NO, It didn't work, what next?

[Jackyl]

Vapo Chill Project 3rd iteration

[greenman100]

you're getting zero flow through the blocks

come on now

[BillA]

Originally Posted by greenman100

you're getting zero flow through the blocks

come on now

lol
naw, its a bit over zero

you bother to calculate your tubing losses ?

[killernoodle]

Lol, I hope there is a divider on that manifold...

LOL, just noticed how the compressor was "cooling"the water. That is the poorest implementation of that cooling head ever.

I thought that it was cooling the resevior, I was wrong.

You need MUCH MUCH MUCH more surface area on that water.

[BillA]

"The total wattage of load here is around 220watts. As you can see the cooling system does not have the effiecencey to disipate this type of load, on proving that my theory on parralel cooling is correct. especially when you introduce the efficency of the liquid going directly throught eh manifold wich is getting directly cooled by the vapochill's head."

some rather vast areas of 'non-understanding' seem apparent
I think that what was proved was that a poorly designed experiment can yield whichever conclusion(s) are desired
a lot of work for nothing

[Jackyl]

The manifold has a devider. it has a hot and cold side. On the side that feeds the block with a 1/2 hose from the pump that then goes to each block via 3/8 hose, cpu/chipset/video. and then the return is the same way. This system has more than enough flow did you happen to see the specs on the pump? The problem is that the vapochill couldn't disipate the heat for the flow rate that is in the system.

I have since sold the cpu/vid blocks and the vapochill and now am trying to figure out what the best way to implament my parallel setup in my current mid tower case, using a 80mm radiator after each block before it goes back into the manifold and gets recirculated.

The pump has 1/2" fittings
The Res has 1/2" fittings
The manifold has 1/2" main fittings and then each hot/cold fitting is 3/8"
The Water blocks have 3/8" fittings.

This system works great but it pulled to much heat from all the components and centralized it. The vapochill just didn't have enough power to disipate that heat. I was toying with the idea of putting the manifold into a cooling and then filling the cooling with ice but then that wasn't a good idea because the ice then would melt and I would have to replace it. That is why now I'm looking at 3 sperate 80mm radiators. Or one big 3 120mm radiator, the one that I was looking at said that it was able to disipate 550watts but that was @ 10c ambient.

So tell me what you guys think.

[BillA]

I don't think you know what you were doing
?s:
compressor unit capacity at its actual operating pressures
heat load being added to coolant
system flow rate and pump heat added
all the various air and coolant temps
heat (cold) losses from no insulation, calculate it
efficiency of heat exchanger (this will be the result of your calcs)

so you add the heat load and line losses and pump heat, and see where the actual compressor capability (at operating temps and pressures -> NOT nameplate) is going to get you given the efficiency of your heat exchanger

it is apparent from your commentary that you have not even attempted to estimate these things
I think you wasted your time

[Jackyl]

When you build something you never waste your time if you want to do it. You only waste your time in other people eyes.

I had a feeling that the vapochill wasn't going to work but that is all that I had at the time. As the vapochill was a 1st gen and was having a hard time cooling my 1.6a @ 3.0ghz at least that is what it seemed like and after I sold it the guy had to flush and refill the system as it wasn't get cold enough.

I wanted to work with parrallel cooling. As I see everyone running series wich makes absolutley not sense in transfering heat from one block to the next then to the next then to a radiator to disipate the heat. The only reasoning for this is space. Well I'm not going to give up on parrallel I just thought that this would be the forum to get more information on what would be is not the best radiator to use in my new venture but I see that everyone is errogant and would rather trash people instead of helping.

Like, that's a waste of time, I see you don't know what your doing. I've built many series cooling systems and they have all worked well but I knew that there was a better way of doing things. and rather than sit and sythetically calculate theoretical outcomes I would rather just do it and try it to see how everything fits and works with each other then if need be calculate.

[BillA]

but your stated 'conclusions are at odds with reality
and the inept testing and data analysis you use to 'support' your erroneous conclusions
nor are you adept at listening
nothing synthetic about what I said, all based on real data
and your lack of calculation has prevented you from understanding specifically where the cooling bottleneck in the system was, hence your erroneous conclusions

I do agree that building and testing things is interesting,
but you need much more rigor than you are applying to this subject before making pronouncements
- that is why I said you were wasting your time

[BillA]

"I just thought that this would be the forum to get more information on what would be is not the best radiator to use in my new venture but I see that everyone is errogant and would rather trash people instead of helping."

you on a different thread ?
you said: "So tell me what you guys think." (post #6)
and the thread title: "NO, It didn't work, what next?"

