Hmm.. this is a good issue to consider, maybe AS5 is not as good as Ceramique goop.
Cause I've been thinking what if that those oozed/squezzed out AS5 between the hsf & ihs will contaminate the oil and that stuff is known to be electrically conductive.
DING DING DING! Now you're with me Bing. The story that I heard on here was that the guy got AS5 floating in the oil, and it finally came in contact with the electronics, thus killing them. Is the Ceramique stuff non-conductive? Also, what do you think about not using a thermal compound at all? Any gaps should allow the oil to flow in (what probably caused the other guys AS5 to enter the flow), and thus allow a nice transfer. This is all theory though, and I would hate to bake a Quad just for trying it.
Coating the capacitors sounds like a great idea, if I can find something that would be safe while keeping the oil from coming in direct contact with them. Good thinking Darkcow. I think you are seeing this similar to the way I am, and that makes me feel a little better about doing it.
I will say that the temps should not be 65c. The Puget Custom reached a balance of 88c without circulating the oil. After they added a pump to circulate the oil through a fanless radiator they were able to maintain load temps of 45c. I feel that is a very good result, and would be great if the Q6600 doesn't decide to boil the oil before it has time to circulate. If that happens I may just grab a potato and make some fries j/k. My plan is to pull the oil from the general area of the heat sink, and return the oil on the other side of the tank. This should allow the heat very little time to dissipate throughout the entire tank.
Also, what do you think about not using a thermal compound at all? Any gaps should allow the oil to flow in (what probably caused the other guys AS5 to enter the flow), and thus allow a nice transfer. This is all theory though, and I would hate to bake a Quad just for trying it.
thats a pretty horrible idea judging by the fact that oil has bad thermal characteristics. but yea, ceramique will probably be your best bet. vasiline might be good for sealing your capacitors.
I doubt vaseline would stay on the caps once it was heated. Some kind of spray on heat resistant coating? Another reason to get a sample to test? Try several different coatings on some of the caps and see which one holds up best.
The problem with using another board to test something like the caps is that I might be able to leave it in there a year before I actually see any change. At what point do I call it safe? See what I'm saying?
I really didn't think the no TC idea was a good one, but wanted to throw it out there to see if anyone thought it might be an option. What if I use high temp silicone to seal the gap around the sink once it is mounted? This might also be an option for the caps.
::Small Update::
The phone is holding up fine in the oil. I can see a few spots forming on the LCD which I believe to be air pockets. The overall appearance, and usability of the display have not been hurt though.
I picked up the aquarium today and got it cleaned up. I plan to start taking measurements tomorrow, and getting some angle aluminum to paint and glue at the corners to cover the factory silicone that has fogged. I had the water as hot as possible while cleaning it to see how the seals held. No change at all even after soaking in water too hot to put my hand in.
Bonus! I picked up a free UPS from work. It may need a new battery, but I'm fine with that.
Phone test:
Here you can see the oil level inside the LCD case, so it is definitely soaking.
It is hooked up to the charger with the screen saver running 24/7. I think the LCD idea is going to work just fine b/c the spots are there, but wouldn't bother me over the Folding client. Also, the spots are more obvious in the pics than they are to the naked eye.
*EDIT* I wanted to remove the case, but Sanyo requires a tri-driver that sux to locate. I didn't want to drill it, or just break it out b/c I wanted to make sure it was 100% working in the oil.
I also forgot to add that Puget Custom Computers seems to have vanished. I sent an email early last week, and still no reply. Today I tried calling both of the numbers they have listed on the site, and neither of them worked. I just wanted to get some direct answers about what to expect here.
*EDIT* Finally got in touch with Puget, and they were a great help. See later posts...
just read through the whole thread. that's a really interesting idea. i live about 50 miles away from kent, wa so almost want to go check on that company. that's quite a drive though and i hate seattle traffic. hop on the local branch of craigslist and see if you can find someone to do it for you or knows for sure that they exist.
as far as allowing the mineral oil to replace the thermal compound, that's probably a bad idea. the thermal conductivity of the regular white grease even is like .7-.8 W/m*K and it's 2-3 W/m*K for the good stuff like AS5. the thermal conductivity of mineral oil is about .15 W/m*K. so it's about 5 times more restrictive to heat flow. also, unless you seat the heat sink and processor while submerged, the interface between them will probably stay relatively dry and sealed.
if you could get some decent flow over the processor, i wonder if you could just leave it naked. maybe if you get a junker system to test first that would be something to try.
for the caps, i know sometimes caps will buzz and vibrate and get nice and annoying. electricians have some kind of goop they can put on them to damp out the vibes. it at least isn't electrically conductive. i have no idea how it does in oil though. might be something to look into.
Thanks for the detail on the thermal conductivity of the oil DJ. I had not looked into that so you added some really good info here. I have a friend stationed at McChord AFB up there, and he has agreed to check it out for me, even if he has to go to their door.
Or if they will work with unbalanced cards I will probably do one 16gb and one 4gb for a nice even 20GB drive. On the same topic, if 16GB will be enough for folding & Linux then I will just do the single adapter.
Yes Ceramique is non-conductive. I prefer it to AS5 for this reason (even though I don't go smearing it all over the place)
From Arctic Silver's Site:
Electrical Insulator:
Céramique does not contain any metal or other electrically conductive materials. It is a pure electrical insulator, neither electrically conductive nor capacitive.
I was lucky b/c he owes me a favor anyway. Thanks for the offer though. So if heat transfer is your thing, see if you can suggest a fan-less radiator setup for me. I was looking at a few PC radiators, but I am open to anything that will not raise the price too much.
i couldn't really spec out a good one for you. i said it was interesting, not that i'm a pro at it. i did design one for a project once. that was all on paper though. however, if you're trying to decide between a few, i could maybe tell you which would work best. in a case like yours where i assume this would just be sitting on a shelf next to the tank, you'll want something with as much surface area exposed to the ambient air as possible. not necessarily fin surface area but "outside" surface area if that makes sense.
I will worry about the radiator later. Update w/ pics later today. Not much to see, but I will be getting the aquarium ready.
EDIT: I just got off the phone with Richard @ Puget Computers. He, and the sales rep I spoke with, were very helpful and willing to talk about the project. He was able to shed some light on many of the issues that have been debated, and also brought up a few new ones that had not been considered yet. After our conversation I am 100% prepared to build this project, and am really looking forward to seeing it finished.
Again, sorry for the poor quality of the pictures, but this was all I had time to do today.
This is the tank I am using, and you can see that the factory silicone has fogged so it is very distracting. I cut some pieces of angle aluminum, and painted them black for the corners. I am going to start prepping the bottom, and back of the tank for painting tomorrow (outsides).
These are not glued into place yet, just sitting on it to give you an idea how it will look.
I was thinking a car radiator for price, but I want it to be small enough to not get noticed sitting behind the tank while still having enough surface area to not need fans.
BTW, I have painted the bottom pane of the tank, and it came out great. I am about to head to work, but it should be dry enough to tape off the back and get it painted when I get home tonight. I should get to start ordering parts by the end of the week.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
j/k. My plan is to pull the oil from the general area of the heat sink, and return the oil on the other side of the tank. This should allow the heat very little time to dissipate throughout the entire tank.
