mercury cooling system

thaltek

Limp Gawd
Joined
Sep 15, 2009
Messages
206
Bill, notice I focused ONLY on motors... Not on actual E&M. "The idea that your pump will mess up your disks at all is a complete fallacy." I didn't say that magnetic fields couldn't mess up your disks. ;)

On another note, do HDD's have faraday cages in them? I don't see one through the window of my WD RaptorX. Or is the steel casing just enough to act as one?

if the plater is electrically isolated from the disk housing then the housing can act as a Faraday cage..... however this only remains the case so long as the magnet is not brought into contact with the disk housing.....as was pointed out earlier putting a horseshoe magnet onto the drive has a negative effect.... well.. lets consider two things... 1. the lid of the hd is made of steel which happens to be a magnetic material.....2. point focus of the magnetic field.. which makes the field more directed....... as for the magnets inside the drive you will notice they tend to always be oriented in a specific way, there is a reason for this.....

ok now can we please get back to talking about diamond thermal interfaces again......
 
Last edited:

BrainEater

[H]ard|Gawd
Joined
Jul 21, 2004
Messages
1,087
Back to diamond interfaces in a min....

:p

---------

Here's the deal with a harddrive :

The magnets are setup in such a way that exterior magnetic fields are minimized....(a 'core' if you will.....) (( Same idea as a 'shielded' speaker))

I'm working on some pics right now.....I am very curious as to hdd magnetic tolerance.....

----------

Diamond CVD is something the home experimenter can do.There are several methods.

The commecially grown diamond is done in a microwave......Input gases include methane , ammonia , n2 , co2......

It can be done in a plasma arc as well.......

:D
 
Joined
Oct 2, 2003
Messages
680
an idea would be to take them and mix in with molten copper and pour into a mold to make a plate for a hsf or water block.....

I wouldn't do it this way. When you melt non-ferrous metals (might be the same with ferrous as well, I don't have experience in that field) they absorb gases from the atmosphere like crazy. these gases form small pockets when the metal solidifies called porosity. As a jeweler I am constantly fighting this when casting the metals we use. We add boric acid to the metal while it's melting as well as fill the chamber where the metal melts with a mix of gases to form a protective layer and prevent oxygen from reaching the metal. the issue would also be compounded because of the uneven cooling caused by the addition of the diamonds. As the metal cools it contracts and the dissolved gases move toward the warmer still liquid areas. The porosity would cut down on the cooling of the block because of the tiny air pockets.

a better way would be to start with an extruded peice of copper sheet and cut holes for the diamonds and set them down with the table side facing out (to maximize their heat transfer surface) and then flow silver solder around the edges to make it free from leaks.


sorry had to throw that in there, in the computer forums aside from the modding section my expertise rarely comes in handy.
 

Smiththers2

Limp Gawd
Joined
Jun 2, 2009
Messages
501
i had a hard drive that i couldnt format once because i plugged it into a usb to ide controller that didnt support its size. WD wouldnt RMA it for me, so i just set a JL Audio 12w3 subwoofer on it for a while. plugged it in and it showed up as a hard drive again. magnets do make a difference. some of the time however, the field MIGHT not be directed at the hard drives, thus not causing damage. either way, i dont use any non shielded speakers near my computer anymore.
 
Joined
Oct 1, 2008
Messages
687
Well, looks like I just mad an ass of myself. I was making the assumption that a magnetic pump designed to move metal with enough force to actually make a difference in a mercury cooling system would generate a significant enough field to cause issues with the hard drives. Apparently I am wrong.

And I have taken apart a considerable amount of hard drives, I use the magnets and a couple of metal strips to hold all of my tools up in a lab.

But I do concede to being wrong about the mag pumps. Pump away!
 

BillParrish

Supreme [H]ardness
Joined
Aug 25, 2006
Messages
7,520
Last edited:

Zoson

Supreme [H]ardness
Joined
Aug 4, 2001
Messages
6,113
Yes I have actually, I have one of these: http://www.grainger.com/1/1/9049-approved-vendor-6xy59-horseshoe-magnet-max-pull-30-lb-alnico-5.html and it completely screws up a HD if I set it directly on the top of the drive for a couple of minutes. I use this to destroy data on customers unwanted/damaged HDs before tossing them.