YOU need to get a grip on what you actually say, and to accept criticism - particularly when you ask for it
grow up

[Jackyl]

The cooling bottleneck was the vapochill and lack of evaporator head mating surface area to the manifold. I know what the bottlneck was. I also know that the system will work.

I want to go with the following components
Swiftech MCW6000-P
Danger Den NV-68 OR Innovatek Cool-Matic NV40 rev. 1.0 GPU+RAM Cooler for nVidia NV40 OR Koolance VID-NV2-L06
3x Danger Den Black Ice Micro 80mm Radiator OR Swiftech MCR80 80mm Radiator Assembly

I'll use my manifold/pump/res/chipset block. I want to setup another parrallel setup in my mid tower case, plumbing isn't the issue rather the issue being what would you guys recommend as far as the radiators and the 6800GT water block.

[Jackyl]

Yeah, I'm kinda out of it

Been posting on other forums and doing many other things. This water cooling setup is something that I have toyed with for many years. that vapochill project is 2 years old. So with my new system I wanted to go watercooled again but have yet to be able to find the components that I'm after. And on air my 2.4c will do 285fsb @ 3.4ghz but I know that it can do 320fsb if I could just get the cpu to chill out. as I was able to get into windows @ 300fsb. 267fsb seems to be my cap with my pcpower and cooling 7 year old 425watt power supply. I tried a budies 510 turbo cool sli psu and I was able to do 285fsb "3.4ghz" rock stable. So I have one on order. But I would like to go water cooled again I just can't stomach going series I don't like the idea of transfering heat from one block to the next and next then to the radiator. It dosn't seem very efficient to me, when compared to parrallel cooling.

So this why I posted. So get some input "good/bad" and to see what the best radiators and 6800 water block would be for me to get to where I want to be with this system.

and I do plan on using 5 80mm tornado fans. I'm not going water cooled for noise control. I don't care about that.

[scooterfl]

suggest doing a search on: head/flow, parallel v. series, fan/rad selection..aww crap...everything watercooling related..then ask about components

[Jackyl]

I did a search on those and well most places don't give specs on their products. So whas it the point. And what your saying is that you don't know or you have no experience with the above products.

[scooterfl]

Sorry not gonna spoon feed

[pdf27]

Originally Posted by Jackyl

3x Danger Den Black Ice Micro 80mm Radiator OR Swiftech MCR80 80mm Radiator Assembly

Which will of course be the new bottleneck in your system...

Originally Posted by Jackyl

I just can't stomach going series I don't like the idea of transfering heat from one block to the next and next then to the radiator. It dosn't seem very efficient to me, when compared to parrallel cooling.

Really straining to be polite here (seems kinda cheeky to just copy the great BillA in his responses, but that sort of reply is really deserved here), but do you actually know what the thermal capacity of water is?
I suggest you look it up (any high school physics textbook should have it in there somewhere, as will the internet). Then calculate for the heat load you will be putting into the water and expected water flow rates what the temperature difference across the water blocks will be.
Then, when you've done that, come back and explain to us why an effective drop in the water temperature of about 0.1 deg C across the graphics block is offset by the losses of around 5 deg C from using tiny radiators and massively increased cost from making the whole setup more complicated!

[redleader]

Originally Posted by Jackyl

The cooling bottleneck was the vapochill and lack of evaporator head mating surface area to the manifold. I know what the bottlneck was. I also know that the system will work.

There are literally hundreds of people running chilled liquid setups. Its pretty basic stuff. The problem with yours is that you need to make a heat exchanger, and no sticking on DD head on a piece of alumnium is not an exchanger. Its orders of magnitude too small, and most likely the cap tube is too long even if it were large enough to work. Finally, you didn't insulate it.

The easiest way to do something like this is to pick up a Walmart AC for a hundred bucks, drop its evap into a res and add insulation. From there you can work on improving that system by making a custom heat exchanger and removing the res, by tweaking the cap tube or by upgrading the gas used.

[jman1310]

If I remember my high school physics: doubling the diameter of a hose will quadruple the possible flow. Even though you are not doubling the hose diameter, the 1/2 inch hose should push a significant flow though the 3 3/8 hoses. Having said that I'm going to put on my flame retartent suit...
Using 3 80mm rads is not a good idea, as you will get better efficency out of one large rad then 3 small ones. If you are hell bent on individual rads I would go with 3 120mm BPIs (roughly similiar costs) but less flow resistance.

[Butcher]

You do remember right - flow is propoertional to the crossectional area of the tube, which is propoertional to the square of the radius (or diameter).
1/2" tubing has 16/9 times more flow than 3/8" tubing. So 2 3/8" tubes is similar to 1 1/2" - additional losses due to friction with the walls of the 3/8" tubes offset the increase diameter.

