hello! i know that copper is better than alluminium, when doing water blocks. but how is silver? 97% silver. pure silver is too soft to work with... can anyone shed some thoughts on this matter? thanx, Ark!
Silver is the best material of all really, but it's somewhat expensive... If you can afford a solid-silver block, just for kicks, go for it. Otherwise make do with copper, there's virtually no difference, and you can easily bypass the results by making the loop more efficient.
thank you! i cant afford the material! but i have a few kilos of heirloom silver - a set of utensils, but they were damaged in transit, so i was thinking of melting them and doing something worthwhile with the stuff. the material is 96.5% silver, 1% tin, 2% copper and 0.5% gold god knows how/why the gold came in, but there you are... will this material work suitably for water blocks? thanx, Ark!
Maintenance on a silver block might be a triffle awkward, what with having to have the butler polish it every two weeks. Though, if you HAVE the material and really want to melt it down, and have the ability to actually cast it, I say go for it! But you have to post pics of the finished thing
Please note that if you melt the silver it will oxidize quickly. The process will introduce lots of contaminants to your alloy. You need some smelting flux to isolate the molten silver from air and to make the contaminants rise to the top so you can scoop them out.
copper don't expand. using material that is capable of thermal expansion is bad 'cause the contact surfaces covered by thermal paste will gradually degrade. dry thermal paste with larger thermal cycle variance (power on and off) ain't exactly good. natural vertical cracks will appear in the paste reducing heat transfer efficiency.
hello all! thanks for all the replies! @cinnander: sad to say, i dont have a butler, and i dont fancy taking apart my computer every 2 weeks to polish my blocks. i was thinking of coating the blocks in gold - the process is known as gold watering here, they dip the stuff in a gold salt, and do electrolysis, very much like anodising. it deposits a few microns thick layer of gold. - and its quite cheap. only a few hundred rupees. like 5 or 6 USD. another option is to coat the non working surfaces with a clear coat of epoxy or PU please let me know of your thoughts. i dont have the necessary metalurgical skills to melt the silver and cast it. so i am taking the help of the local jeweller, whose labour force is quite skilled. - he will be melting the silver, casting it into cube shapes, remelting the scraps, and doing the gold watering i dont have a lathe or a mill yet. but i have ordered a lathe (it should arrive by mid-february), and i am currently gathering money for a mill, so i was just sounding off my ideas off the many people here that are definitely more experienced than me. i shall definitely post pics when i do this. @Javerh: i shall convey this to my jeweller, though i think he knows this already. @ryanjleng: are you saying that silver expands?? or that it expands more than copper??? thanx for all the help guys, i really appreciate it! Ark!
Coating with gold sounds like the best idea imo, as you'd need to do the inside of the water channel, to avoid/delay corrosion there as well.
Did you think of how silver/gold will react with your cooling fluid? Sorry to add to your list of questions.
Technically, diamond would be the best material one could use, but that's rediculously out of the question. I am a strong supporter of copper as far as price/performance goes. What sort of thermal paste do you plan on using? May I suggest this: http://hackedgadgets.com/2007/12/05/diamond-thermal-grease/
Silver is toxic, which is good for killing any algae which might grow in the loop. So now you want to gold plate some silver? Then you have to think about the other things. You're going to need to thread the waterblock to accept some kind of barb or fitting. Can this be easily done to silver? Will it hold? If silver has a much higher thermal expansion coefficient than typical water block materials, how will this affect the seals? Now you have to worry about whether your barbs will loosen up or leak. If anything, maybe you should just stick to copper and silver-plate the block of it's possible. This is not entirely a bad idea, but you may be in over your head. A jeweler may be able to help you out with the silver, but you'd be best to find someone who has worked with silver in an industrial setting. In terms of thermal conductivity, the material doesn't matter. Copper and aluminum have slightly different properties, but in a real world scenario, you'd be hard pressed to see significant results. The difference between silver and copper may in the end prove to be negligible as well. Does the end justify the means?
