Please post here if your motheboard died with a Phenom II X6

I think I have a great way. Drill some holes next to the 80mm fan mount to accept a dual 80mm radiator. I just have to figure out a way to bleed the system so I don't have to use a reservoir.
 
MaximumOverkill,So I looked at my board and its a Revision 1.0 board. What revision did you get with your RMA board?

I just bought a Corsair H50 so that will be my next adventure, trying to retrofit it to the Antec fusion.
 
Well got my RMA replacement back. Still revision 1.0, but the serial number on my old mobo was 1003 series. This is 1005.

I'm going to let her break in for a day or two and then get to flogging. I have a high-rpm 90mm fan blowing at the MOSFET heat sink.

I redid my Corsair H50 so that I've got:

38mm-->Radiator-->25mm instead of:
38mm-->25mmSpacer-->Radiator

Idle temps are 20C and Load were 41C at stock voltage and clocks with case cover off.

APC reported 172 watts at idle and 297 watts at load (LinX), which interestingly enough, is 125 watts difference.

We'll see how much more wattage she pulls when I crank her up to 3.8-4.0 Ghz.
 
I am still trying to figure out why my temps are so high. I still have 40C at idle. I cleaned and re-did AS5 and my room temp is much cooler now.

Even if I turn down the fan speed to 50%, it goes from 6k rpm to about 3500-3800rpm the temps still stay the same. I wonder if I have a bad sensor of if it sill my BIOS settings. I just have the bios set to defaults, that came with the board. Is there something I am missing? There are so many BIOS options I get lost in there.

@MaximumOverkill I hope your new board bodes well.
 
What are you using to monitor temps? Try core temp and/or speedfan or see what the Bios reads under H/W monitor.

I'm pretty sure that with the hex cores you still only need to apply a rice grain amount of TIM. AS5, while not "the" best, is still a great top tier performer. It takes several days of normal computing for it to cure, but I think you can speed up that process if you burn in the cpu with Prime95 or Intel Burn Test although not sure if either supports six cores.

What are expected temps for stock voltages/speeds/cooler for the 1055t?
 
I think mine was In the 30s at idle and mid 40s under stress
Posted via [H] Mobile Device
 
Thanks Hashih, I'll get to OCing this weekend. The thing about AS5 (among others) is that you really need to have perfect mating surfaces for it to work as intended. Shin Etsu and other grey compunds don't need the mirror finish that AS needs. Sure you'll get cooler temps with each TIM on flat lapped surfaces, but AS is more dependant upon it and suffers worse from imperfect surfaces.
 
MassiveOverkill said:
Idle temps are 20C and Load were 41C at stock voltage and clocks with case cover off.

APC reported 172 watts at idle and 297 watts at load (LinX), which interestingly enough, is 125 watts difference.

We'll see how much more wattage she pulls when I crank her up to 3.8-4.0 Ghz.
Click to expand...

Idle temps are 22C and Load were 54C at 1.4V and 3.8 Ghz with case cover off.

5 Runs of LinX at 1.4V. PWM was slightly warm with the 90mm blowing on it. It's amazing that my system pulls 50 more watts with just IE, Chrome and Windows Mail open, where it shows 186 watts via APC .

a1d0c820.jpg


So for this run I was pulling 135 watts idle and 361 watts load for roughly 226 watts difference.

Forgot to add in the first post CnQ and all power saving features were off.

It seems my results are in line with below in that my OC'd PhenomII X6 is pulling nearly twice as much power at load.

http://www.xbitlabs.com/articles/cpu/display/amd-phenom-ii-x6-1055t-overclocking_9.html#sect0

My take: Current AMD boards that claim to support OCing, were designed with Ph II X4's in mind and thus power circuitry implemented accordingly to only handle a maxxed out 4 core (I'm talking non $200 boards meant for extreme OC).

I'm not going to push her past 3.8 Ghz as going 4.0 Ghz and above requires alot more voltage and as it stands now I have to actively cool the PWM area for my current OC.

I'll try and do some longer stability test with the case cover on tommorrow.
 
Last edited:
what BIOS are you running in the board..?

I still have yet to flash mine and get it running...

I skipped from page 1 to the last...what PSU are you running? as I have teh same HX520 you had I wonder if its enough or yours was just bad?
 
Running 1.6, which is the latest and what it shipped with. PSU is in my sig. Just got finished running 10 runs of LinX with the case all buttoned up. I tried various fan configs pointing at the MOSFETs but nothing got it stable until I stole the Antec Spot Cool out of my GF's computer. Never again will I build another rig without installing a Spot Cool.

8f7fedc6.jpg
 
kuripot said:
I'm still amazed by the idle temps of the X6's!
Click to expand...

the temp sensors are not calibrated correctly....they typically read around 7 C lower than true temp(at idle core temp is reporting 26 C in a room that is 86 F, which is 30 C....my cooling is good, not that good...i figure its idling at ~34, which is slightly higher than where my PII X2 idled)
 
MassiveOverkill said:
Thanks Hashih, I'll get to OCing this weekend. The thing about AS5 (among others) is that you really need to have perfect mating surfaces for it to work as intended. Shin Etsu and other grey compunds don't need the mirror finish that AS needs. Sure you'll get cooler temps with each TIM on flat lapped surfaces, but AS is more dependant upon it and suffers worse from imperfect surfaces.
Click to expand...

You must be right because I just installed a Corsair H50 in my system and my idle temp is still 39C! So I should buy Shin Etsu instead? Do I apply to the CPU or the cooler?

According to HW monitor my processor is running at 141.60W, that doesn't seem right at all.
 
