The disassembly problem could be solved by having the retention mechanism separate from the actual block. Think MCW6002. If the CPU could be installed and removed witout detaching the block, you could leak test with the CPU outside of the system.
Printable View
The disassembly problem could be solved by having the retention mechanism separate from the actual block. Think MCW6002. If the CPU could be installed and removed witout detaching the block, you could leak test with the CPU outside of the system.
I'm not sure what you mean. The IHS is replacing the blocks base plate. Out of system testing would be difficult unless someone jiggs up a pseudo mobo to mount to.
What I'm saying is you could test leak with the cpu unmounted; just hanging off the end of the waterblock. Youll have to be careful not to bend pins and yadda yadda.
It'd also make remounting the CPU with the system full POSSIBLE, although difficult. You'd have to unmount the CPU with the block attached, and keep it upside-down, level, and at the highest point of the loop for the whole remounting process. A PITA, no doubt, but better than having to drain and refill your loop with every remounting.
Also, I don't think the gunsmith's whatevers will work well. You'd have to keep the affected area very small so as not to disrupt the block's seal on the IHS. Low-grit sandpaper (de-lapping?) would work better.
And how do you propose to keep the CPU stuck to the top through all that?
That I didn't really think of. A decent adhesive around the o-ring might make it a little more leak-proof, but it'd make the o-ring a pain to remove.
So, yeah. Drain the loop to remove the block is the only option?
All of this is based on conjecture of course. And the potenial lower thermal resistance of a naked or modified IHS. Pursuing this course in 1 direction. brings us back to No IHS and conventional water blocks. Which we allready have aplenty. Interesting thread though. It has really made me think.
Example: So we decide our Direct/IHS doesn't cool well enough. How do we make it werk better? Like D_O_S pointed out.(Real early in this thread) I should have been thinkin'. We can make it bigger. Cure the sealing problem. whadder' we gonna' make it out of? Copper is Ok. Silver is better. A billet cut diamond would be best.(Maybe abit impractical though) We make this new IHS as best as we can. Cut groves or drill for jets. Or??Mount it to the other parts. Cure the sealing problem.But now step back and take a look at it. What have we got? LOL
As to the issue of the O-ring leaking etc, it may be possible to spit this block into two parts:
A lower part that is glued permanently by epoxy (viod warrantee) to the IHS
An upper part screwed to the lower part with O-rings in between. The upper part then would be fixed to the motherboard via an additional plate like the appogee.
But I see a bigger problem: without the extra surface needed to transfer the heat from the IHS to the water, simply passing a stream of water over the IHS will not be enough to absorb the heat.
The water needs a lot of surface to take the heat away. Think all those pins in the apogee for example. In the absence of those pins, a strong jet focussed to semi circular cups may be an alternative. But that is not possible with an IHS. So this is no go either.
Given what the gentleman has alluded in his posts so far with regards to his device properties:
a) will produce IHS surface temperatures very close to ambinet.
b) will have no impact on the pump selection or flow
then the only thing that comes to mind is latent heat of evaporation. It is thoeretically possible, but there would be some difficulties in such a small device. For example vaccum breaks, initial start, non disolved gases and the like (think condenser condensate tanks and ejectors). Who knows, will be a marvel of minutrization.
@Clay: I'd like to think this thread has turned away from a review of NorthWizards claims and towards a "brainstorm" (I hate using cheap marketing terms) of ways it'd be possible to outperform modern water block design limitations. IE, "What's better than slapping a piece of copper on your CPU and running water over it?"
@MeshMesh: Like phase, only using water instead of refrigerant? It's possible, but you'd have to keep it at low pressure to make water boil at near-ambient temperatures. How low of pressure, I can't say, since I don't have a physics reference handy. It seems like it'd be really difficult to keep low pressure on the CPU block and high pressure in the rest of the loop
Going back to the heatpipe ideas, Heatsinks use heatpipes to move heat from the CPU to the fins. In a water cooling loop, the pump moves the heat (hot water) to the radiator, where it is dissipated. CPU blocks are all about transferring heat from the CPU into the water, and a Heatsink's heatpipes are simply soldered to a copper base. Thermaltake made the "Volcano," and it looks like it tries to implement the ideas of a heatpipe cooler into a CPU block. I imagine it sucks.
