The Sound of Silence - Everything Water cooled - Update Feb 23

Test fitting the CPU block

Time for a little update.

As I am planning a large Fan-Duct from the side panel to the fan/radiator, there is very little space above the CPU.

So I made an angled tube fitting. It is a bit big, but I works just like I want it to.





I also fitted a DangerDen FillerPort into the top of the case.

 

Test fitting the Splitter block

I have mounted these alu parts to hold the splitter block behind the Floppy drive, and the bottom 5.25" drive space.



As it turned out, the tubes from the PSU needs to go up, or there would not have been room for the Fan-Duct.



Some tubes access the splitter block from the top, others from the buttom

 

looking good alvin, not sure you need to water cool your ram though. Unless its overclocked it genrates nominal amounts oh heat (my crucial pc3200 DDR 512mb) is cool to the touch.

 

Hi alvin

Looking good. have you tested the blocks/splitter to see if they leak at all? if not you could have one hell of a puddle to clean up!

hopefully that splitter works nicely for you.

will be nice to see if the single rad can handle all the heat for you....

take care bud...

 

Hi Alvin,

Thats looking great. You have been very inventive

You dont happen to work at Studiedata do you?

keep up the good work

 

Thanks. You know what they say: "Necessity is the mother of invention". And I simply NEED for this system to be very, very quiet..... and not cost too much to set up.

lol, no I don't. But I do shop there, as I find them to be the best stocked PC modding supplier in Denmark, and their shipping has so far been impeccable.

I'll probably wait with Ram cooling. But AutoExec might show us in his next update how he has set up RAM cooling.

Until late last night I had only tested the CPU add-on block.
It turns out that the Splitter seal system is quite a b.... to try to get to seal up propperly.
It is simply too hard to get propper access to it once it is mounted inside the pc.
So I have decided to alter it, and glue in plastic stubs in place of the otherwise fancy o-ring system. As every component with the watersystem MUST be absolutely sturdy and reliable.

Yes it will. By now I am quite impatient about getting some test results for the different components.

Unfortunately the splitter block failed its first lab test. So back to the drawing board on that one.

 

The Splitter Block - Take 2

The first test of the splitter Block, using O-rings, did not end up as I had hoped.
It may have worked out if I had just continued to worked on it, but I want something that is absolutely dependable and that will continue to work without leaks for the next couple of years.
And though I liked the O-ring idea a lot, it just took too much fiddling and twinkering, and there was just too little space to allow all my "ten thumbs" to work in there.

So I implemented plan B for the Splitter block. And glued in plastic fittings. Here is a few pictures of what it looks like right now.

From the Top



From the Buttom

The layers of the splitter block are 5mm thick each. Two layers for the top, and thus 10mm of glue thickness to hold the plastic fittings in place.
Three Layers for the two compartments for supply and return, and then again two layers for the buttom which has the buttom fittings glued in.

Overview


This design certainly does have larger and more conventional seal surfaces. Lets hope it works now. I will probably be able to let you know in a few hours.


I managed to find these small clamping straps in a garden section of one of the local hardware stores.


They are soooo small, and yet soo effective. And a perfect fit for my 10mm OD (Outer Dimension) tubes. I just love these little wonders.

See you all a bit later.

 

Wow this is a hardcore build. I wish you gr8 luck. Keep us posted.

 

what does that junction block do for your flow?

i imagine it dramaticly decreases your flow?

 

Thanks. I will make sure to keep you posted.

Well, liquid flow is a funny thing.
Among other things liquid flow depends on pressure, volume of the pipes, resistance in the flow, temperature, internal movement within the liquid and the liquids own surface resistance.
Read on in my next post for a few thoughts about liquid flow, and my splitter/junction block.

 

Notes about flow, pressure and the Splitter/Junction block

I am no expert, only an amature trying to use some common sense, some very basic math, and some observations I have made from places where liquid flow has been involved.
So far I have just juggled the principles in my head, but I think I'll use this oppertunity to jot some of it down. Mainly for my own bennefit, but maybe you will have some comment.

The main principle I have used is a physical constant that say:

Konstant = Pressure x Volume x Temp​

The next "logical" principle is:

A chain is only as strong as the weakest link
Or in terms of water flow:
The max flow through a loop will be determined by the smallest passage in the loop
Edit - Or yet more correctly (as has been pointed out later)
The max flow through a loop will be determined by the hardest to pass passage in the loop
/Edit
​

My largest tube is 1/2" ~ 12mm. If at some point I were to try to push this through a pipe half the size, I would cut the water flow of the entire loop to one half of what the 1/2" tube is capable of.

Now the amount of water that is able to run through a pipe, is determined by the area of the cross section, meaning the area size of a circle, or the area size of a square, or whatever shape the tube/cylinder we want the water to flow through.
So to compare the Water Flow ability of two tubes one would have to know the area size of the cross section, not only the diameter.

Here I have calculated the area of some of the tubes/pipes I use.


So going from a 12mm ID tube down to an 8mm ID tube, I would effectively cut the flow in half.
Going from 12mm ID to 6mm ID would cut the flow to only one fourth of the capacity of the 12mm tube. (113/28 =~ 4)

Thus the main job of the Splitter/Junction block is to help maximize the flow through the entire system, by trying to allow the maximum flow through the pump and radiator.