Last edited:
Shin Etsu x23 is the pre-applied TIM that came on the H50 pump! Since it sounds like you already scraped it off, you'll want to apply the TIM to the cpu IHS
 
@ W.Feather: in everest I would always see 2 temps reported for the cpu, a core temp and a cpu temp. I had read that the core temp is usually more accurate since its a sensor built into the core vs the cpu temp which is a sensor reported by the motherboard. The core temp is usually higher and I'm guessing what we should be basing our temps on especially in relation to the max 65c temp.

I had an H50 initially installed, but didn't like the buzzing noise of the pump. So I got rid of it and put a Sunbeam CCt back in. Overall temps at idle and load are lower since I'm no longer intaking hot air into the system.
 
Last edited:
if you need some cheaper, good quality thermal compound, OCZ Freeze is very good, no cure time, works on imperfect surfaces just fine(it doesnt mind my HDT cooler), and keeps my temps nice and low
 
I'm using OCZ Freeze for the first time on my 5770. Ilike the fact that you don't need to wait for it to cure and its non conductive, unlike AS5. Have yet to try on a cpu.
 
i read through this thread and it seems the boards that are having problems have the 890 chipset. has anyone reported this same problem with the 880gma boards?
 
homerandabe69 said:
i read through this thread and it seems the boards that are having problems have the 890 chipset. has anyone reported this same problem with the 880gma boards?
Click to expand...

specifically those with the 890GX chipset, the 890FX has not been hit by this mainly because they have the needed power phases, plus the cooling for them.

the GX boards tend to have a lower power phase(4+1) and no heat sync on the mosfet's


880's are not seeing the problem really(as far as I know)
 
W.Feather said:
specifically those with the 890GX chipset, the 890FX has not been hit by this mainly because they have the needed power phases, plus the cooling for them.

the GX boards tend to have a lower power phase(4+1) and no heat sync on the mosfet's


880's are not seeing the problem really(as far as I know)
Click to expand...

well i will be getting my new noctua hs in so i will let you guys know if my board gets fried when i oc it.
 
well i tried to do a mild oc to 3.5 today using the stock fan. system posted fine and worked fine for about an hour. i was going to run linx on it while i was out of the house but 5 minutes after i started it up the system cut out. now i get absolutely nothing out of the system. no post or even indication that it has power when i press the button. just seems like the connection to the power is gone. looks like i have to rma now.
 
just to add to the information available.

my Gigabyte GA-890FXA-UD5 board has been running a 1090T at 3.6GHz since i built it a month or so ago.

it happily runs Heaven/Prime95/Steam-download when overclocked for over an hour as long as i bump the memory voltage up to 1.67v.
 
the 890 boards were designed with the x6 CPUs in mind, the older boards were not. I'd be surprised if that gigabyte board couldn't handle it.
 
MassiveOverkill said:
What board do you have?
Click to expand...

880gma-e45 with 1055t. ran it up to 3.5 on stock voltage with nb and ht at 2000 and memory on loose timings and clock. ran intel burn test for 5 minutes and the board went out. rmaing the board now and i just bought an open box crosshair IV for 200 shipped on newegg. ill try to sell the board i get back from rma.
 
OmegaAvenger said:
the 890 boards were designed with the x6 CPUs in mind, the older boards were not. I'd be surprised if that gigabyte board couldn't handle it.
Click to expand...

only the 890fx and some of the 7xx series have the 8pin power i think. the 4pin power connector is what is causing the boards to burn up
 
the connector isnt, the fact that it is a four phase power setup is what is killing the boards(in some cases)
 
The majority of boards that are dying are MSI boards, at least based upon this thread's reportings.
3.5 Ghz on stock voltage should not kill a board after 5 minutes in IMHO, but then again that board has no MOSFET cooling. Did you have a fan blowing at the MOSFET area?
 
so, 8pin 12v ATX or 4+1 voltage phases, which is it to be?

curious as to whether any reports on this board have arrived:

http://www.scan.co.uk/Products/Giga...(x16)-DDR3-1866(OC)-SATA-6Gb-s-SATA-RAID-mATX

only has 4+1, but has 8 pin ATX and is part of Gigabytes UltraDurable3 brand with heatsinks on the mofsets.

if this board falls over with X6's then it must be the 4+1 voltage phases that are insufficient............?
 
Last edited:
its not the actual plug that makes a difference.....its the actual power phase.....only boards that ive read/seen with the issue of dieing when overclocking a x6 are those with 4+1 power phases with NO cooling for the mosfets.....and even then its hit or miss, if you read this thread in its entirety you will see that there are people here that have not had an issue overclocking on some boards
 
im rmaing my msi 890gma-e45 soon and i just got a crosshair IV in which i have installed now. it looks like this one doesnt have the nb problem so i think im in the clear now.
 
I really don't think number of power phases have anything to do with the issues in this thread - 790/890FX boards all have 4+1 phases just like 890GX, 880, etc.

It's more a matter of how hot/stressed do the power regulation components get..

I just did an X6 1055T build for a friend on an 870 board which has no PWM cooling whatsoever and it's running beautifully, but this is a completely stock system which won't be overclocked.
 
The Pharaoh said:
I really don't think number of power phases have anything to do with the issues in this thread - 790/890FX boards all have 4+1 phases just like 890GX, 880, etc.

It's more a matter of how hot/stressed do the power regulation components get..

I just did an X6 1055T build for a friend on an 870 board which has no PWM cooling whatsoever and it's running beautifully, but this is a completely stock system which won't be overclocked.
Click to expand...

the issue occurs when it is overclocked. Something with the powerphases on 890GX boards is causing the issue, as far as I know 880 boards are not affected....just 890GX boards when overclocking
 
You must log in or register to reply here.
Back
Top