The motor is a direct drive DC servo spindle type and the outer motor casing acts as the motors shield.

Still got all your fingers? ;)
 

BillParrish

Supreme [H]ardness
Joined
Aug 25, 2006
Messages
7,520
mIssing a toe, I was pulling the thing off of my bench vise when it kinda got away from me because of the slick paint and droped the damn thing on my foot. Could not walk for a week. Got smart and put an bolt thru the hole and attached a wooden handle.
 

Riccochet

Fully [H]
Joined
Apr 11, 2007
Messages
27,968
I'd be interested to see some coolers made of the tile material on the space shuttle. While it acts as an insulator on the shuttle it has rather extraordinary heat dissapation properties. Run some heat pipes through it. AFAIK Lockheed is the only company to make the stuff and they won't send me a sample. Wrote them a letter and tried. :/
 

thaltek

Limp Gawd
Joined
Sep 15, 2009
Messages
206
I wouldn't do it this way. When you melt non-ferrous metals (might be the same with ferrous as well, I don't have experience in that field) they absorb gases from the atmosphere like crazy. these gases form small pockets when the metal solidifies called porosity. As a jeweler I am constantly fighting this when casting the metals we use. We add boric acid to the metal while it's melting as well as fill the chamber where the metal melts with a mix of gases to form a protective layer and prevent oxygen from reaching the metal. the issue would also be compounded because of the uneven cooling caused by the addition of the diamonds. As the metal cools it contracts and the dissolved gases move toward the warmer still liquid areas. The porosity would cut down on the cooling of the block because of the tiny air pockets.

a better way would be to start with an extruded peice of copper sheet and cut holes for the diamonds and set them down with the table side facing out (to maximize their heat transfer surface) and then flow silver solder around the edges to make it free from leaks.


sorry had to throw that in there, in the computer forums aside from the modding section my expertise rarely comes in handy.

>_> there are ways that prevent air bubbles from even happening.... :D

HydroBudz said:
AFAIK Lockheed is the only company to make the stuff and they won't send me a sample. Wrote them a letter and tried. :/
lol
011.gif

http://www.thespacestore.com/spshthti.html
 

keenan

2[H]4U
Joined
Aug 5, 2009
Messages
2,695
OK apparently no one here knows how magnetic fields and Faraday cages work. Stop it.

Since Faraday cages don't necessarily affect magnetic fields at all, I'll just leave the rest of your posting alone and not deconstruct it any further.

Magnetic shielding is difficult and certainly not employed in hard disk drives. What's happening is that the read/write head is so close to the platter and the induced field so tiny that its field strength is astronomical. Generating similar field strengths at any practical distance from the platter would be basically impossible (well, okay, it's possible but not in this kind of situation).

Edit: Okay, I tried to find out some actual numbers, which is a bit difficult since the field strengths involved are going to be implementation specific. However, I did find this patent from 2004 that describes a magnetic write head, presumably for a hard disk. They quote a saturation magnetism of > 2.45T, which doesn't mean a value that high is actually used, but since it appears to be a desirable attribute, I would suggest that fields in this order are used. Given that MRI machines produce larger fields of this order and are considered extremely strong and dangerous, you're not going to want a field that strong in your house, let alone in your computer. For reference, the surface strength of a NIB magnet is around 1T, and falls off quickly (and larger ones are quite dangerous, did you see the xrays of the guy who smashed his finger seriously being careless with one? - again not something you want just laying around). And even these magnets would have to be pretty much placed on top of the platter to make a difference. Magnets of the sort used in motors are an order of magnitude weaker yet, though their fields tend to extend further.
 
Last edited:

BillParrish

Supreme [H]ardness
Joined
Aug 25, 2006
Messages
7,520
You are correct I mispoke although taken to extream (as the wavelength approaches 0 ) they both (magnetic and electromagnetic sheilding ) both resolve to a solid ferrous box or plane. A copper Faraday box would be a fairly useless magnetic sheild and in that respect I was wrong.

