Pro/Forums - View Single Post

Ultraviolet · 02-06-2006, 01:41 PM

Thanks for all the responses--I'm pretty excited about this project. To answer some of the questions and give a little background as to why I am doing this: There is simply no other place to stick a loud server where I am currently living. I have a nice big basement---and it floods constantly and is 99% humidity when it's not flooded. There are certainly more practical approaches to this problem--I'm not really all that interested in them. This is a hobby and I'm looking for a challenge and this seems to be a match. Additionally, another hobby of mine is woodworking. For the last few years, I've actually been constructing the cases of my computers out of wood (last one has a wine holder and 15 cd holder built in) as I despise big hunks of white and black plastic laying around my house. Now I'm thinking far enough outside of the box that I want this server to be included in a functional piece of furniture (probably an endtable constructed out of cherry). The wood will further dampen any noise. Originally I was approaching this from the standard watercooling perspective and was designing waterblocks for the hdds, mosfets etc as I did not want a single fan in this system (and as this piece of furniture is not going to be designed with venting heat in mind). Then I stumbled upon this submerged cooling idea....it solves so many problems (while creating so many more

The end goal however is a 0db RAID-5 server (or as close to 0db as I can get) enclosed in a gorgeous piece of furniture. This will not look or sound like a computer.

I've done a lot of searching on hdd submersion in Fluorinert and everyone says it can't be done, but it doesn't appear that anyone's done it. I have some old hdds about so I will definitely give this a shot. If it doesn't work, I would probably then move on to ricecrispi's suggestion and see if I can somehow seal them up, but still have them submerged. Maybe have a snorkel coming up for some breather hole??? Who knows...I do have a few questions though:

1. Came across this seller on ebay http://cgi.ebay.com/Fluorinert-Elect...QQcmdZViewItem who is selling Fluorinert FC-40 and FC-43. All mentions I've seen of people using Fluorinert in a computer application were using FC-77. Can anyone see why FC-40, 43 would not workhttp://products3.3m.com/catalog/us/e...er/output_html? At this price, the project actually becomes affordable.

2. This is the part I was really struggling with in my original design--any ideas of a radiator for complete passive cooling? There will be no overclocking and likely no video card (certainly nothing hot in any case). Maybe this part needs to wait until I see if hdds will work in this stuff as that would add a lot more heat to the system.

Thanks again for all the help and comments.

edit: Looks like FC-40 may have about 2.5x the viscosity of water, and FC-43 nearly 3x if I'm looking at this correctly. That certainly is not going to help the hard drive's chances in this....

02-06-2006, 01:41 PM	#22
Ultraviolet Cooling Neophyte Join Date: Jul 2005 Location: US Posts: 22	Re: Galden ZT 150 Thanks for all the responses--I'm pretty excited about this project. To answer some of the questions and give a little background as to why I am doing this: There is simply no other place to stick a loud server where I am currently living. I have a nice big basement---and it floods constantly and is 99% humidity when it's not flooded. There are certainly more practical approaches to this problem--I'm not really all that interested in them. This is a hobby and I'm looking for a challenge and this seems to be a match. Additionally, another hobby of mine is woodworking. For the last few years, I've actually been constructing the cases of my computers out of wood (last one has a wine holder and 15 cd holder built in) as I despise big hunks of white and black plastic laying around my house. Now I'm thinking far enough outside of the box that I want this server to be included in a functional piece of furniture (probably an endtable constructed out of cherry). The wood will further dampen any noise. Originally I was approaching this from the standard watercooling perspective and was designing waterblocks for the hdds, mosfets etc as I did not want a single fan in this system (and as this piece of furniture is not going to be designed with venting heat in mind). Then I stumbled upon this submerged cooling idea....it solves so many problems (while creating so many more The end goal however is a 0db RAID-5 server (or as close to 0db as I can get) enclosed in a gorgeous piece of furniture. This will not look or sound like a computer. I've done a lot of searching on hdd submersion in Fluorinert and everyone says it can't be done, but it doesn't appear that anyone's done it. I have some old hdds about so I will definitely give this a shot. If it doesn't work, I would probably then move on to ricecrispi's suggestion and see if I can somehow seal them up, but still have them submerged. Maybe have a snorkel coming up for some breather hole??? Who knows...I do have a few questions though: 1. Came across this seller on ebay http://cgi.ebay.com/Fluorinert-Elect...QQcmdZViewItem who is selling Fluorinert FC-40 and FC-43. All mentions I've seen of people using Fluorinert in a computer application were using FC-77. Can anyone see why FC-40, 43 would not workhttp://products3.3m.com/catalog/us/e...er/output_html? At this price, the project actually becomes affordable. 2. This is the part I was really struggling with in my original design--any ideas of a radiator for complete passive cooling? There will be no overclocking and likely no video card (certainly nothing hot in any case). Maybe this part needs to wait until I see if hdds will work in this stuff as that would add a lot more heat to the system. Thanks again for all the help and comments. edit: Looks like FC-40 may have about 2.5x the viscosity of water, and FC-43 nearly 3x if I'm looking at this correctly. That certainly is not going to help the hard drive's chances in this.... Last edited by Ultraviolet; 02-07-2006 at 05:40 AM.