Phase9 (3 evaps unit, 2.5 HP + 3/4 HP)- Single-stage WorldRecord ?= -66.6*C @ 215W

Printable View

11-03-2006, 04:26 PM
piotres

Phase9 (3 evaps unit, 2.5 HP + 3/4 HP)- Single-stage WorldRecord ?= -66.6*C @ 215W

Hi

Some mounths ago, Symphy, nice memeber of XS.org forum ask me about build unit to cool His incoming, serious hardware (Conroe X6800ES + 2* high end CrossFire config) . So we have decided on 3-evaporators system - with 2 rotaries.
One rotary = one CPU evap controlled with CPEV, second rotary = 2* GPU evaps, controlled with 2*CPEVs.
One condensor, 3.0 kW [!] :slobber: with 34W fan.
Oilseparator for CPU loop :fact: .
There were many complications, eg. case was cutted 2 times - first was too small for further rotary (first plan was two 9k BTU rotaries), first that had been 1.9 kW condensor.
Symphy really want monsterous rotary for CPU loop, first I've bought Matsu 28000 BTU [!] rotary which didn't work, finally I've found even better (= stronger :D ) compressor for CPU loop -8.3kW, around 2.5 HP [!] :slobber: SIAM NN33VAAMT.

Final specs :

Compressors: SIAM NN33VAAMT for CPU loop (2.5 HP), and LG QB134PE12A for GPUs loop (3/4 HP)
evaps : piotres's custom
Expansion valves: 3*CPEV Fach "PZ" (each evap controlled with CPEV)
condenser:3.0 kW with 34W fan
Refrigerant : R507 & R410a blend for CPU and GPU loops
Case : Chipboard case with wheels, painted red
others: REFCO 4*panel gauge, ShineYear oilseparator with float for CPU loop, homemade receiver for GPU loop, sight glass for GPU loop, one EBM 30W fan on the back wall

Let's allow to talk pictures - full story of build :) .

http://www.members.lycos.co.uk/piotres1/phase9/1.jpg

First plan : ordinary 9000 BTU rotaries + 1.9kW condensor :p: .

http://www.members.lycos.co.uk/piotres1/phase9/3.jpg

http://www.members.lycos.co.uk/piotres1/phase9/4.jpg

http://www.members.lycos.co.uk/piotres1/phase9/5.jpg

First case ...

http://www.members.lycos.co.uk/piotres1/phase9/6.jpg

Very nice REFCO panel gauges :) .

http://www.members.lycos.co.uk/piotres1/phase9/9.jpg

http://www.members.lycos.co.uk/piotres1/phase9/10.jpg

http://www.members.lycos.co.uk/piotres1/phase9/11.jpg

Evaporators and suction lines :D .

http://www.members.lycos.co.uk/piotres1/phase9/12.jpg

http://www.members.lycos.co.uk/piotres1/phase9/13.jpg

And now we have started to go crazy :D ...new 3.0 kW condensor with 34W fan.

http://www.members.lycos.co.uk/piotres1/phase9/14.jpg

http://www.members.lycos.co.uk/piotres1/phase9/15.jpg

Is it happy ? :rolleyes: ;)

http://www.members.lycos.co.uk/piotres1/phase9/16.jpg

Homemade receiver for GPU loop (before 2 CPEVs) .

http://www.members.lycos.co.uk/piotres1/phase9/17.jpg

http://www.members.lycos.co.uk/piotres1/phase9/18.jpg

28000 BTU Matsu and oilseparator appeared :slobber: ..that time it's stared to go XTREME :D .

http://www.members.lycos.co.uk/piotres1/phase9/20.jpg

http://www.members.lycos.co.uk/piotres1/phase9/22.jpg

http://www.members.lycos.co.uk/piotres1/phase9/23.jpg

http://www.members.lycos.co.uk/piotres1/phase9/24.jpg

More and more pipes :D .

http://www.members.lycos.co.uk/piotres1/phase9/25.jpg

Just after that, I've put some R507 into CPU loop, and I saw that Matsu don't want to start ...seems like it's broken :( :( :mad: .

