NO, It didn't work, what next?

[redleader]

University surplus auctions are another great place for this sort of thing. Everthing from chillers to cryogentic freezers get auctioned when labs upgrade. Its a great way to pickup parts cheap, if not whole systems.

[BillA]

the very best,
people make a living off of selling this kind of surplus
problem is ya got to be there

[bobkoure]

Originally Posted by unregistered

http://cgi.ebay.com/ws/eBayISAPI.dll...ME:B:EOAB:US:6

And the billadams111 who got this one for $9.99 would be someone we know?
congrats - looks like a nice piece of gear for cheap!

[BillA]

it is a beauty !
I have 'added' it to the environmental chamber as it is a pass through design (air inlet/outlet same side)
with this and another mod I can now handle 600W ! (needed to test big rads)

[MaxxxRacer]

Man that is a hell of a find Bill.. I take it your going to make a waterchiller setup soon?

[Jackyl]

If there were ever a time and place that I would say something like "you guys are TOO HARDCORE!" This would be that time and place.

I'm not going to be one to say that you should do this or that, nor would I expect someone to say that to me. But jesus, when you start to throw in flowrates of 1/2 vs 3/8 hose and pumps and other things of that nature ya know it really takes the fun away from it.

And from what I have seen so far for parallel vs series cooling well it's a space issue most of the time and other times people do not want to spend the extra time on the extra needed hardware.

Oh and for flow rate. if you happen to see the pump that I'm using well yeah, it's something like 3,000 gph. It's a pond pump.

And someone said something about what is the thermal load differance when the water goes from block to block in series, well it depends on the flow rate/inside hose diameter/ block restriction. So in all reality the key would be to figure out how long to keep the water in the block before moving it onto the next block without restricting flow and carrying exessive heat from block to block. Yet making sure that the liquid can carry the heat fast enough to not hold it in the block and cause issues. So can you have to much of a flow rate? or no matter how fast your flow is it will still remove the heat.

[Butcher]

The only issue with higher flow rate is extra heat dumped in the loop by the huge pump you're using to achieve that flowrate. Flowing water faster through the blocks/rad will improve performance. The whole water has to stick around to collect the heat thing is a myth.
In light of that, series will give you higher flow through your blocks which improves performance over parallel.

3,000 gph pump says nothing about it really - need max head as well to gain any real feel for performance.

Where's the fun in having a badly designed system that doesn't work? Better to get some real knowledge as make something decent.

[pdf27]

Originally Posted by Jackyl

Oh and for flow rate. if you happen to see the pump that I'm using well yeah, it's something like 3,000 gph. It's a pond pump.

So what? That will be against zero head. What counts is how much water it can push through your system (more than a few GPM doesn't gain you very much at all), and how much heat it dumps in your loop (at a guess, I'd say that it is probably dumping more heat than your CPU is into the loop right now). If so, a smaller, lower flow pump will give you better temperatures.

Originally Posted by Jackyl

And someone said something about what is the thermal load differance when the water goes from block to block in series, well it depends on the flow rate/inside hose diameter/ block restriction.

No it doesn't - it depends only on the working fluid, mass flow rate and heat dumped to the loop by the blocks. Throwing extraneous factors in like that to try and seem clever just shows you up as a complete numpty.

Originally Posted by Jackyl

So in all reality the key would be to figure out how long to keep the water in the block before moving it onto the next block without restricting flow and carrying exessive heat from block to block. Yet making sure that the liquid can carry the heat fast enough to not hold it in the block and cause issues. So can you have to much of a flow rate? or no matter how fast your flow is it will still remove the heat.

We've had this arguament before. It's utterly pointless, and a full thread discussing it can be found in 8-ball's sig. The mass flow rate through a block does not make a blind bit of difference to the average water temperature in a closed loop system - the difference in CPU temperatures is caused by the improved heat transfer between the block and the water alone.

Originally Posted by butcher

Where's the fun in having a badly designed system that doesn't work? Better to get some real knowledge as make something decent.

At a guess, so he can come on here and pose with his "amazing system" while making himself look a complete fool with his total lack of understanding of even the most basic thermodynamics.

[JamesAvery22]

hehe dont think this guy is gonna listen to anyone. Sounds like a guy who needs to see real world results, something right infront of his face. Which isnt really a bad thing. I tend not to believe people I dont know off the bat either, so to each their own. Just sucks for the stubborn ones as they'll never learn.