Now, something like Asetek's Vaphochill Micro might be a better place to look for ideas. That heatsink uses what looks to be heatpipes filled with refrigerant. THOSE should be able to transfer heat better than water-filled ones, and could be used to transfer heat away from the CPU block. That way, you could transfer the heat to the water away from the block, also allowing for a surface not restricted by die size.
Reading over all that again, it seems ridiculously impractical. Something revolutionary is going to have to happen to CPU blocks (Go, NorthWizard, Go!) if water cooling is going to survive. The milled-copper CPU design has reached it's limits. Heatsinks are getting better, and phase is becoming more popular. There are still a few of us who are looking for something that allows great overclocking with low-noise, and can be done by just about anyone. Phase may be easy for those with a double-major in overclocking and HVAC, but the insanely low temperatures and the high-pressure refrigerants make it a little too dangerous for a newb like me.
EDIT: Water @ .042 ATM would evaporate at 32C
[QUOTE=oshox]@Clay: I'd like to think this thread has turned away from a review of NorthWizards claims and towards a "brainstorm" (I hate using cheap marketing terms) of ways it'd be possible to ouperform modern water block design limitations. IE, "What's better than slapping a piece of copper on your CPU and running water over it?"
This morning and last night I really thought that direct IHS was what he was up to. And truthfully it really looks that way. But when I started to think about modding a IHS to work better. It brought me to removing the IHS and making my own. From there it got sealed as 1 assembily. About this time I noticed my sig and said WTF I allready have that! As far as Norths claims go we're really going to have to wait and see.
Low pressure, or high pressure oshox.
Have the inlet close to the IHS so there is only a very small gap between them (1mm?).
You'll need pump with a decent amount of pressure, but you need that for effective direct die cooling anyways really.
What may limit spilling during CPU changes is a "quick disconnect" feature. A main inlet feeds jets similar to what the storm uses (basically holes in a plate) which shoot the many smaller jets onto the main IHS surface area. A small plate can be pulled out slightly to close these smaller jets and the outlet so that only the water trapped on top of the IHS would be able to escape during removal.
You'd end up with a multi-teired baseless storm pretty much.
EDIT - G8 "Summit" for the name? hehe ;)
That would totally work, yes. Just a mid-plate that can. But the surface area issue raised earlier does put a dampener on the whole idea, doesn't it?
I recieved an e-mail back from North Wizard:
Quote:
Hey there Joshua...
Thanks for the email, and I am well aware of the thread over at extremesystems. I have no problems discussing the block, but as you well know, that thread has been 90% bashing... very little of it is anything other than that.
In the thread over at Xoxide forums, I have given every hint I can to the people who have come over from Xtreme - ad I will tell you the same thing.
I agree 100% when people say there is no way to improve or get such a big difference in temps when you are talking about inptoving on current block design... Cathar's Storm is about the best there will ever be, he himself said he would only be able to improve upon it minimally, but the cost would not be worth it... and he is totally correct about that. I can see no way to make any changes to the way blocks are made right now, and get any improvements that would make it worth releasing something.
Now, the fact is, you do not have to follow along current lines of thinking when it comes to making a water block, and if you are able to just drop the whole thought of how they are made right now, and start over from a different point of view - start from thinking of the best way to cool, not of how can I improve, it opens up other avenues - which is all I have done.
What I am working on, like I have said, is neither new to water cooling, nor is it some great, genius, never thought of before, idea... it has been done before, just with very limited success - although the cooling aspect of it worked perftecly.
I am just taking an already proven cooling method and making it something anyone can use.
The other thing is this... I am not a huge company here... and the reality is, if I posted all my drawings and plans, the stuff that gives away what I am doing... if Swiftech or Danger Den decided they liked them and could easily produce them, what would stop them from taking it before I have a chance to finish producing them... they have far more resources and money than I do and could make things happen at a far faster rate - I would be left sitting back with a bunch of drawings, knowing that it was my idea, but never being able to do anything about it.