Theoretically silver would be the best material for a water block - although soft it is durable enough (after all a water block doesn't take much stress) and it is the best thermal conductor which can be acquired. The only problem is it's price - it is a few times more expensive than copper, so it's just not worth it. After all copper's thermal conductivity is not much worse. I myself am currently designing a set of water blocks for my machine - i'm using 10 milimeters thick copper plate in which I intend to cut channels for the coolant and seal them with a plexy cap - pretty standard design. I plan to put water blocks on almost anything that generates heat - CPU, northbridge, memory, GPU, VGA memory, hdd...
hello all! thank you all, for your thoughts. i just want to mention that cost is not a factor here, as i have the silver sitting in my cupboard right now. now that, that is aside, i found some figures: Silver Thermal conductivity [/W m^-1 K^-1]: 430 Silver Coefficient of linear thermal expansion [/K^-1 multiplied by 10^6]: 18.9 Copper Thermal conductivity [/W m^-1 K^-1]: 400 Copper Coefficient of linear thermal expansion [/K^-1 multiplied by 10^6]: 16.5 Alluminium Thermal conductivity [/W m^-1 K^-1]: 235 Alluminium Coefficient of linear thermal expansion [/K^-1 multiplied by 10^6]: 23.1 so silver expands more (bad) and conducts more (good) and copper expands less (good) and conducts less (bad) so the million dollar question: in all of your experiences, assuming you do HAVE a supply of silver, would you cast a block of silver, insteaad of copper? @cinnander: i have access to very pure water, but your idea is quite nice, coating the insides would decrease corrosion related damages. @thacrudd: i want the questions! better to realise the problems now, and solve them, than an "oh f*ck" moment later! my choice of cooling fluid would be plain and simple H2O. i have access to calibration grade water, that is first reverse osmosis'ed and then double vacuum distilled to remove trace gassess. i dont know much about cooling fluids, so if you can suggest a better fluid, i am all ears isopropyl alcohol is too flamable, kerosene doesnt carry enough heat, ethylene glycol - i dont know, i would love to hear you thoughts on this matter. i dont think gold (pure gold) will not react adversly to water (also quite pure). @allockse: wow! diamonds? i think that is way out of my league. in the article you posted, they say: i wonder how does graphite figure in the scale, as i currently use graphite goop as a thermal paste. 30 USD would equate to 1200 of my currency (INR) it wouldnt be THAT expensive. but what can be the carrier? silicone grease? or hard petroleum jelly? definitely an avenue worth venturing! but http://www.heatsinkfactory.com/ic-diamond-7-carat-thermal-compound-15-gram-p-16605.html is even better!! but is this good? @stuey: silver is toxic, yes, but silver oxidises to quickly too, thats the reason for opting for gold plating the silver i have, is quite hard, and i think it can take threads, taps and other delicate machinings quite well. after reaching the point, where i actually start doing the machining and stuff, i will have had spent very little. so if my silver doesnt take the torture, i can simply fall back to copper... maybe i can melt a small part (maybe a goblet or a side dish or some such) and test out the hardness myself, eh? as to leaking, silver is 18.9 and copper is 16.5 (thermal expansion coef) in percentages, silver expands by 14.54% more than copper. will silver blocks really leak? i havent worked with silver and i dont know anyone who has worked with silver, but i am going to try. if it fails, i can always melt it down and try again. cant i? @Dare Devil: well, i have the silver with me, gathering dust. so the thought. i would love to see the pics of your progress. are you posting your modlog here? thanx again! Ark!
Of course it does. Given the operating temp of a waterloop I'd think that thermal expansion would be a negligible consideration.
Good question regarding graphite, I don't know the specs off the top of my head, but I assume it wouldn't be too far off of the materials we've already discussed. I like the idea of having a diamond based thermal paste because it isn't conductive, and one could avoid the offchance of a short caused by improper application. The article said any silicon based carrier containing both Polydimethylsiloxane and Polytetraflouroethylene will work fine. I'm sure its available online. I clicked on that link you provided, definitely worth picking up if you don't want to mix it up yourself. Pros and Cons on either side. Good luck with your waterblock!! Remember that your block design should be relative to the limitations of your water pump. I remember reading some specs at swiftech describing how different blocks work more efficiently depending on pump speed/pressure.
