Refrigerator unit for liquid cooling systems!

Check this out:

http://img209.imageshack.us/img209/1...ileaultio7.jpg

Fount it here: http://www.aquatuning.de/product_inf...leistung-.html. It's an external (obviously) cooling unit that can handle up to 790W!

A quick estimate of power output by some common components:
QX9650 @ 4.0GHz/1.4V: 217W
8800GTX 3-SLI: 3x170W=510W
MB: 3*20W (NB, SB, mosfets): 60W
Total: 787W

So what about one of these to cool your system? It also comes in two smaller versions, rated to cool 265 and 395W respectively. Power consumption of the biggest unit is 375W.

Some other specs (google translate from german beware!):

"The smallest unit of the series, the Ultra 300 produced just 20 dB and is quieter than a household refrigerator"

"There are temperatures of up to 4 ° C is possible"

"Alternatively, can the Chiller at a value just below ambient temperature to the problem of condensation is to be avoided"

"Due to the low water temperatures are unprecedented opportunities for overclocking"

"We recommend the Chiller as an additional source in a refrigeration system in addition to an existing radiator"

"Recommended water run: 1200-3000l / h"

Oh, yeah and price listed on this site is €400=$608

Try the Chilled water forum, not here ;)

someone is using that exact unit over there too

Quote:

---

Originally Posted by Xilikon

Try the Chilled water forum, not here ;)

---

Sorry, was quick to post when I saw this unit, got excited!

Quote:

---

Originally Posted by evil-98

someone is using that exact unit over there too

---

camouflage ? I just saw his worklog and it use the same unit.

How does one move 790w of heat while only drawing 375w from the wall? Shens?

Quote:

---

Originally Posted by Polizei

How does one move 790w of heat while only drawing 375w from the wall? Shens?

---

How can a triple-fan ThermoChill dissapate ~500W of heat with roughly 20W going to fans? The 375W specification on that chiller will be to run the compressor. If the chiller drew 800W, yet could only dissapate 790W of radiated heat, the darned thing would overheat itself!

Ah, the possibility of ONE unit to cool the whole system... water temp of 4C... no more triple 3.120 radiators... ...less case modding... all in one circuit... low noise + high overclock combined... QX9650 @ 5.0... *dreaming* ...3x Laing D5 pumps in series :eek:

I might just be willing to splurge €400 on a unit like this :yepp:

on a side note...

Dictionary: splurge (splûrj)

v., splurged, splurg·ing, splurg·es.

v.intr.

1. To indulge in an extravagant expense or luxury.
2. To be showy or ostentatious.

v.tr.
To spend extravagantly or wastefully. :(

how big is this thing? can't tell by the pic

Quote:

---

Originally Posted by disruptfam

how big is this thing? can't tell by the pic

---

I'm interested too, I may have to save up for one of these things.... :D

Quote:

---

Originally Posted by disruptfam

how big is this thing? can't tell by the pic

---

475x360x490mm (LxBxH)

This unit has made me completely rethink my setup... I'm in chilled liquid-cooling heaven :cool:

edit:

i rechecked the specs.

ummm 400 UK lbs. Which is rougly 600-700 dollars.

Umm.. no Nol, Jinu, and almost any other custom builder could probably supply you with a cheaper and better solutions for that kind of cash.

Quote:

---

Originally Posted by zLynk

How can a triple-fan ThermoChill dissapate ~500W of heat with roughly 20W going to fans? The 375W specification on that chiller will be to run the compressor. If the chiller drew 800W, yet could only dissapate 790W of radiated heat, the darned thing would overheat itself!

---

Dude... you need to rethink what you just said.

The guy is correct. There is no way a something that consumes 375W can disapate 790W of heat. If that was the case, then you would have created energy. Using 375W to burn off 790W of heat.

Were also talking about compressors here. Compressors superchill a liquid to below freezing. This liquid is then transfered to a heat exchanger where the liquid is cooled. If you look at REAL chillers, you'll see that this is a TOY.

This reminds me of a Salt water tank chiller. These units ARE VERY BAD in the computer world. Why? because when you have over 1.5gl/m flow, it takes roughly 350W of heat to raise or lower the water temp by 1C.