I don't care about making huge money off of anything like this either, because the fact is... if I bring out a block that can easily beat the storm, I could demand the same price, or more for mine, and people would happily pay it... what I am trying to do, though, is make this block, and still keep it in a fair price range... which is another obstacle to overcome... I want it to retail around the 50$ mark.
If you have ever decided to get something made, you would know what is involved in getting it from an idea, to paper, to actual autocad drawings, to prototypes - then, finding flaws, doing minor changes or complete redesigns... it is not easy, nor is it cheap - the prototyping for the blocks was close to 1500$ - of course, when being made in a production run, when all the programming has been done and the shop is running them in orders of 500 or so... the price drops a huge amount... but to get a single block done, something that is very intricate and detailed, and you can ask other people about this as well, just to verify, it is extremely costly - I have put out all of the money for everything - out of my own pocket.
Another thing.... think about that... I have been the ONLY person putting any money into this idea... no one else. So, how can it be anything other than an idea I actually believe in. I have never went to anyone to ask for support, financially or otherwise - I have been doing this on my own (well, there is another person involved in the designing part of it) and in the end, if the block works, it wil be something that i created and did out of my own pocket, and I may make a few bucks... or it will be a complete flop and I will lose every minute I have put into it, as well as all the money I have invested in it - I already know this, and I am following through with it because I believe in what I am doing.
I am not 'slamming' the Storm when I use it as a reference... I am saying, quite clearly that it is the best out there, so it is the only thing worth using as a reference. As for conventional water blocks (as conventional as the storm can be LOL), Cathar made an awesome block, and no, nothing will probably ever come out that will beat it, if it is based on the same principles all current blocks are based on, but if people want to move ahead and take cooling farther, then maybe it's time to quit thinking the way everyone does when it comes to making a water block - you look at every block out there... from the start... let me describe them, easily, for you... a hunk of copper, mounted to the cpu with some thermal paste, some sort of design on the copper, whether it be channels or pins, or whatever... there have been tones of internal design ideas, a top of either acrylic or brass, or copper, or aluminum... and some people have incorporated nozzles and all sorts of things... that is the basics of water blocks right from the start - well, I am not following any of that for the most part - instead of starting with a block and seeing how it can be made better, I am starting from the point of a hot CPU and how can I cool it using water/coolant, and how can I do it the very best way.
The fact is, I have complete competence in what I have designed and know what it does and is capable of - but at the same time, I am not an idiot and will just put the whole design out there so it can be scooped up by someone else - i have hundreds of hours invested, and lots of money - I am not about to just give that all away... I hope you can understand that.
I am not some kid-n00b, like posts in the extremesystems forum thread have called me - I am 41, have a wife and kids and make a good living outside of water cooling - water cooling has been a hobby of mine for years, and all I am doing is taking an idea I have and making it become reality... the block, once done, will stand on it's own, or will be a complete flop - and it will do that by being tested and reviewed by the sites out there that have the proper equipment to test blocks... not tested by being put on a system and temps taking... but where all variables are controlled and it is just the performance of the block being tested - where they keep the coolant the same temp coming into the block and can vary the heat under the block... where it is not dependant on air temps, changing coolant temps, rad sizes... that sort of thing... and there are a couple places out there that have a test bed that can do this exact type of testing, and they are the people who can test and review a block, and give repeatable results - those are the ones I will be happy to send to for testing - that way, when the results are back... it will have nothing to do with me, or with how it was installed, or with rad size... it will make it or die based on it's actual performance.
I am keeping any posts and updates to the Xoxide Forums... they are the forums I belong to and the place I have chosen to post anything to do with this - all of you are welcome to join and post, whether good or bad... but please, keep the slamming/bashing out of the forum - I do not mind criticism at all, but posts just to bash are a waste of time for everyone.
I do not expect people to believe... especially those who have not seen the inside design, which I have kept private for the most part, but I will say this again... if you quit thinking about how blocks are made right now, you may see that there are other ways to cool a CPU, and then may be able to see that what I have said is not only possible, but pretty easy to accomplish.
Devin Sittler
Quote:
Originally Posted by oshox
Not so much actually. Jet impengment works due to the pressure against the surface area. If you have low flow, low pressure and tons of surface area you've got a big ass deisel motor (just begging for propane).. if you have low surface area, high pressure, and low-high flow rates you've got the equivalent of a NOS fed 4-banger (it's going to perform, but you cant feed it much more) :D
Things like chemical or physical etching can be done..
