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

M

MassiveOverkill

[H]ard|Gawd
Joined
Jul 8, 2005
Messages
1,748
Please post your CPU, OC, if turbo was on, voltage, cooling, mobo (w/ revision), and symptoms:

I want to see if this is a Phenom II X6, MSI or some other trend going on ( maybe Turbo Core ramping voltage up too high).

I'll start:

1055T at 3.8 Ghz w/ turbo set to 15.5 with 1.4 V (turbo ramps up to 1.45V)
Corsair H50, maximum load 52C
MSI 890GXM-G65 rev 1.0
Was priming LinX and after 2 runs it powered down like it was in sleep mode. After normal troubleshooting it would power on for a fraction of a second and then die. If you removed the 4 pin power connector to the motherboard it would power up but not POST. Same behavior with everything removed including CPU and RAM. Blown PWM/MOSFETS seems to be the autopsy results. I've been overclocking for over a decade and this is the first motherboard in all that time that has died from overclocking. I don't think 1.4V on water is extreme.
 
Last edited:
i do know a guy on the folding team has killed 3 MSI 785G motherboards..
 
I'm starting to think the power requirements of the AMD 6 core is overloading the 4 phase power circuitry on these mATX boards. I'm seeing alot more reports of MSI deaths with a couple of Gigabyte thrown in. I have yet to see an Asus death.
 
Last edited:
MassiveOverkill said:
I'm starting to think the power requirements of the AMD 6 core is overloading the 4 phase power circuitry on these mATX boards. I'm seeing alot more reports of MSI deaths with a couple of Gigbyte thrown in. I have yet to see an Asus death.
Click to expand...

+1

You said it :p
 
I had an MSI NF980-G65 motherboard died on me during an overclocking attempt. And I only maxed out a 1.38v in the bios. And this is not an mATX motherboard, it's the best (and only) AM3 SLI motherboard.

Phenom II 1090T
Turbo-core disabled
4.0ghz by changing multiplier only
1.38 v

It just went pop, and then there was a smell of burning electronics. Tracing the smell, I got to the MOSFET heatsink on the motherboard.
 
My (signature) 790GX board with 4+1 phases was nicer to me and simply shuts down when the load gets too great to handle with my 1090T. With Turbo & Cool N' Quiet disabled the delta from idle to full load could be well over 200W. The largest delta I saw with my Kill-A-Watt was an idle around 180W and then a run of Everest -> FPU Julia with produced numbers of 414-415W seconds before shut down. Upgrading the power supply from 550W to 750W had no effect. Upgrading to a 890FX board did. I caused this shutdown easily 20+ times without any apparent damage to the board.
 
I have been waiting to go pII x6 just to make sure bios and boards are up to snuff. Are Asus the only boards that seem to not be effected?
 
If you're a member of overclockers forum, I made a post over there as well...

I'll add, if you remove both CPU and RAM, I get the same behavior. With the +12V auxilliary power connected the power supply or board acts like there's a short and powers up, and then immediately powers down for a fraction of a second. If you remove the +12V auxilliary power, the board powers up and stays powered up, but does not POST.

I'm posting from another Fry's combo I got with the G6950/MSI H55M motherboard. This motherboard has the same Active Phase Switching technology that my AMD board had. I have it disabled on both boards so the PWM doesn't deactivate on low loads, but I'm wondering if MSI's implementation of this feature is somehow causing the problem/flawed.

I have my power saving feature set in OS (Sleep after 1 hour), and woke up at 4:30 to this temp. setup just cycling power to the board, like it's trying to go to sleep, but can't.

I took a picture of the PWM/MOSFETS of my dead board:

662bebfa.jpg


Compare that to say this:

http://www.newegg.com/Product/Image... 6Gb/s USB 3.0 HDMI Micro ATX AMD Motherboard

and you can see there are less, but what seem like beefier components.
 
MassiveOverkill said:
I'm starting to think the power requirements of the AMD 6 core is overloading the 4 phase power circuitry on these mATX boards. I'm seeing alot more reports of MSI deaths with a couple of Gigabyte thrown in. I have yet to see an Asus death.
Click to expand...