Now how these units work in salt water tanks is this way. Large masss of water, with evap conditions going on, and also adding a chiller, you can bring your tank serveral C's below ambient and actually keep it there for a while.

On a PC system however thats completely changed. There is no evap, and in a sealed loop without a radiator, the water temp has no direction but to climb.

Now if that chiller cant keep up with your heat load, and im sorry, but if you ever seen a real unit that can do 790W, and THERE NOT SMALL, You'll laugh and say nice one.

In short, you'd have to be almost an idiot to pick this unit over a custom builders, and even more stupider to hook it up to a quadcore with tri sli.

Quote:

---

Originally Posted by NaeKuh

edit:

i rechecked the specs.

ummm 400 UK lbs. Which is rougly 600-700 dollars.

Umm.. no Nol, Jinu, and almost any other custom builder could probably supply you with a cheaper and better solutions for that kind of cash.

---

Price listed on german site is 399,99 €, which is not bad, if it really can deliver what it promises. And what if it's originally meant for fish tank cooling? The original pumps used in computer WC setups came from the aquarium business (read: eheim). Not to mention i DO have a 220L fish tank (although it doesnt actually require cooling as I live in northern Europe :p: )

It's an interesting idea. I will spend some time in the chilled water section reading up on the concept though, before I go ahead and splurge...

Quote:

---

Originally Posted by zLynk

How can a triple-fan ThermoChill dissapate ~500W of heat with roughly 20W going to fans? The 375W specification on that chiller will be to run the compressor. If the chiller drew 800W, yet could only dissapate 790W of radiated heat, the darned thing would overheat itself!

---

Find me a single stage phase unit anywhere on any forum that draws less power from the wall than heat it moves.

They move about 250-300w of heat, and draw 500-600 from the wall.

Not to mention, cascades trip 30A breakers @ 120v, which is 3600w, all used to move the same 300w of heat output by a quadcore processor.

polar bear, yes i implore you to do some research.

The dumbest thing you can do is buy something without any knowledge on what it is, and how its suposed to work.

Well if your britney spears, thats another story...

But once you grasp how chillers work, you'll seriously laugh. I have a 1/3HP salt water tank inline chiller, and it about 3 time this size on a 90 gallon reef. I took the system down, but i still have the chiller. I wouldnt ever think about dropping it, because

1. the unit is too masive and would look ugly
2. the temp probe can only take coolant down to 55C the lowest, not bad.
3. id need to wear ear muff's to bed because its loud.

Here is what i have:
http://www.desertcoralaquatics.com/a.../prod_956.html

Dimensions: 15.7" x 13.7" x 17.7"


So i ask you guys, think and do some research on things like this b4 you press the gun. Dont ever underestimate the stupidity of a company when its desperate for revenue.


Also you guys see the minimum flow requirements. 1200L/H -3000L/H ~ 20L/M ~ 5.2 gal/M ROFL~! who here even has a system with 1.5g/m+ ?? i know i do, but do you? Let see you pull that off with a MCP655 with 3 ek blocks + 2 chipset + 1 CPU.


As i said, you'd have to be pretty stupid.

Quote:

---

Originally Posted by NaeKuh

In short, you'd have to be almost an idiot to pick this unit over a custom builders, and even more stupider to hook it up to a quadcore with tri sli.

---

Ok ok point taken no need to rub it in dude ;)

So it probably IS too good to be true. Well darn. Thanks for the advice though, that's what these forums are for.

Quote:

---

Originally Posted by polar bear

Ok ok point taken no need to rub it in dude ;)

So it probably IS too good to be true. Well darn. Thanks for the advice though, that's what these forums are for.

---

hey... im only here to help you out.

i wasnt directing the comment at you. I was more directing at future comments from trolls an idiots who will say good things about it.

Well there isnt any. We told you in the world of compressors, and TEC's which is sub ambient cooling, there is no way you could move more heat then power required.

As i said if this was the case, AC units wouldnt be power hogs, and could be modified to provide unlimited energy then by pulling heat out of the enviornment.

Quote:

---

Originally Posted by NaeKuh

hey... im only here to help you out.

i wasnt directing the comment at you. I was more directing at future comments from trolls an idiots who will say good things about it.

Well there isnt any. We told you in the world of compressors, and TEC's which is sub ambient cooling, there is no way you could move more heat then power required.

