P5WDH Phase change - dielectric grease a must [corrosion]

[eva2000]

After my 12 day run with E6700/Asus P5W DH on Kayl Black Frozen SS http://www.xtremesystems.org/forums/...d.php?t=115040 I decided to leave unit off overnight and next day unmount the evap and see how things are underneath. Motherboard was dry under the insulation, a bit damp like on top of the sticky back foam pictured below



Other than that everything looked fine. Removed insulation and plucked out the cpu and noticed these brown marks on back of cpu's gold contacts. Looks like a bit of corrosion, or something ?

Brown marks shown in below pictures mainly at

(click image for larger version)


top left side from
4th contact on top row
3rd to 5th contact on 2nd row from top
5th contact on 3rd row from top

bottom right side from
7th contact from right side on bottom row
7th and 8th contact from right side on 2nd row from bottom row
7th contact from right side on 3rd row from bottom row
7th contact from right side on 4th row from bottom row

The location of the brown marks on these contacts correspond nearly the same to where the below pics show that the motherboard LGA pins have from the naked eye misaligned or the pin arrays are slightly darker in those pics than the rest. But the photos show they're lined up somewhat better i.e. here and here and here ?

Pictures 2, 4, 5, 11, 13, 14, 15, 16, and 17 at http://fileshosts.com/intel/Asus/P5W...rks/index.html illustrate it best.

http://fileshosts.com/intel/Asus/P5W.../photos/2.html

http://fileshosts.com/intel/Asus/P5W.../photos/4.html

http://fileshosts.com/intel/Asus/P5W.../photos/5.html

http://fileshosts.com/intel/Asus/P5W...photos/11.html

http://fileshosts.com/intel/Asus/P5W...photos/13.html

http://fileshosts.com/intel/Asus/P5W...photos/14.html

http://fileshosts.com/intel/Asus/P5W...photos/15.html

http://fileshosts.com/intel/Asus/P5W...photos/16.html

http://fileshosts.com/intel/Asus/P5W...photos/17.html

The centre is covered in nail polish and dielectric grease which I have since cleaned off.

Both E6700 and P5W DH pictured above still work fine as i tested them both without problems. So was I lucky to catch this early before something bad could have happened ? Should i even bother cleaning off the cpu's gold contacts' corrosion ? And if i clean them how should i go about it ?

Guess there's a lession to be learnt, insulating LGA775 sockets is hard to do so using dielectric grease in all of the LGA775 cpu socket covering pins and all is probably a good idea for long term runs on sub zero cooling! I only used dielectric grease in the inner square socket but not on the LGA775 pins in the motherboard :o

For now plucked E6700 into one of my Intel D975XBX Badaxe boards for more extensive testing to ensure cpu is fine. So will have some E6700 + D975XBX Badaxe benchies later I hope

[AndrewZorn]

now i know...
one less thing to go wrong when i finally get phase

[Speederlander]

I thought the generally accepted rule was load grease on everything that could possibly suffer from condensation, most especially the pins.

[MrDeeds]

Thats messed up. Thanks for the heads up, i was considering testing a p5b deluxe i just bought without any dielectric grease because i didnt want to clean up the mess if i ended up not liking the board. I'm sure i wouldnt have done that much damage over a couple hours but no condesation is better than some condensation.

There has to be a better way to insulate a LGA775 socket.

[^don.k's^]

I ll see this weekend how it goes under my 6700, it has been some weeks with the phase change...

[eva2000]

Yeah some folks say if you have an air tight seal you don't need dielectric grease, but how do you know if you have an air tight seal ?

I think i will use dielectric grease from now on all of the socket pins too!

Update: seems E6700 looks okay so far



@3600mhz at 1.394v

[freecableguy]

always use grease on everything, middle of socket, socket pins, bottom of CPU...insulate everything from air...

[Durzel]

I assume that even applies on Vapochills with the heater on the back of the motherboard? Asetek specifically state not to use grease in the LGA775 socket, and the manual for the socket kit doesn't mention it at all!