Now quite long time with no work, new rotary searching and .....

http://www.members.lycos.co.uk/piotres2/phase9/1.jpg

http://www.members.lycos.co.uk/piotres2/phase9/2.jpg

SIAM already mounted, temperature on 215W heatload, uncompleted insulation .

http://www.members.lycos.co.uk/piotres2/phase9/3.jpg

Some frost on GPU loop :)

http://www.members.lycos.co.uk/piotres2/phase9/4.jpg

http://www.members.lycos.co.uk/piotres2/phase9/5.jpg

http://www.members.lycos.co.uk/piotres2/phase9/6.jpg

http://www.members.lycos.co.uk/piotres2/phase9/7.jpg

http://www.members.lycos.co.uk/piotres2/phase9/8.jpg

http://www.members.lycos.co.uk/piotres2/phase9/9.jpg

http://www.members.lycos.co.uk/piotres2/phase9/10.jpg

http://www.members.lycos.co.uk/piotres2/phase9/11.jpg

Completed, case screwed, ready to some tests :toast: .

http://www.members.lycos.co.uk/piotres2/phase9/12.jpg

TESTS in 20*C room :) :

SLI/CF evaps went into vaccum :p: .

http://www.members.lycos.co.uk/piotres2/phase9/13.jpg

CPU evap went into vaccum :p: :D .

http://www.members.lycos.co.uk/piotres2/phase9/14.jpg

SLI/CF evaps loaded with 115 W each (2* 115W ).

http://www.members.lycos.co.uk/piotres2/phase9/15.jpg

IT'S DEVIL :eek2: :rehab: :shocked: .

215W heatload on CPU evaporator = -66.6*C - Guys, is that a WR in Single-stage temperature ? :sofa:

http://www.members.lycos.co.uk/piotres2/phase9/16.jpg

even -67.7*C with 215W :banana:

http://www.members.lycos.co.uk/piotres2/phase9/17.jpg

All pressures in load (215W for CPU, 330W for GPU loop) :D .

http://www.members.lycos.co.uk/piotres2/phase9/18.jpg

http://www.members.lycos.co.uk/piotres2/phase9/19.jpg

http://www.members.lycos.co.uk/piotres2/phase9/20.jpg

Some tests on ancient Barton 2500+ mobile ...+3.05 ghz/2.1v idle around -74-77*C, stress -71-74*C :fact: .

http://www.members.lycos.co.uk/piotres2/phase9/21.jpg

Ready to go .

Stefan told me : " - 70.1*C at the evap during 3D Mark 06 CPU Test with retail E6700 and 1,76 Volts . " And that was only first play with unit, I think CPEV can be tunned bit better even :p: . Soon he'll put X6800 + 2*X1950XTX under evaps :toast: . I hope he'll write sth more here :) .

Regards
Peter
11-03-2006, 04:42 PM
filmbot

Wow, very nice work. Its so big for a SS. I can only imagine how much this cost. Again, very nice work :toast:
11-03-2006, 04:46 PM
metalazzo

such monstrous power :|
i guess it's pretty noisy right ? :D
11-03-2006, 04:52 PM
afireinside

2.5HP that's MASSIVE! What's the power draw on it? Noise? Very impressive :toast:
11-03-2006, 04:55 PM
Nosfer@tu

Quote:

Originally Posted by metalazzo

such monstrous power :|
i guess it's pretty noisy right ? :D

Both points are of no importance! It is Xtreme systems dude :)

That is simply cracy Piotres :p :nono: Greenpeace will FRY you :D HAHHA
11-03-2006, 05:03 PM
K404

WOW! Just....wow!
11-03-2006, 05:24 PM
wdrzal

well I see its already done ,but you could have used one compressor with 1 cpev(constant pressure expansion valve)possibly 2 to have available the same pressure to all,cpu +2 gpu. Then use Epr valves on the suction to adjust suction pressure for each individual head. as you know its the suction pressure that controls temp.I didn't read the thread yet ,or do the actual design to determine if and how well it would work,but ideas for someone in the future. I think it would work just fine.

This applies to all engineering and design: If you can't make it work on paper,it won't work in real life.

what that means if you can't draw(drafting) it (what ever your making)you can't make it & if you cant make the numbers work(formulas for thermodynamics or pressure or energy or structural strength or what ever) You can't build it with 100% certainty it will work and work safely as is supposed to be designed to do.

Walt alway work safe !!!
11-03-2006, 05:44 PM
Master_G

Sheesh, that is a crazy piece of work, damn impressive!

G
11-03-2006, 06:25 PM
osiris999

great job
11-03-2006, 06:39 PM
damnnit

That is freakin awesome. Great work. I can only imagine the pricing for that unit.
11-04-2006, 12:09 AM
piotres

Thanks guys for all positive comments :toast: .