The question was concerning why the fairly strong magnetic field of the disk drive motor itself does not affect the magnetic platters. I submit the disk drive's motor housing is the sheild and to some extent for external fields the outer case itself and it is fairly effective as the practical experiment by the previous poster with his pump motor in very close proximity to his drives. ( However a strong enough field will certanly do the trick all one has to do is google "HD degausser" for commercial products. Basically just strong electromagnets that totaly erase a drive without opening the case.

Numbers are great, used them all my life, just next time you have an old drive still readable just "stick:" it to the back of a large magnet (one from a decent woofer should do) for an hour or so.

Me personally I keep the pump away from the drives but unless it is a big pump or other strange circumstances I have yet to see a post that it causes any major issues.
 

octoberasian

2[H]4U
Joined
Oct 13, 2007
Messages
4,082
Wasn't a computer company working on such a device using liquid metal to cool a laptop's CPU?

I think it was Fujitsu, but I have to look for the article again. Their idea was to use something OTHER than mercury, and the pumps would be electromagnetic. It'd be practically silent.

However if this is a homebrew DIY project, you'd have to come up with a method of how to safely contain the liquid metal; pump it without it destroying the pump; and be effective at dissipating heat. If one can make a homemade EM pump and implement, better post your project here.

Though, don't patent it, that computer company may have already patented it. You know how companies are nowadays with patent infringements... >_>
 

thaltek

Limp Gawd
Joined
Sep 15, 2009
Messages
206
Wasn't a computer company working on such a device using liquid metal to cool a laptop's CPU?

I think it was Fujitsu, but I have to look for the article again. Their idea was to use something OTHER than mercury, and the pumps would be electromagnetic. It'd be practically silent.

However if this is a homebrew DIY project, you'd have to come up with a method of how to safely contain the liquid metal; pump it without it destroying the pump; and be effective at dissipating heat. If one can make a homemade EM pump and implement, better post your project here.

Though, don't patent it, that computer company may have already patented it. You know how companies are nowadays with patent infringements... >_>

no.... if they want to have rights on the design they will already have it patented..... in that case they have little room to gripe......
 
Joined
Oct 1, 2008
Messages
687
BrainEater said:
Diamond CVD is something the home experimenter can do.There are several methods.

The commecially grown diamond is done in a microwave......Input gases include methane , ammonia , n2 , co2......

It can be done in a plasma arc as well.......

Wait.. Really? Actual synthetic diamond that has the same heat transfer and capacity as would be grown in a massive lab?

Looks like I have a new project..
 

osrk

[H]ard|Gawd
Joined
Jan 10, 2003
Messages
2,010
:rolleyes: next door...... that will remain my answer until orsk decides to read the first 5 posts.....

Well then I'm sorry for your terminal cancer and that you only have ten more days to live :|. Better get started on your mercury cooling.

Also to those who said you it can't be moved yes it can, caterpillar drives are in theory possible but you'd basically need superconductors and using liquid heluim isn't really particle. Also unless you're using floppy disks to store your porn your hard drive would not be affected.
 

BrainEater

[H]ard|Gawd
Joined
Jul 21, 2004
Messages
1,087
Wait.. Really? Actual synthetic diamond that has the same heat transfer and capacity as would be grown in a massive lab?

Looks like I have a new project..

Yes.

Now it's certainly not easy, it would take some foolin around I'm certain.

Making CVD diamond is a careful ballet of plasma properties and input gas adjustment.....the pro's can actually select the plane they want to grow the diamond in...

It's on my project list too.

:D

---------

I've seen those numbers before myself Keenan...And my experimental data appears to back that up.Your everyday normal magnet (especially 'cores') won't do it easily.

It's certainly possible tho.....My AC electromagnet degausser does'nt have any issues erasing HDD's....

I've heard HDD cases are actually a mu-metal alloy .....true ? false ? I'm not sure.

-------

Also to those who said you it can't be moved yes it can, caterpillar drives are in theory possible but you'd basically need superconductors and using liquid helium isn't really practical.


MHD pumps don't need all that.You can make one in your workshop...magnets + current =force , as per the Right hand rule....

The biggest problem with MHD pumps and liquid metals is the fact that the liquid metal dissolves the electrodes....they have an insane failure rate.