This could be true with overclocking. Remember a typcial 4 phase is not really an overclocking board. I personally only use 8 Phase when I want to overclock.

I know my Gigabyte 770 and 870, both 8+2 phase have had zero issues with overclocking of a 1055T to 1090T.
 
i had a msi 750a sli borad go on fire!!! the thermal pads on the mosfet got on fired and burnt some parts of the board. this was when i ovclocked the 1090t.
 
flexcore said:
I have been waiting to go pII x6 just to make sure bios and boards are up to snuff. Are Asus the only boards that seem to not be effected?
Click to expand...

it seems at the 890GX chipset, only Asus has not been affected(as of now), however all other 8** series chipsets have not had the issue(so far as ive seen), though if your on lower than an 890*X your probably not overclocking, and the 890FX boards all have adequate power phases
 
Interesting - when I picked up my 1055t, I was going to throw it in a 785g mATX board that I had, but was worried about the power requirements. When I picked an Asus 890gx board, main reason I got the one I did over a similar Gigabyte version was because the Asus had 8+2 power phases, Gigabyte had the 4+1. Thought I was just being a baby worrying about it, because if the board supports 140w it shouldn't matter whether its 4 or 8 phase power. Looks like I made a good choice. :D
 
motherboard in sig. is a 4+1 power phase and normal voltage is 1.475.I have run it at 4.1ghz without problems.I also use the finger touch method to check componets;they barely get warm.At this point processor speeds and voltages are turned to 3.6ghz ans 1.375v. this fine for me and still leaves headroom.

sorry to hear there are problems.
 
I have same motherboard as the OP.

I tried 1.56V yesterday and it was working great, but very hot during stress testing,
 
Asus M3N78-VM here and at stock volts doing 3360 + turbo is on. Still going strong.
 
I killed a Gigabyte 4+1
On a stress test w/o OC
Board was hot hot especially the Fets

Compared to a ASUS 8+2 which was a replacement
Stress test w/o OC board was warm at best not even close to hot
 
adobian said:
I have same motherboard as the OP.

I tried 1.56V yesterday and it was working great, but very hot during stress testing,
Click to expand...

How high was your OC and what board revision do you have? What were you using to stress with?

Omegas said:
Asus M3N78-VM here and at stock volts doing 3360 + turbo is on. Still going strong.
Click to expand...

I doubt you'll have any problems. I don't think that's high-enough of an OC to overload the MOSFETS/PWM

iPearl said:
I killed a Gigabyte 4+1
On a stress test w/o OC
Board was hot hot especially the Fets

Compared to a ASUS 8+2 which was a replacement
Stress test w/o OC board was warm at best not even close to hot
Click to expand...

Was that Gigabyte a ATX or mATX board?

I don't think I mentioned this but the board ran fine for about a week just in the open as I was waiting for my case to come in. I had a 120mm fan blowing down over the MOSFET/PWM. Inside the case there was still some air getting to that area, but obviously not as good as in the open air.
 
Last edited:
Mine tested at 1.56V V core but only at 3.8ghz. It test prime95 for a while but failed after about 30 minutes. Since then I ran 1.41V at 3.5ghz, but I think I am going to return this MSI 890gxm-g65 .

A friend of mine had the same board, he toasted it yesterday while downclock from 3.99ghz to 3.83ghz. When booted up, the blue LED on the MB flashed three times as though it was warning of overheat. Then it shut down. Booting back up, the fans ran and the LED was lit but nothing else. He said he smelled something burned around the CPU area but he doesn't think it's the chip. He'll get another board in a few days to test the chip.

He also mentioned that he used the Cooler Master Hyper 212 Plus cooler which has a metal backplate but a very thin and soft non-conducting layer against the motherboard and when he screwed on too tight, the sharp pins underneath the MB might have pierced through to touch the metal plate and shorted out something around the CPU.
 
