Xtreme External Cooler

[CyberDruid]

Hi guys I think I finally built a LC set up worthy of posting here

Spec:
Iwaki RD 30
Mean Well S150-24
ThermoChill PA 120.3 dual shroud 6 Panaflo 86.5 CFM
Bonneville Heatercore and two POS fans

I am still working on the thermal controls for the fans (25W rheostat for now) and will rework the Bonnie with a shroud and some better fans in a bit...still in progress...

Plan to use a manifold and split the loop into three in and three out to make plumbing Case-Side easier.

Appreciate any input


NO LEAKS!


Even the G 1/4 O-Ring barbs get the treatment


Rad assembly comes out easily for service...it is suspended in the frame from well nuts. I plane to replace the crap poly fitings the rad ships with and use real ones...I wish ThermoChill would just charge the extra $10 and ship real fittings.




NO LEAKS! And no Elbows or restricitons either



Plumbed



The suspension system keeps the rad from moving very much but isolates any vibration.



The lower compartment is laid out to allow good airflow... the Bonnie is underslung and supported on foam strips in aluminumn angle--the fans push air out so dust should not be that big an issue. I will be fabricating a proper shroud for the heatercore in a bit.


The reservoir comes off and allows for service of the fans/Filters...






The size of the cordless drill might put this beast into perspective...

THe
The rear still needs the backplane--but I am waiting until I work out the disconnects and wiring before I do that.





Detail shots of how I mounted the Bonnie.


Here you can see how I isolated the pump while still securing it firmly to the chassis.


First case I ever felt the need to put wheels on...


The undercarriage details.The bottom will be lined with Sonex.



It's all business in there.



Another look at how the rad assembly mounts. The well nuts assure no metal to metal contact.


Reservoir--this res will be like a geyser once the Iwaki get's running And it is above the 30CM spec called for on elevation. I may fab a 36 CM cylinder res just for fun...


The acrylic is fun to look at--but soon will be lined with White 1" Sonex.



What a crazy thing to build hunh?

Loops goes like this inlet>reservoir>pump>bonnie>PA130.3>outlet

Work to do: Backplane, Electrical, Manifold, Lines to PC(s)/\

Question: what would be the best way to put a load on this beast? I was thinking maybe some pelt blocks just to generate a constant load--but I am a but foggy on how to figure the exact wattage of heat...

[Grinch]

that looks kick ass! good job!

[XS_RoB]

Holy ShiZZLe!

I am loving it! At first as I was reading the title I keyed in on External Unit... then as I am watching the build progress I thought it was in a PC Tower then , I see towards 3 qtrs of it that it is actually a stand alone unit!

Mad Skillz! I love the whole freaking thing! 2 shrouds Push Pull - NICE! Thermochill & RD30 No expense spared!

Great and keep us posted! (I'm jealous) :-/

[ghent915]

CD:
May I ask why you are combining both a bonnie AND a pa120.3? Is it to eliminate the heat from the Iwaki, so that the PA can just kill the heat dump from your components?

Why did you choose the bonnie anyways? I'm not trying to be critical at all, I'm just seriously curious on the design rationale you have for this.

I will say that your fab skills are quite good, and the system looks to be absolutely killer.
Good luck, and I look forward to hearing your comments on this!
-Ghent

[[XC] Synthetickiller]

Now that is sexy! I want one! Looking good.

I'm also curious about the bonnie.... Would even an MCR320 work better than a bonnie? Just wondering....

[DarthBeavis]

why not just drop in a phase change unit if you are gonna go thru all that work? it looks great but the performance should match the work u put into it. You went way beyond to just maintain ambient temps. There is a point when you go to the next level. I think you have reached it bud. Time for you to head over to the Vapor Phase Change forum and get to the next level.

[MrToad]

Question: what would be the best way to put a load on this beast? I was thinking maybe some pelt blocks just to generate a constant load--but I am a but foggy on how to figure the exact wattage of heat...

Hmmm. I'm not sure what you mean with "put a load"...

Do you mean to have some load on the PSU to stabilize the 12V rails?

If that's the case, you can use one of these (attached to a heatsink):

http://uk.farnell.com/jsp/Passive+Co...sp?sku=1174281

Edit: Forgot you use a RD-30, which runs @18V. Somehow I don't think we are talking about the same thing...

[ghent915]

I think he means to put a HEAT load on the system, so as to figure out the exact heat removal abilities of the system.

I have an idea on how to calculate that, but it's probably WAY too complicated (the nerd in me making it too hard). As such, lemme kick it around a bit and see what I can come up with as a better solution.
-Ghent

[Waterlogged]

Even the G 1/4 O-Ring barbs get the treatment

I certainly hope there is nothing in that thread sealant that eats O-Rings! It's also a waste of some perfectly good thread sealant.

..And no Elbows....

In a case that size, I'm sorry but that's no great feat. I could use my quick fit's with 12mmx10mm nylon tubing and not need any elbows in that enclosure. Now, if it had been the size of SG03 or Microfly, then I would be really impressed.

Other than those little things, looks good.