Heck, the direct die idea I mentioned above would actually work ok with a pelt maybe....
If someone would tell the block designer (Devin Sittler) to mail a copy of the documents to himself he will have free copyrighting of his idea and no one can steal it because he will have a legal document dated and allowable in court so that he can prove it is his idea.
most likely he's waiting to finish his block, test it, have some unbias and precise testers have a go at it, then if its a good deal, copyright it and sell for a few...
the problem is how do you take on a company such as Swiftech if they were to copy this.....he's a one man operation and is talking about $1500 he invested from his own money as if it's a big amount....i don't think he's got time or money to chase around copycats until it's properly patented.....but you'd only want to patent something if you know it's worthwhile.......for now he thinks it is and we don't as these results seem unbelievableQuote:
Originally Posted by phelan1777
Results can be predicted using modelling (CFD) before going anywhere near final production / prototype. Whether any modelling has been done, I dunno... not seen any meself.
Quote:
Originally Posted by dinos22
what I am saying is that his copyright would only cost 37c and the price of the envelope and when someone tries to copy his idea he could take it to court.
Which still places the financial burden on him and more than one good idea has been swiped due to lack of funds.Quote:
Originally Posted by phelan1777
I wish North had just started all this by saying something like "Hey!, I've had a real cool idea and initial results seem to indicate that it works quite well.
I'll keep you posted as progress is made."
Without the direct comparison to the Storm all of the flaming could have been avoided.
If it did do as well as he claims- and he had testing to prove it- then he could have blown his horn far and wide without being questioned.
On the other hand he's getting a lot more attention now than if he wouldn't have mentioned storm. Now hes got everyones attention so if he comes up with something good he won't have a problem selling it.
I hope that he comes up with something good, but based on what he's showed so far I doubt it.
Ive been thinking this through aswell...quite close to the ideas of j. as aboveQuote:
Originally Posted by andersson.j
If there is a copper base that has been impinged,then had liquid gallium,the stuff that's better than AS5 but corrodes Al. is poured into the base and impingements(on the underside) then another piece of copper base is press fitted onto that,obviously machined and then lapped after,that would definately provide better heat transfer.
the jet idea i had was similar too but, well,have you guys got an electric toothbrush?The vibrating mech thats inside that that can be modded in such a way inside a top delrin piece of a water block with all the neccessary seals and of course the ability to move from side to side on top of a jet assembly and would 'pulse' jets of water onto the base from the above.The electric for it can be provided simply from the cpu fan header.
I think you know what i mean. :)
Pulse cooling is used effectively by the fire services and is a proven method of extinguishing fires more efficiently.
Most of you will realise that whenever there is vibration,nuts and bolts do undo themselves,this is simply remedied by spring washers/springs/nyloc nuts where needed.
Phew ....
However, my moneys on the 'no base design' .
I'm surprised no one over there has asked: "what material is the base?"
Given North's repeated assertion that the target price is less than the Storm you can bet that no exotic materials/machining will be involved.
and maybe he should have gone with a simple title and not that 10c better than storm or close/near/surpassing ambient temps with his new block.
and with that line he should have finished it before going online with his idea.
but if it works then good for him, if it flops then it'll suck for him and tarnish his rep.
a 10 celcius difference under NORMAL conditions can simply not be true.
Of course, 10 celcius delta dosn't tell us anything unless we also know the delta between core temp and water temp. naturally the larger the detla between water and core, the bigger the difference you will be able to measure between a good block and a bad one. The heat output over time also matters.
The fact of the matter is that (assuming a normal system where water is rad cooled by ambient air) there simply isn't much left to improve on a block once you have a "simple" pinmatrix in place. Sure, you can allways tweak it, but the differences are going to be small.
Evne if the water in the loop had direct contact with the exposed core, I think a 10celcius difference compared to a storm would be a lot to hope for (again, given normal condition, for example going from a core temp 40 to 30 in a room with 20celcius ambient). its just not gonna happen.
-Stigma