Lasty ,

I have to say using mercury is not the way to go...The dangers waaay outweigh the cost of using Galinstan , for the same performance.

Solve the amalgamation problem first......People will sponsor you for the Galinstan after.

:D
 

Yuritau

n00b
Joined
Sep 9, 2009
Messages
59
Heh, if you truly solve the amalgamation problem, make sure you patent it before looking for sponsors. :p
 

Yuritau

n00b
Joined
Sep 9, 2009
Messages
59
you realize that with a vacuum pump you'd vaporize some of the mercury and expel it into your environment? The previous post was referring to something more contained.

That's basically what I was getting at. I apparently need to work on my clarity, hehe.
 

BrainEater

[H]ard|Gawd
Joined
Jul 21, 2004
Messages
1,087
"Ease down Ripley."
:p


the amalgam problem can be solved by taking atmosphere out of the equation.......

Negative.This is not correct.

Atmosphere has nothing to do with it.Liquid metals are solvents for solid metals, period.....It's just like paint and paint thinner.

Even using refractory metals has issues.MHD pumps are fully functional in nuclear reactors right now though....they can and do work.

:D
 

thaltek

Limp Gawd
Joined
Sep 15, 2009
Messages
206
"Ease down Ripley."
:p




Negative.This is not correct.

Atmosphere has nothing to do with it.Liquid metals are solvents for solid metals, period.....It's just like paint and paint thinner.

Even using refractory metals has issues.MHD pumps are fully functional in nuclear reactors right now though....they can and do work.

:D
you might want to explain that to the heat pipes in my system...... not that you are wrong......
 

BrainEater

[H]ard|Gawd
Joined
Jul 21, 2004
Messages
1,087
I honestly could'nt tell you what reactors.I would suggest they are of the experimental /research type....monju maybe ? FFTF ? hard to say...Even if I had the clearance to know who was doing what in this field , I'm fairly sure I couldnt tell you. :D

Fast flux reactors typically use liquid sodium metal / liquid sodium- potassium alloys , as the primary coolant.....If you do a little gooogling on 'em pumps' you'll find several references to this use.


you might want to explain that to the heat pipes in my system...... not that you are wrong......

Heat pipes have nothing to do with liquid metal.(at least the ones in computers don't)

A heatpipe is essentially a copper tube with a drop of 'refrigerant' inside.

See here : http://www.benchtest.com/heat_pipe1.html

:D
 
Last edited:

osrk

[H]ard|Gawd
Joined
Jan 10, 2003
Messages
2,010
I honestly could'nt tell you what reactors.I would suggest they are of the experimental /research type....monju maybe ? FFTF ? hard to say...Even if I had the clearance to know who was doing what in this field , I'm fairly sure I couldnt tell you. :D

Fast flux reactors typically use liquid sodium metal / liquid sodium- potassium alloys , as the primary coolant.....If you do a little gooogling on 'em pumps' you'll find several references to this use.

*shrug* maybe the fast breeders there are only a few in the world that are actually up. Other millitary reactors use water as a moderator and coolant so there's no need. The russians used sodium metal though...
 

e39dinan

n00b
Joined
Sep 11, 2009
Messages
53
Come on guys, I'm sure there's at least one dangernerd here with the resources to pull this off.
 

BrainEater

[H]ard|Gawd
Joined
Jul 21, 2004
Messages
1,087
Heh.

I'm busy using my 'dangernerdyness' in my effort to do anhydrous alcohol submersion cooling... :p

------

Yea fast breeder reactors are the only earthbound ones considering/testing/using this kind of pump.....I know NASA/DOE are looking into them for space uses (ALIP program)...

:D
 

thaltek

Limp Gawd
Joined
Sep 15, 2009
Messages
206
I honestly could'nt tell you what reactors.I would suggest they are of the experimental /research type....monju maybe ? FFTF ? hard to say...Even if I had the clearance to know who was doing what in this field , I'm fairly sure I couldnt tell you. :D

Fast flux reactors typically use liquid sodium metal / liquid sodium- potassium alloys , as the primary coolant.....If you do a little gooogling on 'em pumps' you'll find several references to this use.