[CyberDruid]

Hmmm let me formulate a proper reply here...

[Waterlogged]

It's funny you should blame the O-Rings not being "captive" because I have plenty of 3/8" I.D. tube barbs here from DD to do a loop that I use for my testing and not a single one has ever leaked. I've used them on several different BI's, an Alphacool top, Csp-Mags, and I've milled the captive grooves off all my old Apogee's. If you know what your doing, you won't have problems. The trick to sealing O-Rings properly is to tighten them with your fingers until the O-Ring provides very light resistance, then use a wrench and add 1/2 turn at most. Another trick is to not set your loop up where the tubing puts a torsional twist on the fittings. Do those 2 things, and a O-Ring will never leak on you.

[CyberDruid]

Thanks for all the interest guys!

Why the Bonnie? Well....it's not like I don't have plenty of other choices...I have two BIP 240 and 3 Xflow 240 and 3 PA160 laying around...

Partly because the Bonnie is low restriction and a good form factor for the app at hand. Partly because it has been sitting in a tote for about 3 months doing nothing and I thought it would be interesting to test. And partly to take out the heat dump from the Iwaki and give a bit more headroom. Once the side panels are on you will never see it...so why waste a tasty looking rad that could go in someone's rig?

RE "go phase"/"costs not equal to results"/POV...thanks I already have phase (a Chilly1 ) and regarding your opinion...well I just don't agree and really did not build this to meet anyone's budgetary expectations. I have no interest in building phase units...subambient temops are great fun and I play there too...but this is something different perhaps you might consider how nice this would be for Pelt loops...

RE wasting sealant on O-ring fittings ...The particular reservoir I chose does not have a land or detent to trap the O-ring and keep it from slipping off the tiny land on the barbed fitting. It makes it impossible to properly tension the fittings--a slight twist of the line could either overtighten and make a leak or undertighten and make a leak. Although the probabilty for twistng is about zero (considering how hard it was to get that tubing over the Iwaki ) I am paranoid of leaks. No the sealant does not effect the nitrile.

TheO-ring seat issue is both a defect in the reservoir design and a defect in the barb design. I could change those out and use poly NPT fittings...but I don't like them either and they could stress the acrylic if overtight. The solution: fab my own cylinder res with a proper base milled with a landing/detent to receive the O-ring barbed fittings. For test purposes though I think it will suffice.

RE case size/no elbows no challenge...yeah no kidding.it's HUGE. I have been into LCing small cases for a while..I love a challenge...I built this humongous rig just to break out of the habit I have been in of jamming 25# of stuff in a 5# bag lol...but guess what: it's starting to fill up

RE load. Yes I mean applying a dummy load to the loop so that I can gauge how many watts this rig will shift. Any help with this would be great.

Thanks again for checking out this project! I value everyone's advice.

It is still underway and I will be altering/improving/complicating it as I go along. The final outcome will likely not resemble this hastily constructed (pop rivets )aluminum/acrylic structure at all.

In the end I hope to have an easily reproducable "product" in a proper powdercoated steel enclosure. This is really a "mock up" toward that end to see if my ideas about sound dampening/compartments/baffles (yet to come ) and thermal control circuits for the fans and play out in "real world" tests.

Ideally the unit should be about silent and fairly remote from the PC(s) it cools. This would allow for a more pleasant computing environment as I have become extremely intolerant of fan noise/compressor noise but still want Xtreme performance

More in a while

[CyberDruid]

It's funny you should blame the O-Rings not being "captive" because I have plenty of 3/8" I.D. tube barbs here from DD to do a loop that I use for my testing and not a single one has ever leaked. I've used them on several different BI's, an Alphacool top, Csp-Mags, and I've milled the captive grooves off all my old Apogee's. If you know what your doing, you won't have problems. The trick to sealing O-Rings properly is to tighten them with your fingers until the O-Ring provides very light resistance, then use a wrench and add 1/2 turn at most. Another trick is to not set your loop up where the tubing puts a torsional twist on the fittings. Do those 2 things, and a O-Ring will never leak on you.

Ahh the advice is so much better with a little sarcasm...

How long have you been cooling with liquid? And never had a leak from an O ring
Ooooookay..you must really know what you are doing then...lol.

FYI I fab in acrylic for a living and making a res is no biggy--I have all the tooling and equipment to make one from scratch..this PPC res is too thin a material to safely make the detent and still have sufficient threading. It was handy though...so I used it.

Appreciate the tips...but the attitude I dunno about that.

If your entire point of concern is A: I wasted sealant B: I don't know wth I am doing you need not worry...

[Waterlogged]

FYI, I've been using water since '03, so I'm no slouch either.

[Jimmer411]

I love the layout of that enclosure.


Only thing that I would change is the plastic barbs. Ive already had 1 crack in the past and that was enough to make me switch to all metal.


Ignore the haters, if you want to use thread sealant on your o-ring barbs then go for it, I still use thread tape on all mine. You never know, and personally I would rather be safe than sorry.