[pdf27]

Problem is, he's actually seen real world results on his rig (and they're dire) but he won't listen to advice from loads of people who've seen the problem before and is convinced he's right. What I reckon he really needs to see is results from a well-designed rig make his look really, really dire (which they are). Then he might start to get the message.

[redleader]

Originally Posted by Jackyl

If there were ever a time and place that I would say something like "you guys are TOO HARDCORE!" This would be that time and place.

I'm not going to be one to say that you should do this or that, nor would I expect someone to say that to me. But jesus, when you start to throw in flowrates of 1/2 vs 3/8 hose and pumps and other things of that nature ya know it really takes the fun away from it.

And from what I have seen so far for parallel vs series cooling well it's a space issue most of the time and other times people do not want to spend the extra time on the extra needed hardware.

Oh and for flow rate. if you happen to see the pump that I'm using well yeah, it's something like 3,000 gph. It's a pond pump.

And someone said something about what is the thermal load differance when the water goes from block to block in series, well it depends on the flow rate/inside hose diameter/ block restriction. So in all reality the key would be to figure out how long to keep the water in the block before moving it onto the next block without restricting flow and carrying exessive heat from block to block. Yet making sure that the liquid can carry the heat fast enough to not hold it in the block and cause issues. So can you have to much of a flow rate? or no matter how fast your flow is it will still remove the heat.

I'm not trying to be a dick, but you really need to do some reading. You don't seem to have a grasp of the basics here, which is a real problem given your interest in advanced stuff. Do some reading, work on getting an understanding of what effects performance and why, and then resume your projects. I think you'll understand why your chiller didn't work, and much more with a little effort.

[killernoodle]

That chiller would be better off attached to a good waterblock; screw that manifold and the parallel lines. Get rid of the resevior because that isnt helping anything, only dissipating whatever decrease in temperature you are getting.

Even better, bolt the cooling head to a huge heatsink and drop it in the resevior. I'd go for one of those swiftech mcx478-v and just plunk that in your resevior. That would be far better than the way you are going. Then, insulate everything and run everything inline. That should be a huge benefit.

[JamesAvery22]

Originally Posted by redleader

I'm not trying to be a dick, but you really need to do some reading. You don't seem to have a grasp of the basics here, which is a real problem given your interest in advanced stuff. Do some reading, work on getting an understanding of what effects performance and why, and then resume your projects. I think you'll understand why your chiller didn't work, and much more with a little effort.

Heres a good place to start.

http://www.overclockers.com/articles1088/

[Jackyl]

Hey pdf27 thaks for being so harsh, I mean I don't spend days on generating synthectic numbers in a synthetic enviroment just to come on here and smash on people that are asking for some help.

Also, I know why my vapochill project failed A) the vapochill is unable to disapate they amount of thermal load all the components were putting out combied. B) The evaporator head had too small of a contact area on the manifold.

I'm currently "after reading alot of information on this site and many, many, many others!" coming to the conclusing that I'm going to go with 2x TT Big Rads "39.99 comes with a 120fan" TT NB-Cl-W0010 P4 block as for the cost and performance from reviews it's a worth purchase. Danger Den NV-68. And I'm either going to use my big res that I bought from overclockers hideout many years ago or just use a danger den fillport mounted on the top of the case to fill the system.

I'm going to give series a try using 3/8 line and using my existing pump with a rheostat to throttle the flow to document the difference voltate/flow has on the system with this pump. if it is over kill. or adds to much heat to the loop I'll ditch it and buy a smaller unit.


pump---->y-splitter--->Cpu Block--->Radiator---->Yspliter-->pump
--->Gpu block--->Radiator----------------->pump

I'm not going to water cool the northbride as it is absolutly pointless given I can do a 320fsb and get into windows it's the 2.4c temps that are holding the system back from being stable.

I'm going to see if I can do what alot of you have said already, run synthetically and see how the numbers turn out.

[JamesAvery22]

Originally Posted by Jackyl

Hey pdf27 thaks for being so harsh, I mean I don't spend days on generating synthectic numbers in a synthetic enviroment just to come on here and smash on people that are asking for some help.

If you refuse to listen to those who obviously know more about a subject than you do, then dont expect a warm welcome.

Originally Posted by Jackyl

Also, I know why my vapochill project failed A) the vapochill is unable to disapate they amount of thermal load all the components were putting out combied. B) The evaporator head had too small of a contact area on the manifold.