Big rotary isn't more loudly that small, rotaries are generally bit noisy so You know :D I think it's not 24/7 rig (rather hehe) so that isn't so important .

@Afterinside : Powerdraw of big rotary is around 2200-2800W (I don't remember exactly), smaller 600-800 W.

Price - I want to say : For me this unit is XTREME, but price of it is NOT xtreme :D . I'll sent You prices on PM, if somebody wants to know price, please sent me PM :) .

Regards
Peter
11-04-2006, 01:54 AM
Duniek

how many kg weight this monster SS ??
:eek2: :eek2:
11-04-2006, 02:08 AM
jinu117

OMFG...
And to think I detuned 1st stage of my red cascade to use 600w under normal running condition when I was sick of watching 1200w....
Now this is SS to extremity :) Someone wanted to be able to bench without going 2 stages I presume :)
Nicely done.
11-04-2006, 02:46 AM
Jort

polish guys are =>>>>> extreme for sure :D

yeah thats one fffriggin!!! powereater and a hell of a coooool beast ^^
11-04-2006, 02:55 AM
SexyMF

Personally I enjoy the colour.
11-04-2006, 02:56 AM
Brettbeck

WOW! Awesome build!!
11-04-2006, 03:02 AM
Pete

Man thats cool!!! Should of got a video of it running so we can hear it grone!
11-04-2006, 03:03 AM
LittleDevil

Great unit :toast:

Quote:

polish guys are =>>>>> extreme for sure

Agree with that :p:

Regards
11-04-2006, 03:15 AM
jplikesnos

That looks great, looking forward to seeing the results with X6800 + 2*X1950XTX !
11-04-2006, 03:53 AM
Unknown_road

Don't the rotary's overheat after a while? I build a cascade with rotary's once and when suction pressure was very low the unit couldn't run for more then an hour or something.
11-04-2006, 03:56 AM
dinos22

great just great

this isn't an easy task and you've just created a monster there....i am envious man :toast: :toast: :toast: :toast:
11-04-2006, 04:42 AM
Athens[2004]

Quote:

Originally Posted by piotres

Hi

IT'S DEVIL :eek2: :rehab: :shocked: .

http://img90.imageshack.us/img90/3060/nowaidj9.jpg

:clap: Nice work you going deep in pro :fact:
11-04-2006, 05:03 AM
Jack

The temps are just insane :D I like it a lot!

The powerdraw of that 2,5HP rotary is xtreme aswell!
11-04-2006, 01:33 PM
piotres

Quote:

Originally Posted by Unknown_road

Don't the rotary's overheat after a while? I build a cascade with rotary's once and when suction pressure was very low the unit couldn't run for more then an hour or something.

No, it's well cooled, fan from condensor blow quite cool air on it, and back fan throw warm air away from case :) .

Thanks for all coments guys :toast: .

Regards
Peter
11-06-2006, 09:55 AM
quintus

Damn!!!!!!!!!!!!!!!!!11
thats the THING on this few months with the other cascadeautosuper r14 -130......

beer!!!!

2.5hp u for real?????? i think u are mad u can half that and make a cascade lol 1hp and 1,5hp
do u have some prices for that beast? w8ing for pm...
11-06-2006, 10:01 AM
Big SturL

Could the term "extreme" be more suitable?
11-06-2006, 01:48 PM
symphy

Heyho ;)

I have problems with my 2 X1950... Both are not coldbugged, but overclocking... :X with subzero temps, not even 700 at the cores are possible. I'll have to design a new BIOS... But resultswill come for sure.

The unit works like a charm, i'll open a thread at the xtreme overclocking forum.
I can't validate anything right now; the CPU-Z version i used after mounting the unit (i had to hurry up ^^) was 1.36.

5,2 GHz promised; more like 5,28 to 5,3
11-06-2006, 08:18 PM
Revv23

wow, that is simply a crazy unit. piotres you are the most creative SS builder i have seen, very nice work and it came out really good looking!!

It's amazing that it runs CF and CPU with temps colder then most deticated SS'.

have fun with it symphy.
11-06-2006, 10:25 PM
[cTx]Philosophy

Verry impressive to say the least...

I definately wanna see this thing in action with a beast system..