Also, in this thread can someone PLEASE tell me exactly what FUGGER means by:

On the backside there is nothing in the way except the small pins that must be trimmed like on all motherboards to prevent shorting.

[eva2000]

Yeah i'm scared enough to always use it in future.. such a close call heh

luckily cpu is more than fine @3814mhz at 1.465v on Hydrocool 200EX 1/4" OD tubed water

[GWillakers]

First of all, let me extend my condolences....
I know she isn't dead but.... sorta like a diagnosis of cancer...

But do you really wanna put that chip into yet another good board?
Now I'm not sure about this... but isn't corrosion sorta like cancer or rust....
Aren't you now looking to spread it to the badaxe?

This is one of those interesting posts that will stick with me for a while...

I guess this is one reason why they run computer equipment in dehumidified rooms.

[eva2000]

i don't think corrosion works like that unless moisture/water/condensation is present again..

Since i'm not on subzero cooling but plain 28+C water cooling it should be fine

[Durzel]

eva2000: Are there any socket/clamshell heaters with that Kayl Black Frozen SS?

[eva2000]

no heaters with Kayl unit.. it doesn't get that cold really - the evap read no load -44C, idle @4300mhz at 1.66v -38C and loaded -31C while windows as around -26C idle and 0C loaded

But guess air got in under neath the insulation..

[One_Hertz]

hmm weird I used a 6400 on a P5W DH deluxe for 2-3 weeks before changing to a 6700 and noticed no signs of corrosion or anything anywhere and i am not using a drop of grease. I am using a chilly1 unit and for insulation I have chilly's foam ring, sticky foam type insulation on the mobo and liquid rubber to fill the tiny spots I have left out. I guess I had an airtight fit on the 6400, I just hope it stayed that way when i changed cpus. I spent like 3-4 hours making that insulation.

[eva2000]

Yeah that's the only prob not knowing 100% whether you have an air tight seal...

[jimmyz]

I assume that even applies on Vapochills with the heater on the back of the motherboard? Asetek specifically state not to use grease in the LGA775 socket, and the manual for the socket kit doesn't mention it at all!

Also, in this thread can someone PLEASE tell me exactly what FUGGER means by:

when the solder points on the back of the mobo stick out you need to trim them down some to prevent them from going through the insulation and make contact with the metal backing plate which would be very bad.

[Durzel]

when the solder points on the back of the mobo stick out you need to trim them down some to prevent them from going through the insulation and make contact with the metal backing plate which would be very bad.

Ah right.. thanks!

What's the best way of trimming them down then? (complete newbie question) What tool(s) should I use?

[hipro5]

George the corrosion happens because of these "opens".....Air comes and goes through them when the CPU is warm and then cold.....
Prommie for life..... BEST sealing ever......

[eva2000]

ah so i should further seal those gaps ? or the outer evap enclosure ?

[hipro5]

Whatever must be done is around the socket.....You must firstly use Plastic spray that can handle at least -70*C around the socket's components (I even put this plastic spray underneath the CPU - it's tiny components - ) so as to be sure and afterwards place the foam tape around it - but firmly placed so as to minimize those "opens" - ....

[eva2000]

I see. thanks for the tips

[eva2000]

thank god cpu is fine

Some Super Pi action @3814mhz

32M single


32M dual


Unfortunately, Badaxe board's 2:3 divider doesn't really work that well compared to P5W DH. But Badaxe's 4:5 is better or more stable than P5W DH

[Vassili]

Dielectric grease in your LGA755 socket, doesn't cause that any problems like some pins not making contact to the contactpoints on the CPU, or do I see this wrong?

[Durzel]

Dielectric grease in your LGA755 socket, doesn't cause that any problems like some pins not making contact to the contactpoints on the CPU, or do I see this wrong?

The pins will displace the dielectric grease.

[^don.k's^]

Well, i removed the SS, my 6700 remains like new, i used a lot of vaseline, look at the pics :







Since i got the same corrosion as you with a 146 on my first dry ice session, i always use vaseline.

Regards!