If you realize that the evaporator head wasnt making good contact with your manifold, hence not cooling efficiently what so ever, why do you think it cant handle the load when you just said it never had a chance to properly cool it??

Lose the parrallel crap loop, and read up on immersion chillers. And, IMHO, turning a direct die chiller into an immersion chiller is a giant waste.

A vapochill can handle the heat load, just wont give you below freezing temps. The BTU load of a vapochill is around 1000 (please someone correct me if Im wrong). The rating on window A/C units, the ones that people (who know what they are doing) use to make chillers, are around 5000 btu.

Originally Posted by Jackyl

...
I'm going to give series a try using 3/8 line and using my existing pump with a rheostat to throttle the flow to document the difference voltate/flow has on the system with this pump. if it is over kill. or adds to much heat to the loop I'll ditch it and buy a smaller unit. ....

#1 The pump in your pic looks like mag3 or higher. 3/8 tubing is a horrible idea. Will restrict the input of the pump way too much.

#2, its an AC pump, you cant "throttle" AC with a resistor.


You really need to read some basics about pretty much everything mentioned in this thread. Don't start off with modding the heck out of a vapochill unit and barking random things on a highend VERY technical forum. Sounds like you should go to some place with a lot of good beginning reading. heres a good place for you to spend a few days:
http://www.ocforums.com/forumdisplay.php?f=127

The block and the radiators you are looking at are pretty bad choices. Im sure they will work to an extent, but there are cheaper and better solutions out there. Not to mention easier to find. Im done spoon feeding for the day so read those stickies and search some forums if you dont understand why.

[pdf27]

Originally Posted by Jackyl

Hey pdf27 thaks for being so harsh, I mean I don't spend days on generating synthectic numbers in a synthetic enviroment just to come on here and smash on people that are asking for some help.

You weren't asking for help, you were telling a lot of people with massively more knowledge and experience than you that they were wrong and you were right. You then started posting a lot of garbage that we see here regularly. All things considered, I'm quite impressed with the restraint shown around here that everyone was polite and helpful for all of the first page and much of the second despite your attitude.
The harshness was a deliberate attempt to get you to face reality, and it does seem to have worked at least to some extent. You're still barking up the wrong tree, but at least you're in the right forest which seems to be a start.

As mentioned before, this is a very technical forum. I've got an MEng in fluid mechanics/thermodynamics and am among the least qualified people on the forum, not to mention the VAST amount of experience someone like BillA has to back up their theoretical knowledge. If you start coming in here spouting rubbish and claiming you know better than us (like you did) you're really asking for trouble.

Originally Posted by JamesAvery22

#2, its an AC pump, you cant "throttle" AC with a resistor.

Not sure about pumps (never tried it), but I've had excellent results running two Sunon AC fans in series across 240v AC to undervolt them. Ferdb will be able to give you more details, but they are the best silent fans I've ever come across for watercooling purposes.

[Butcher]

Vapos are good for about 200W (~700 BTU/hours) James, so you were in the right ballpark.

[BillA]

Originally Posted by MaxxxRacer

Man that is a hell of a find Bill.. I take it your going to make a waterchiller setup soon?

I assume you're referring to the Haake C11, yea
I have 2 Haake A82s but wish to free up one of them by using the C11 with a variable speed pump
not too sure that I'll be able to control the output that way as the nameplate calls for -40°F freezing point on coolant
may have to just vary the fans' speed - using it to chill air in an enclosure

[redleader]

Not sure about pumps (never tried it), but I've had excellent results running two Sunon AC fans in series across 240v AC to undervolt them. Ferdb will be able to give you more details, but they are the best silent fans I've ever come across for watercooling purposes.

AC motors have the rpm fixed by the frequency of the voltage. So you can't change that with a resistor. However by limiting voltage or current, or inducing a phase angle, you can make the slip a little off some cycles. This limits torque, so it will cause lower speed or the thing to not work. I'd imagine you could also damage things, though I've never seen it happen. Which exactly depends on the motor and how its designed.

[redleader]

I'm going to give series a try using 3/8 line and using my existing pump with a rheostat to throttle the flow to document the difference voltate/flow has on the system with this pump. if it is over kill. or adds to much heat to the loop I'll ditch it and buy a smaller unit.


pump---->y-splitter--->Cpu Block--->Radiator---->Yspliter-->pump
--->Gpu block--->Radiator----------------->pump

Why try it? You know roughly what the pump can do, take a conservative guess on what the flow will be. Then plug in the thermal load, and the specific heat of water (4.186 j/g). Take a look at the difference you're dealing with, and then make an informed decision.

