Central cooling

After reading about a guy who set up his farm in the garage and used a couple of car radiators for cooling I thought I would try something similar but for me it will be with less rigs......

My expectation for this is that I will not have to buy any more rads or pumps etc for any future rigs just what cooling blocks and barbs are necessary for each one (and maybe some hose). I further expect that for most of the year I can run passive but there will be a fan (a serious fan) for high summer use

This log will take a while cos I don't do fast :) ...I think I have mostly everything I need now so here goes...

I spent an hour yesterday cutting and folding a sheet of 1.5mm galvanised steel
to form the sides of the fan/rad box...I'm trying to use secondhand where I can but this had to be new :shrug: can't win them all

There will be those among you who are perfectionists...well, look away now....Lets just call this a proof of concept...no use of straight edges no fine measurements just rough and ready cos Heath Robinson is my hero

In the first pic we have Today's efforts setting up a unistrut frame (bolted to the floor and the wall) then fitting the fan box with a channel to hold the rad. In the channel I have placed closed cell foam to protect against possible vibration issues



Second pic is the box finished off (mostly) with the rad fitted



If the weather holds I will add more tomorrow...and a bit more detail too

Hey Grumpy this looks not to bad to be honest, so this is the outside of the house where the rad is going to be?

awesome...can't wait to see more. I love these "projects"!

Oh this is cool! grumpy you know that i have been waiting for this thread with much curiosity to see the full design and the final results. I realize that due to the capacity of the rad and the climate that most or at least a good portion of the year this setup may work passively.

I am curios though about the warm summer nights and the noise level during the night? Also what about the cool winter nights when you reach freezing temps, what will you use for coolant as you want it to disipate heat easily but also not freeze.

This is a really cool idea!

Great pictures Grumpy. Where is the fan going to be? House side pushing or outside pulling? Glad to see you are making progress.

Because I want to use the parts that I have on hand where I can I elected to run this using 22mm (~3/4") copper pipe. The rad had 2 1/2" inlets so here in the first pic is reducing from 2 1/2" to 2 1/4" to 22mm



Now I would like to introduce you to the fan specs ... 63cm, 650rpm, 240v, 0.35Kw and fitted to the rad probably in excess of 2400 cfm



Ladies and Gentlemen I give you the MOAF

note the yate loon 120mm attached for scale purposes....Is this the Mother of all Fans???


It has just started to rain so I will start to assemble the pump and associated pipework. inside ...more soon.

fantastic how it is all coming together Grumpy!
I would leave the 120mm fan in place as it looks great!

put the 120 on the inside of the fat ass fan to get rid of the dead spot LMAO

Hey grumpy, how is all the plumbing inside the house coming along?
have you got any of the machines up and running yet?

Odd that you should ask today..... For the first time in a week or more I got to do some more today .... Most of the hot side pipework is now in and the Fan and controller are good (yet to run it though) just need a thermostat and a few fittings to finish up at least to run everything in a flushing cycle....sometime this week

We will be looking forward to the updates!

Considering the problems Savage is having priming his system, Grumpy is going to have a real ball.... :D:

Oh lets hope not!!!

Wow...that is looking great! I think you will be fine priming up that system...you just might need a diesel powered air compressor

Someone asked for photos before this is all finished so here are just a couple to give you all a feel for iteration 2.2 (just crappy cellphone camera I'm afraid)

parallel loops pic. 2 spare loops on the end



Fan controller, pump pressure guage and big 'ol return pipe (that's 2 1/2" copper :Kewl: )



Return leg of parallel loops joining return pipe (which doubles as a res)



using EK res as a sight glass for water level



And, under the bench.....The pump.




This will all work quite well as is but is on the limits of the pump...in that I need to add 2 more blocks to see how it performs then.... I might have to get another in series. It will lend itself quite well to some experimentation with series versus parallel performance.

Need a few more taps first. just now I waiting on a flow gauge to arrive so the fun can start.

That is impressive grumpy...thank you for sharing. That looks outstanding and appears to function well!

Man that is quite the undertaking Mike nice work M8.

Wow, nice work Mike. Hope it works as planned.

Savage, do you think he will have a problem shaking and rotating the system to prime it... :HAHA:

bob, you bastard

Mike that has got to be the coolest computer cooling setup I have ever seen! Can't wait to see the temps when you start testing...

Thanks for sharing! :

Most excellent Grumpy! :Cool:

SavageNFS wrote:bob, you bastard

Hee, hee, hee.....

Nice burn LMFAO

Thanks for the comments guys

The original Idea was to have a setup where any new rig only required a cpu block and maybe a chipset fan to get going but as each iteration of this build has come and gone I find that by spending on a few more bits and some more 1/2" clear tubing I should be able to experiment with parallel, series and series/parallel setups.

When the flowmeter gets here I can see how much flow I am getting as things are now and, by closing taps, get the results for each individual leg or combinations thereof. I have a feeling that I am below 1 gall/min for each block mostly due to the fact that the pipework and rad seem to need a little over 2m head but that said, that figure is with the flowrate up around the 9 galls a minute mark.

Trying to do the calculations based just on pressure and a pressure/flow graph just does not cut it. Hence the purchase of a flow meter

So, as the build and the experiments progress I will add stuff here, probably starting with the final base setup next. I may even move all the power outlets for that. :D:

Some time before next winter I will be installing a digital fan thermostat too so that I can get results on how fast this heats up. It may then be the case that I add a 45gall drum of coolant to the circuit as a means of running as nearly fanless as possible.

Results so far show that with around 4galls circulating and the fan off it takes around 1 hour for the coolant to raise by 10 degrees (ambient around 15) so you can see that moving to 48 galls of coolant we could be talking 12 hours with the fan not running and, in view of the fact that when the fan does run the 4 galls of coolant drops 9 degrees in 2 mins the fan cycle may only be less than an hour per day.

I guess it is fair to say that my approach is not that usually associated with watercooling threads but hey, rough science is good science too right?

EDIT: bob, the reason for the iterations to date is just that.... I could not get it to self bleed the way I had it to start with. The return/reservoir pipe works a treat though.... but man have you seen the price of copper lately?

Great thread Mike. Yes I have seen the price of copper go up in wiring so I can imagine what it would be for pipe. On this side of the pond there have been much use of PVC instead of the copper. A friend of mine is a plumber and he has gone to 100% PVC. He has a tool to make up all the connections, no more looking for the correct fitting. He says it is great.