be possible:




also:








Using fan to cool VRM heatsink near CPU. No need to worry about FET temperatures on the rear of the board.

-Raja

[Semer]

Using fan to cool VRM heatsink near CPU. No need to worry about FET temperatures on the rear of the board.

-Raja

Well, your words are soothing. The fan on the heatsink near the CPU now put.
It's not as bad as it seemed before.
Thanks for your help.

[Semer]

Raja, tell me more, I read reviews about the board and saw that many people have the reverse side of the board was covered with a thin layer of sticky material. When I built a computer that also saw that the board is covered than the sticky. But not the whole board is covered, and in some places. On the part of the CPU was not, was a pure PCB.
What could it be?