All measureing it is going to tell you is how much your meters suck

[Jackyl]

pdf27 sorry for ruffling your feathers, But can you show me some pics of this well designed system that your refering to? I would like to see something that is superior to what I have built. I'm not saying that not such a thing exists but rather to see with my own eyes what you are talking about. I've seen many seires systems and well that was the basis of my build, it's not another thumb, sure it failed but failed in what term? I tell you this it didn't fail in the fact that it is the most complex water setup that I have done yet. Nor, did it fail in testing out somethings that I have been wanting to do. But it did fail in disipating the heat that all the componets that I was trying to cool generated. So I dismantled the unit, sold the vapochill and the chassis and the blocks and kept the pump and res and manifold and the documentation that I had generated. Now will I go series, this time probably but always in the back of my mind I'll be thinking about a way to disipate that thermal load in a parallel setup. That is my big beef peopl always want to keep doing things that ways that work rather than inovate or invent knew ways that are generally more complex but have the same goal to make the way that works more simplistic than it currently is. And because of this I come off like not wanting any help or barking up the wrong tree or saying that people don't know what they are talking about. While infact this is not my intent, I'm asking for help, input, peoples personal evalutaions on what they have done and what they recommend. And not read the stickies or read this or read that. Actual helpfull material.


So with that said. I have a few questions.
First off: My pump is a Pondmaster 5 now as for GPM and Ft/Head? From what I have read about pumps, The pumps that are used for different applications have different ratings per application. So I'm going to assume that the Ft/Head is the term used to relate to how much flow the pump can push for a certain lenth or run of tube. So my question is: Is this pump to big? As I stated before I was going to throttle the flow rate but without doing that inline by restricting the flow I can't do it on the A/C site as only D/C can be throttled with a rheostat. I was thinking about model train controller but that is an a/c-d/c converter that is why it can throttle the d/c after it has been converterd.

Now jamesavery22, you said that the TT water block and the TT radiators are poor choices? Why is this? What would you recommend over the TT rad's a heatercore? or a permacool radiator with turbulator technolgy "oil cooler"?

I have always like swiftech and my last 2 blocks were swiftech and I'm glad that they went away from that push junk on the fittings and went back to acually post fittings.

And still reading, and reading. As now I'm not sure that series is the key, because when it really comes down to it what is the bottome line, I keep reading it in every post, Every thread, Every review. It's heat dissipation. So as long as you can dissipate the heat then is the system flawed? I would think not but rather someone might come along that can disipate the same heat but with less hardware. ahh, isn't that what this whole thing is about simplicity and acheiving the same goal, Heat disipation in the most simplistic but yeat efficient way?

[pdf27]

Originally Posted by Jackyl

pdf27 sorry for ruffling your feathers, But can you show me some pics of this well designed system that your refering to? I would like to see something that is superior to what I have built. I'm not saying that not such a thing exists but rather to see with my own eyes what you are talking about.

Best place to look is the general gallery on this site, and read through the comments in the threads. There will be a few particularly good ones out there, but I don't have the chance to sort through them and highlight examples.

Originally Posted by Jackyl

I've seen many seires systems and well that was the basis of my build, it's not another thumb, sure it failed but failed in what term? I tell you this it didn't fail in the fact that it is the most complex water setup that I have done yet. Nor, did it fail in testing out somethings that I have been wanting to do. But it did fail in disipating the heat that all the componets that I was trying to cool generated. So I dismantled the unit, sold the vapochill and the chassis and the blocks and kept the pump and res and manifold and the documentation that I had generated.

That's good. Maybe I overreacted to your early posts (if so, sorry for that) - I've been having a fairly dire week and that may have coloured how I read your posts.

Originally Posted by Jackyl

That is my big beef peopl always want to keep doing things that ways that work rather than inovate or invent knew ways that are generally more complex but have the same goal to make the way that works more simplistic than it currently is.

Partly, many things have already been tried and people know the results - you may or may not actually be innovating and not know it. Partly, with complex systems unless you have the sort of test equipment BillA has lying around you won't be able to work out where the problem is in a system.

