That didn't take long...

P

Patriot

[H]ard|DCer of the Month - March 2011/June 2013/De
Joined
Dec 15, 2010
Messages
2,502
A picture cries a thousand words...
IMG_20130126_014218.jpg


Anyone know how to RMA SM boards?
 
Pretty straight forward. Get the model and serial number off the board. Usually printed on white labels. Email their tech support address stating you want a RMA for the board and list the model, series number, and what when wrong. Turn around from request to RMA number is usually a day.

Sorry to see that happen. What board was it and what happened? I am guessing a little too much magic go-go jiuce?
 
Firstly,
arguecat.png


Secondly, was the failure due to general operation of the equipment, or was it operated outside of normal parameters? :p

Oh, it should also be noted that this guy was running cine benches on this box....this is what happens when you cine bench instead of fold. :)
 
"Secondly, was the failure due to general operation of the equipment, or was it operated outside of normal parameters?"

I would say normal operation as that will eventually kill a board.

You Know Big Brother reads these boards.;)
 
Do you know what the surface temperatures on those components were? I wonder if high temperatures led to the failure or if there was some other kind of electrical failure caused by high voltage or current that caused it to burn like that.
 
Must have been the latter as the component released the smoke at the power-on...
 
Zink said:
Do you know what the surface temperatures on those components were? I wonder if high temperatures led to the failure or if there was some other kind of electrical failure caused by high voltage or current that caused it to burn like that.
Click to expand...

It was cooled adequately ...
chips were watercooled vrms had copper sinks and 4 fans on them...

This was as you can see on the back of the board... fan sitting right above it halfway over board to give airflow to the backside as well...


I was pushing it.
But others have had the same thing happen under much much less stress.
It lasted 9 wu's at ~1400w AC load... 3.48ghz normal folding operation.
Set a few WRs on it last weekend.

http://hwbot.org/benchmark/wprime_-_1024m/
http://hwbot.org/benchmark/cinebench_r11.5/ (I have a higher one in pocket for the first guy that passes me ;) )

It was acting wonky since I rebuilt it... before any tweaking... reboots at stock.
I only got it behaving last weekend and it was still finicky on windows and linux installs... it seemed off.
Then it was folding rock solid for 9 wu's then smoke.

I was running the AC at 68, and keeping the chips on load under 60...
Now I get to test all of the individual components... Joy.
 
Were you using OCNG bios? Doesn't that void warranties?

Or was that BIOS conjured up in a way that isn't easily detected? :p
 
If SM send you a new board then use standoffs on the case, so that air can get to and cool the rear of the board.
 
Spazturtle said:
If SM send you a new board then use standoffs on the case, so that air can get to and cool the rear of the board.
Click to expand...

... if you don't use standoffs you will short the board...

There was a fan blowing on the part that busted...
 
It must have had cereal for breakfast? Post Toasties, specifically?
 
jmilcher said:
OP- you were using standoffs correct? Doesnt look like it at all in that pic
Click to expand...

Just standard standoffs...else I couldnt use the expansion slot.

I might consider getting taller ones like they use for open builds...but that would mess up my spacing with the radiators...

I might however make a few more cutouts on the mobo tray and mount some 120s back there.
 
Yeah from the looks of it the board was too close to the case, now remember higher temps = higher resistance, so if some components were getting too hot it could cause others to get hotter and die.

Also remove the rubber grommets.
 
Last edited:
Patriot said:
Just standard standoffs...else I couldnt use the expansion slot.

I might consider getting taller ones like they use for open builds...but that would mess up my spacing with the radiators...

I might however make a few more cutouts on the mobo tray and mount some 120s back there.
Click to expand...

Standards are fine, no reason to raise the board abover the 1-2mm the standoffs do. Maybe introducing airflow back there could help though.
 
Spazturtle said:
Yeah from the looks of it the board was too close to the case, now remember higher temps = higher resistance, so if some components were getting too hot it could cause others to get hotter and die.
Click to expand...

facepalm.... keep in mind this rig is watercooled with vrm sinks and fans on those sinks
AC set to 65F ....

there was actually a fan that would have provided airflow to the place in question...

More so... while that socket would have had airflow to its driver circuits... the other sockets would not have.
Driver circuits really shouldn't get hot... I agree with Core32 and tears analysis... most likely a passive component died and that damaged the IC... pop...

http://hardforum.com/showthread.php?t=1619088&highlight=harbringer

upper right is the part of the board that had issues... notice the fan... at the time it was adjusted even closer and halway over the board....

Those copper sinks were barely warm to the touch thanks to all the air movement.
IMG_20130118_225951.jpg
 
Last edited:
Patriot, Patriot, Patriot...
 
Patriot.
I am sorry to see that there have been an accident with your SM mobo, but you are not alone you should know.

Something similar happened with my SM motherboard for a little over 2 months ago. Luckily it was only the plastic shield that burned, so the motherboard is still folding on further at 3.0GHz. The strange thing was that before this happened, I had folded flawlessly for some time at 3.5 and 3.6GHz, while the accident happened later on just 3.4GHz. http://hardforum.com/showpost.php?p=1039349302&postcount=162

I had not mounted heatsinks when the accident happened, but I have now. http://hardforum.com/showpost.php?p=1039367180&postcount=246
 
Well unfortunately in my case a resistor went out making a short... (I think)
And it will create a 6 inch long fireworks display if I give it a chance :)
 
You must log in or register to reply here.
Back
Top