And prices would Own also :)

Not that I could ever afford one, but it would be good to know how much JIC..
11-07-2006, 06:57 AM
Xtasy

that unit is awsome, i was testing that unit for a week with sympy and the rest of the overkill.oc member. the temps are huge :D piotres you made a very good job :)
11-08-2006, 07:05 PM
Benjie

*changes in to ninja suit and hunts down symphy*

Actually if i did that i dont think i could move that thing! Its HUGE! At least you can use your electricity meter as a Fan when you switch that bad boy on :P
11-08-2006, 08:33 PM
wdrzal

Quote:

Originally Posted by piotres

Thanks guys for all positive comments :toast: .

Big rotary isn't more loudly that small, rotaries are generally bit noisy so You know :D I think it's not 24/7 rig (rather hehe) so that isn't so important .

@Afterinside : Powerdraw of big rotary is around 2200-2800W (I don't remember exactly), smaller 600-800 W.

Price - I want to say : For me this unit is XTREME, but price of it is NOT xtreme :D . I'll sent You prices on PM, if somebody wants to know price, please sent me PM :) .

Regards
Peter

Large hp produces more mass flow but not lower temps in the small evaps used in cpu cooling. By using large hp compressors,you just un-necessarily added a higher amperage draw on the electrical end=higher power bill for no reason.Plus higher total load.

The higher hp compressors also produce a higher "heat of compression" figure that makes a larger condenser necessary to reject total heat load(BTU's) for no reason.

A lower hp compressor with smaller condenser could do the same job,cheaper & better.while I,m sure your unit works great,in refrigeration, bigger is not always better. it's not like auto racing were pure hp benifits..Is the boiling point of the gas that produces the low temperatures & enough mass flow is needed to carry away the heat. each gas can carry a given amount of heat per pound,the flow rate needed and heat a given gas can carry can be calculated..but you must stay within certain internal velocities for the system to work correctly. To large a compressor just adds extra heat and probably more mass flow than needed.While I didn''t take your components and produce a design to prove what I,m saying,just by looking and reading what you posted ,I beleive you over sized the compressors. Running most standard compressors in a vacuum to produce lower temperatures is not a good idea from a longevity of compressor standpoint. I don't know if you did this or not.

There is a learning curve as refrigeration is more complex than it appears at first. Take the time to learn how to calculate all the parameters needed to build a great design that does not wast electricity.

Not discounting your work only giving advice so you can produce even better units. At this site I mostly try to give engineering advice in layman's terms if some wonder about my advice. Also I skiped many other important details in design and even some explanations are not complete as I don't intend this post to be a complete book.

keep on building :woot:
03-29-2007, 05:39 AM
piotres

Some shoots from CEBIT 2007 :D :

http://www.extremecooling.net/index/...cebit_095_.jpg

http://www.extremecooling.net/index/..._img_0030_.jpg

http://www.extremecooling.net/index/..._img_0031_.jpg

Loaded Conroe X6800 @5 ghz :banana: .

http://www.extremecooling.net/index/..._img_0028_.jpg

Overkill-OC Team in action

Regards
Peter
03-29-2007, 06:45 AM
1badhabit

Wow, nice job Piotres!!
03-29-2007, 06:48 AM
sacha35

nice job Piotres, very good job there mate.
03-29-2007, 07:32 AM
Ssilencer

Impresive unit Piotres!
You're great :up:
03-29-2007, 07:41 AM
Pete

Amazing work and temps
03-29-2007, 09:46 AM
[XC] gomeler

Those rotaries must hate you for the amount of vacuum you are pulling :p: My 1/2HP rotary would hit 100+ celsius running 15" HG, can't tell how much you are running but it looks way down there. Amazing temperatures though, I'm glad I don't pay for the power to run that thing!
03-29-2007, 10:14 AM
pinto

Great work peter!
My unit is really good but this one is really beautiful!
03-29-2007, 10:36 AM
Barnsley-Bill

Your the man Peter :) realy impresive unit and scores well done mate.
03-29-2007, 11:02 AM
TopherTony

Thats beautiful, its a work of art. Great job
03-29-2007, 11:07 AM
n00b 0f l337

With the oil sep he could add something like r11 to separate with the oil and cool the compressor.
04-25-2008, 07:35 PM
manithewindlord

wowwww.... i was wondering what would be the power consumption of that system ? 1000watts ? must be very healthy on electricity bill.
04-25-2008, 08:01 PM
Planet

lol at dragging up a 1+ year thread