Originally Posted by Jackyl

And because of this I come off like not wanting any help or barking up the wrong tree or saying that people don't know what they are talking about. While infact this is not my intent, I'm asking for help, input, peoples personal evalutaions on what they have done and what they recommend. And not read the stickies or read this or read that. Actual helpfull material.

Fair enough - that's a good way of going about it. You might get more help/a friendlier reception in future if you're a little more careful in how you phrase it however - just about everyone thought you were saying something you didn't intend to say.

Originally Posted by Jackyl

So with that said. I have a few questions.
First off: My pump is a Pondmaster 5 now as for GPM and Ft/Head? From what I have read about pumps, The pumps that are used for different applications have different ratings per application. So I'm going to assume that the Ft/Head is the term used to relate to how much flow the pump can push for a certain lenth or run of tube.

OK, by "Pondmaster 5" do you mean Model 5 Magnetic drive, code number 02515? If so, it seems to have pretty similar specifications in the flow regime you're likely to be running in to the Danner Mag 3, one of the more popular watercooling pumps. This is going by Phaestus' pump comparison article and the spec sheet you linked to, but should be reasonably accurate. (Pondmaster seem to be a trade name for Danner anyway).
All the water blocks, tubing, radiator and the like in your system will have some resistance to flow. This will take the form of a pressure drop between entry and exit, and is measured in terms of what vertical height of water will give the same pressure difference between the top and bottom of the column. In the case of what I think is your pump, the pressure difference it can sustain seems to be 10'6", or a bit over 3m.
What you usually have is a line on the pressure/flow rate graph showing the pressure/head characteristics for various pumps, and another line showing the same characteristics for your loop (i.e. due to tubing, blocks and the like). The point they meet at will be the operating characteristic for the pump, and using this you will be able to predict the most suitable pump for your system in terms of water pushed versus heat dumped into the system by the pump. Cathar has done a very good thread on that around here, with some very useful graphs.

Originally Posted by Jackyl

Now jamesavery22, you said that the TT water block and the TT radiators are poor choices? Why is this? What would you recommend over the TT rad's a heatercore? or a permacool radiator with turbulator technolgy "oil cooler"?

Not sure about the radiator (distinct lack of radiator testing about), but the waterblock doesn't look promising. Internally it is basically a "maze" style waterblock (i.e. water flows in, around the block through a channel and out again). What ProCooling test data there is (for the DD Maze 4) shows it to be beaten by several degrees by many more modern blocks, and my gut feeling (not backed up by any test data) is that the Maze 4 will have a higher performance than this ThermalTake block.
Link to the comparative block testing here: http://www.procooling.com/html/pro_testing.php
If you want the rationale behind my thinking the maze blocks all behave similarly, go to Overclockers Australia and read Cathar's WhiteWater development thread (there's a link in one of the stickies). It's something like 45 pages long but quite entertaining and explains why the newer "impingement" blocks perform so much better very well indeed

Originally Posted by Jackyl

As now I'm not sure that series is the key, because when it really comes down to it what is the bottome line, I keep reading it in every post, Every thread, Every review. It's heat dissipation. So as long as you can dissipate the heat then is the system flawed? I would think not but rather someone might come along that can disipate the same heat but with less hardware. ahh, isn't that what this whole thing is about simplicity and acheiving the same goal, Heat disipation in the most simplistic but yeat efficient way?

Series isn't the key - for instance if you had three replica loops, each dedicated to cooling either the CPU, GPU or Northbridge that would perform better than a single one of those loops cooling all three. The thing is that the difference isn't all that great and to do this you've trebled your cost and space requirements. If you start compromising, in most cases the series loop will win, as the performance differences are minor and the cost differences quite major in comparison.

[BillA]

you guys are now wasting your time
someone who already has the answers is not "seeking help"
he wants to debate
but he is so ignorant that such is pointless, the guy needs to go to school
but instead of reading he wants to post
'nother kiddie with a keyboard

[Butcher]

You know what they say bill - a little knowledge is a dangerous thing. Jackyl is living proof.

[JamesAvery22]

Originally Posted by pdf27

snip

Yeah he's got a Mag5. If he has a giant radiator and he's doing CPU,GPU,and a hot NB then I guess it wouldnt be too bad.


The TT rads are like those old cube bigmomma radiators, but smaller. More like a old condenser design, one long stretch of copper tubing thats winded back and forth tons of times so it has like 20 passes. Very small ID, very restrictive. Dunno what TT was thinking on that one? Guess its more like one of those German low flow systems.