Last edited:
I want to add that on the day of this MB failure, he added the EVGA 8800GTS card to the system and played some games for a about 5 minutes. then stress tested Prime95, failed the test,rebooted to downclock in bios, then rebooted and this happened.

For me, I might go back to Intel and build an i7 system.
 
Last edited:
griff30 said:
glad I didnt buy a 6 core just yet.
Click to expand...

you realize this only occurs on 890GX boards(as far as we have seen), and mainly mATX i believe(too tired or somethin to remember it all :p ), but anyhow nothing is wrong with these chips(atleast mine), but rather select boards
 
Yikes! I was looking into buying a 1055. Maybe this is a good excuse to get that Asus Crossfire IV I really want but was having trouble justifying the $220 price tag. :D
 
Aw crap, now I'm all paranoid my new Gigabyte 890GPA is going to fry. There is no indication that this will happen (voltage is spot on with no ripple according to OCCT), but I'm scared nonetheless. And to the guy mentioning going Intel to solve this issue: They have the same issues. There was quite a hubbub especially over lga 1156 mobos going up in smoke a while back.
 
I am not running one, nor have I killed a board, but I recall seeing something similar to this in a Toms Hardware 1156 motherboard roundup where their attempts at pushing OC voltage over a certain barrier caused "cheaper" boards to blow their MOSFETs in a short period of time. I think MSI had fixed the problem on a later revision or BIOS update by limiting voltages. I think the only boards not to have this problem were Asus and Gigabyte ones.

Probably a related issue.
 
Nobody should be surprised about this happening to motherboards without heatsinks on the mosfets. If you don't have a good heatsink or a strong fan blowing directly on your mosfets you're just asking for trouble no matter how many phases you have. Strip the heatsink off those 8+2 motherboards and leave it to the mercy of passive cooling and I wouldn't be surprised if a Thuban could kill them as well. We saw the same thing with the 9850BE/9950BE. It correlates that a manufacturer would be more likely to add in the cost of a heatsink if they are building a more complex power layout.

http://www.irf.com/technical-info/whitepaper/TP-071123.pdf
Page 30 for a Barcelona server 5+1 phase layout under artificial load ~180W total (144W vCore section).:eek: Being server parts they survived but I would expect your standard desktop mosfets to die a quick and horrible death at those temperatures.
 
I had a MSI 785G-E51 that would just shut down as soon as I ran a stress test using all 4 cores on my P II 555. I RMA'ed the board and they sent me back the newer E65 board with the heatpipe cooler on the NB and mosfets. I haven't had time to test it yet, but it looks like it should help a lot- pretty big cooler on a mATX board!
 
The thing is that this is happening to decent 890GX boards WITH heatpipe solutions on the mosfets.
 
I blew up a MSI 785G-E45 with an X6. I think it would have lasted longer if I wouldn't have overvolted to about 1.4v and overclocked to 3.8ghz

It only lasted a few hours of stress testing at those settings
 
MSI Customer support is finally responding to claims on Newegg. We'll see what happens. Email address they gave is bouncing back. I already shipped my mobo off under RMA anyways.
 
Last edited:
MassiveOverkill,
Just out of curiosity, did you find the serial number on the motherboard? Or is it just on the carton box?
 
Is this happening to boards that are NOT overclocked as well? I have a 1090T I have no intention of overclocking...
 
Robstar said:
Is this happening to boards that are NOT overclocked as well? I have a 1090T I have no intention of overclocking...
Click to expand...

no, you should be fine....whats happening is that when you start overclocking more power is being pushed through the 4+1 phase, causing them to fry(key word is should)
 
W.Feather said:
no, you should be fine....whats happening is that when you start overclocking more power is being pushed through the 4+1 phase, causing them to fry(key word is should)
Click to expand...

Thanks! This is what had me worried too.
I wasn't going to OC just wanted a rock solid 6 core motherboard with 2 PCIe slots for crossfire or maybe SLI cant decide which graphics cards to go with. Oh and USB3 and lots of SATA 3...
 
You must log in or register to reply here.
Back
Top