Also, why run 5 screaming loud 80mm fans, when a couple of good 120mm will push the same air for less cost and less noise?

Have to say I agree with bill - you need to listen and think more, rather than just leaping in and getting burnt.

[bobkoure]

Originally Posted by Jackyl

I just can't stomach going series I don't like the idea of transfering heat from one block to the next and next then to the radiator. It dosn't seem very efficient to me, when compared to parrallel cooling.

How much heat rise is there in the water as it goes through each block?
How much flow loss is there by using parallel flow? What does the C/W of each block look like? Is the trade-off of lower flow for slightly-cooler coolant worthwhile?
The general consensus here is that it is not - but by all means go research it yourself.

As far as radiators go, IMHO think frontal area. Three 80mm rads is 192 sq cm. If you run them in air-series, you're effectively back to 64 sq cm.

People here are being very polite. Remarkable behavior - pat yourselves on the backs for restraint...

[Marci]

Discharge phasechange unit. Cut evaporator head off phasechange unit. Replace with 1/2" OD Copper Tube, lots of it, coiled inside a bucket. Replace capillary tube on the phasechange unit with about 1metre of 0.028 ID. Replace condenser with an old BeCooling Aquacoil and 100CFM fan. Recharge with r507 or r404a. Reinsulate. Fill bucket with coolant. Run pipe out of bottom of bucket to pump inlet mounted outside and below the bucket. Run pump outlet to manifold. Run return from blocksetup back into top of bucket. Insulate the bucket. Insulate all tubing. Insulate the blocks. Measure coolant temps AFTER the blocks but before returning to the bucket. If it's lower than ambient, don't use any rads in the watercooling side of things. If it's higher than ambient, add rads to bring it back down to ambient. NOW you have a waterchiller that's worth having. A stock Vapo with less than a 1" diameter contact area with a waterblock is about as effective in the setup above as a flea trying to move a juggernaught by leaning against the back wheel.

However, then you're back to the bog standard waterchiller that everyone's been using for years.... but beware that if you ever switch it off you may have to wait 45 minutes for the whole system to repressurize before it'll start back up as the compressor is ridiculously weak in comparison to 240v compressors. Or stick a solenoid between low pressure side and high pressure side so that when unit is switched off solenoid opens and isntantly repressurizes the system.

Build the above, pipe it in parallel, then compare it to the EXACT same setup piped in series. I bet series beats ya.

Either way, nothing particularly new.... just another DIY way of doing what people have been DIY-ing for at least 8 years, all the info for which is plastered all over the net in great detail on sites like xtremesystems etc.

That's the VapoChill aspect of it all.... but seeing as you've dropped that...

Long & short of it is, using the exact same components, the flow & head loss etc inflicted by running in parallel far outweighs the benefits if compared to series. If you want the facts to back that statement up, read these forums, as it's plastered all over em in various forms in various places in various threads.

S'my opinion!

[bobkoure]

Originally Posted by Marci

Discharge phasechange unit. Cut evaporator head off phasechange unit.

What's a vapochill worth on ebay (or here in the bought+sold area)? I would guess that it would be more cost effective ($ per BTU moved) to sell the vapo and use the $ to buy a more powerful system (like the mentioned-above wallmart AC or maybe a dehumidifier). If you used a dehumidifier you might be able to get the evaporator coil into a bucket/reservoir without having to discharge / cut / solder / recharge.
Of course, I'm totally out of my depth here, never having built a chiller (erm... well other than a chiller for mixed drinks for a frat house I lived in in the early 70's - think "Animal House" but not as dirty - and no "double secret probation", many of the profs and admins came to our bashes ). This was before we had any idea that freon was bad for the ozone layer and it was cheap.
Not to mention (yet again) that Jackyl is moving to non-refrigerated water systems...

[Marci]

Depends on who's selling it. A plantpot will charge £200+ for it, and probably get it as well, but only cos another plantpot has bought it... whereas anyone who prices by performance would logically charge / expect to pay £130 to £180 for it seeing as MachII's are now popping up for £170 - £225 in bought+sold areas of other sites' forums... and GTs for £275+

But either way, can pick up a flash chiller refurbed from a bar for about £45, regas that and get -30 deg coolant.

[BillA]

ah Marci, you are far more serious than this passing dilettante
over here one can purchase chillers of many sorts on eBay quite cheaply,
but using them requires a bit more knowledge than this experimenter had

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

[Marci]

dilettante.... i like that word.... dunno, just looks as good as it sounds *shrug*

anyways... damn customs duty, VAT & import tax to hell!