Heat pipes have nothing to do with liquid metal.(at least the ones in computers don't)

A heatpipe is essentially a copper tube with a drop of 'refrigerant' inside.

See here : http://www.benchtest.com/heat_pipe1.html

:D
http://en.wikipedia.org/wiki/Heat_pipe
second paragraph and on....... i will also take this opportunity to point out that a year or so ago i purchased a hsf that had a mercury warning tag with it that specifically stated that mercury is used in the heatpipes of that hsf.......
 

BrainEater

[H]ard|Gawd
Joined
Jul 21, 2004
Messages
1,087
I have read that before actually.

The materials chosen depend on the temperature conditions in which the heat pipe must operate, with coolants ranging from liquid helium for extremely low temperature applications (2–4 K) to mercury (523–923 K) & sodium (873–1473 K) and even indium (2000–3000 K) for extremely high temperatures. The vast majority of heat pipes for low temperature applications use some combination of ammonia (213–373 K), alcohol (methanol (283–403 K) or ethanol (273–403 K)) or water (303–473 K) as working fluid.

You'll note that the working temperature range given for mercury is 250-650 C . This is outside the normal working range of any computer parts.

I am certainly not saying it's not possible , but I am very surprized to hear there are actually mercury heatpipes in computer parts.What kind of Heatsink is this you have ?
 

thaltek

Limp Gawd
Joined
Sep 15, 2009
Messages
206
I have read that before actually.



You'll note that the working temperature range given for mercury is 250-650 C . This is outside the normal working range of any computer parts.

I am certainly not saying it's not possible , but I am very surprized to hear there are actually mercury heatpipes in computer parts.What kind of Heatsink is this you have ?

its either the stock amd hsf that came with the am2 3.1ghz dual core or this

http://www.newegg.com/Product/Product.aspx?Item=N82E16835119091&Tpk=buffalo

i am a little fuzzy on which one but i do quite clearly recall that one of the two clearly stated the use of liquid mercury in the heatpipes..... as i nearly fell out of my chair when i read the tag....
 

BrainEater

[H]ard|Gawd
Joined
Jul 21, 2004
Messages
1,087
Hmmm

I've looked into this a bit.I can find no confirmation that there is mercury in any computer grade heatpipes , including the Evercools and AMD stock....

It's more expensive,less efficient,corrosive to other metals , etc, etc... with the current push for RoHS compliance as well , I highly doubt it's mercury..If it is get rid of it properly.

:D
 

thaltek

Limp Gawd
Joined
Sep 15, 2009
Messages
206
Hmmm

I've looked into this a bit.I can find no confirmation that there is mercury in any computer grade heatpipes , including the Evercools and AMD stock....

It's more expensive,less efficient,corrosive to other metals , etc, etc... with the current push for RoHS compliance as well , I highly doubt it's mercury..If it is get rid of it properly.

:D

meh..... *shrugs shoulders* it is what it is....... i am starting to wonder what the reactivity with plastics........


[edit] sorry brain eater i am mistaken the mercury warning was actually about the processor and it was more about the lead content not mercury..... and i realllly need to stay off the internet when i am going on 48 hours of no sleep...... >_< collage........ oh and i found the paper thingy.....
http://www.amd.com/us-en/assets/con...ion_-_AMD_Processor_in_a_Box_(PIB)-update.pdf
[/edit]

i got to thinking about it though and if i remember my thermodynamics class correctly a fluid in a such as water in a reduced atmosphere has a lower boiling point, i suspect this is true with mercury as well.... with a vapor operating temp of 523~923 K or 249~650C but this i suspect will vary depending on pressure, that and the pressure associated is not given in the wiki article.....

now if you bring up the potential problem of heat flux with the vapor..... it might just be easier to cave in and go full liquid seeing as mercury is a pretty efficient heat conductor and that would bring down the operational temperature range to be more useful numbers.....
 
Last edited:

thaltek

Limp Gawd
Joined
Sep 15, 2009
Messages
206
ok i am an idiot for not verifying my sleep deprived claim..... sorry bout that... now lets reactivate the thread.


anyone have any ideas about letting thermal convection to move the coolant in the loop move the mercury instead of using a pump....?
 
Top