The following illustration has two different loops. One in parallel, and one in seriel.

Notice how in both loops 6L/min is running through the CPU block, as this is the maximum this pump can allow through the size of the red tubes.

I have made a rough cross section calculation of my splitter/junction block.

It is only a rough calculation because it does not take into account that all the fittings are smaller than the tubes them selves.
Also it does not take into account that the flow through the splitter/junction block itself is easier for some of the connections that other, simply because of the square internal shape of the compartments within the block, and the fact that some of the tubes are connected at the top and some are connected at the buttom of the block.

What the rough calculation how ever DOES show, is that I MIGHT have cut the flow through the CPU block to half of the maximum that this pump is capable of.
So I have taken meassures to help the water to want to "choose" the CPU loop rather than any of the other loops.
The CPU connections are right opposite the pump/radiator connections, so the internal flow in the splitter/junction block has NO obstruction for the CPU loop. Also I have drilled the Inner Diameter of the CPU fitting a little larger than the standard size of the rest of the connector fittings.

So all in all, I am still hoping that things will turn out okay.

 

yer it just looks like it might be quite restricting for flow, look forward to your next update though

*edit* LOL during the time i was writing this post you posted your update

*edit2* wow that's some intresting formulas there. just read through it now good stuff, good luck with the rest of the project

 

It dosent quite work like that im afrade, 1m of 1/2tube will have less restriction than 2m's

This project is relly amusing

slater..

 

Yes, the length of the tube also has an influence, as - for one thing - it has a larger surface that touches the water/liquid. Any bends in the tube will also influence the flow.
Liquid really IS a funny thing, and many, many things will influence its flow.

My brother just told me yesterday, that for the first time ever there is now a mathematical model for the flows through the heart. The Physics department at RUC (the university where he is studying = Roskilde University Center) has developed a model.

Anyway, I just wanted to introduce ONE of the principles of what I have guestimated about, when I made the Splitter/Junction Block.

It should be obvious by now, that I think some, guess - a lot - and then let a lab test show me if I am somewhat right. And so long as I am within a functional interval, I am quite happy with the practical implementation.
I am just cooling a small computer anyway! Not sending someone to Mars.

Its all for my own use, fun and entertainment (and apparently your entertainment as well).

 

That sure is hardcore cooling Alvin!

Very nice and I just love the splitter you have made - give me ideas

Subscribed!

 

just curious are u posting this log on ne other forums cuz im really interested but i havnt been able to see a pic since day one and they still dont wont to work for me just curious to see if it show up on anouther forum or nething?? thnx

 

Thank you. Always glad to give some inspiration back to all you guys on my favorite Modding site: Bit-tech.net

Sorry man, I don't. But I'll see if I can load the pictures on a different server one of these days. Until then, please remain patient.

 

your pc isn't that powerfull so aslong as there is a flow you shoulden't have temp problems.... Also if you ahve problems with the gpu loop you could always stick a huge heatsink on it instead

 

Water Loop details - Final test assembly

The rough Flow calculations I made in post #91, made me take a second look at the fittings I use in the Pump/Radiator Loop.
The Blue plastic fittings had a 7,5 mm ID, and the Radiator metal fitting only had a 6,5 mm ID. Not satisfactory meassurements!

I have now drilled all these fittings to have a 9 mm ID.

Here I have already done one of the fittings on the Radiator. The other fitting (top one in the picture) is halfway done. You may be able to see how much of the wall I am removing.

I would not necassary recommend others to drill out the fittings while they are mounted in the Radiator. A LOT of little metal pieces will fall into the radiator. I had to rinse out the base of the radiator at least 15 times, before the rattle sound of metal was gone from the radiator.
I am however getting very impatient about getting more test results, and moving on with this mod, so sometimes I take short cuts, in an attempt to save time.

Fortunately, I am now FINALLY ready to make a water flow / water leak test.

The Pump vibrates, so I have suspended it in a blue elastic strap.



I did this test without the HD assembly, as the HD box is a very tight fit, and is quite hard to get in and out for these tests.
I have to switch the two HDs anyway, as the ones I need in my case, they are currently mounted in the temp PC I am using while modding the case.

I am HAPPY to Report: NO LEAKS and GREAT FLOW in every loop.

I still have not installed/made a reservoire, so I get A LOT of bubbels when I fill the system with water.
These bubbles are noisy, and I want them Out Of There, but at least they give a good indication of the flow in the different loops.

I feels great to have reached this milestone, so I'll just take a note of the two-three tings I still need to do to this case.

What TO DO next:
- add temperature sensors
- make PWM Fan controller
- design MCU cards to Monitor and Control the PC
- add VU meters in the sides of the front plate
- replace PVC tubes to/from Pump with rubber tubes, to eliminate pump vibration transfere to the case
- add passive radiator/convector
- maybe add wheels to the case
- maybe add some lighting to inside of the case
- make fanduct for the side of the case
- make/add reservoire

- SILENCE wifes PC

Oh well, a least I know I wont be bored for the next couple of days (read: weeks)

See you all next week, as the kids are comming over for a four-five day holliday shortly.

 

If you want some more info on really good advice on watercooling try procooling.com

They've been pioneering watercooling since the idea.