[ownage]

Really nice build, i hope one day to make something of my own.
What are te costs if the external case without the watercooling parts when its finished?

[CyberDruid]

I love the layout of that enclosure.


Only thing that I would change is the plastic barbs. Ive already had 1 crack in the past and that was enough to make me switch to all metal.


Ignore the haters, if you want to use thread sealant on your o-ring barbs then go for it, I still use thread tape on all mine. You never know, and personally I would rather be safe than sorry.

I have had my share of trouble with plastic and I could not agree more. It's a shame that Thermochill ships with 3/8 NPT poly when by all rights should ship with wide flange 3/8 BPSS O-Ring fittings. I guess TC figures if somebody knows what they are doing they will already have the right ones in their bin

Really nice build, i hope one day to make something of my own.
What are te costs if the external case without the watercooling parts when its finished?

OMG don't make me think too hard about that...lol I am about 1/3rd of the way done.

I do not know what the enclosure will cost--the final units will be sheet metal lined in Sonex...simple stuff and far less expensive than this prototype.

Pricing the parts...hmmm...well there are quite a few parts still to go.

I will put together some links to the products I used so far.

[MrToad]

I have had my share of trouble with plastic and I could not agree more. It's a shame that Thermochill ships with 3/8 NPT poly when by all rights should ship with wide flange 3/8 BPSS O-Ring fittings. I guess TC figures if somebody knows what they are doing they will already have the right ones in their bin

AFAIK TC ships the rads with no barbs whatsoever (unless they've changed their policy in the last month). They leave the choice open for the end user.

In your case your supplier (who hasn't done his/her homework if the barbs were NPT, as everyone and the cat knows TC uses BSPP) must have "bundled" those barbs.

[Waterlogged]

Honestly, I didn't post in this topic to get into a ping match, but rather to voice a concern and have a good natured jab at "no elbows", considering the size of the case.


You guys could abandon using DD/EK barbs you know. There's always D-Tek barbs that lets you crank down to your hearts content.

[LinusTech]

Thermochill's North American distributor bundles the poly barbs with all T.Chill radiators that they sell b/c G3/8 barbs are not readily available in North America, and 3/8NPT will do in a pinch.

[Jedda]

The AquaXtreme G3/8 barbs are readily available in NA.
Why DD doesn't just buy barbs from Thermochill along with the rads is anybody's guess. Guess they want to unload their plastic things instead.

[CyberDruid]

Honestly, I didn't post in this topic to get into a ping match, but rather to voice a concern and have a good natured jab at "no elbows", considering the size of the case.


You guys could abandon using DD/EK barbs you know. There's always D-Tek barbs that lets you crank down to your hearts content.

for the same reasons I use paste on the1/4 D-tek fittings on the res I would not be happy with the D-tek 3/8 fittings--the nut is the only "flange" to seat the O-ring and is too narrow--without a land in a properly deep detent that style of barb is unsuited to the application.

Since you have been LCing for 4 years I am sure you are familiar with the wide flange poly BPSS fittings that shipped with the Storm Rev 1.

IMO those type of barbs with a fat high Durometer O-Ring would be ideal. However I have yet to find these available anywhere. The only ones I have are on my Storm Waterblocks...

Those barbs you linked will work fine in a properly detented land--but on the Thermochill there is nothing like that to keep the ring from popping out from under the barbed fitting when tightened. Sorry finger tight and a half turn more is NOT good enough for me. A friction fit is required..one that will resist vibration, handling, shipping, reworking lines, all without a leak.EDIT: The more I think about the more paranoid I am of leaks from that PPC res. ..I will have to make one now. End EDIT.
I have had good results with brass 3/8 NPT though.

And despite the size of the case there are many improved locations available to me for mounting the elements IF I were to use elbows...It may be a large case topside but there is still a fair amount of gear below decks...the decision to eschew elbows set the course for a variety of choices...like mounting the pump vertical despite the Iwaki PDF stating DO NOT mount pump vertical

[CyberDruid]

The AquaXtreme G3/8 barbs are readily available in NA.
Why DD doesn't just buy barbs from Thermochill along with the rads is anybody's guess. Guess they want to unload their plastic things instead.

I need to get some of those. They have the proper width of flange.

[Charles Wirth]

Nice build, hopefully you keep it in the garage or outside so you don't offload the heat in the room.

The unit is still big enough to add a compressor chiller.

[MrToad]

I need to get some of those. They have the proper width of flange.

NPSM seals between a groove in the female thread (the block) and a flange on the male thread.

BSPP with O-ring (Gx/x") has a groove on the nut of the male fitting to hold the O-ring in place and seals against a flat surface on the female thread.

the wide flange poly BPSS fittings that shipped with the Storm Rev 1

The threads in the Storm, as well as the MCW30, MCW60 and so on are 1/4" NPSM, hence he "wide flange".

And FYI, BPSS is many things, but none of them directly related to plumbing. However, BSPP is British Standard Pipe Parallel, (short form: BSP, as if it's not tapered, or BSPT, then it must be parallel).

Now that I've let some steam out, back to my personal war of attrition with the rodents...