As i said if this was the case, AC units wouldnt be power hogs, and could be modified to provide unlimited energy then by pulling heat out of the enviornment.

---

No worries, no offense taken! Time spent in chilled water heaven was disappointingly short, but it felt good though for as long as it lasted :D . Now back to the REAL world of triple WC loop with triple 120.3 radiators and 6x 2400 rpm / 38 dBA fans. I think I'm going to need a drink... :(

http://img142.imageshack.us/img142/4...edvodkano2.jpg

If you look closely, you may notice the QUAD-ice cubes bathing in the CHILLED liquor in front of my TRI-sli setup... :rofl:

vodka is my fav alko :)

Quote:

---

Originally Posted by disruptfam

vodka is my fav alko :)

---

hey, anyone know the coolant properties of vodka? :ROTF:

(sorry, might have forgotten to take my anti-mania drugs again)

Quote:

---

Originally Posted by polar bear

If you look closely, you may notice the QUAD-ice cubes bathing in the CHILLED liquor in front of my TRI-sli setup... :rofl:

---

:up:

oh and incase no one welcomed you

:welcome: to XS. :D

Quote:

---

Originally Posted by polar bear

hey, anyone know the coolant properties of vodka? :ROTF:

(sorry, might have forgotten to take my anti-mania drugs again)

---

heh, it would evaporate

Quote:

---

Originally Posted by NaeKuh

:up:

oh and incase no one welcomed you

:welcome: to XS. :D

---

Thanx, NaeKuh!

Quote:

---

Originally Posted by NaeKuh

edit:

i rechecked the specs.

ummm 400 UK lbs. Which is rougly 600-700 dollars.

...

---

Wrong. The price is in Euros, which is currently at a record level of (checking xe.com...) 1.517 USD; GBP (British Pounds) are 1.989 USD.

Quote:

---

Originally Posted by NaeKuh

Dude... you need to rethink what you just said.

The guy is correct. There is no way a something that consumes 375W can disapate 790W of heat. If that was the case, then you would have created energy. Using 375W to burn off 790W of heat.

...

---

... only if there is a requirement to cool below ambient temperatures. The article is likely using a value for total saturation of the heater core. After all, a higher figure looks better on paper, right?

Tell me, how big of a radiator would you need to keep your water temps at ambient?

Quote:

---

Originally Posted by Polizei

Tell me, how big of a radiator would you need to keep your water temps at ambient?

---

LOL :rofl:

I assume you're suggesting a scenario that doesn't include a compressor, right?

Do you know what youre talking about?

Quote:

---

Originally Posted by zLynk

Wrong. The price is in Euros, which is currently at a record level of (checking xe.com...) 1.517 USD; GBP (British Pounds) are 1.989 USD.



... only if there is a requirement to cool below ambient temperatures. The article is likely using a value for total saturation of the heater core. After all, a higher figure looks better on paper, right?

---

Didnt check recient exchange rates, either way this unit is still too expensive.

uhh... seriously do you know what your talking about? Did you also miss my calculation where i said it would require AT LEAST 5.2gl/m Do you have an IWAKI RD-30? or a little Giant Aquarium pump? :rofl:

Now lets think about it.

IF this unit uses 375W, it would need some form of compressor or TECs. Do you even know how a phase or TEC unit works? Zlynk, you sound like you know half of what you should know. Your concept is all wrong from your first post to this.

And what medium would you use to keep the unit right near ambient? Also, once again FLOW RATE, how do you intend to keep that?

No this unit uses a significant amount of energy which leads me to think its either compressor or TEC. Unless you can think of another way to cool down.

Now, Lets add some numbers because i think you need see where im getting at.

Q6600 G0 @ 1.45Vcore ~ 180-190W of heat
G80 GTX @ LOAD ~ 180-210W x 3

no... it wont work no matter how hard i think.

Yes, I understand how refrigeration works. No, I don't intend to frivilously spend the money necessary to maintain a 5GPM flow. Besides, racing season is about to open, and $1000 weekends add up quick.

Take a further look into the initial link... you'll find pictures. The device uses a compressor and heater core.

Polizei, I laughed above because it is impossible to maintain an ambient coolant temperature in a traditional water-cooling setup. There will always be a positive differential when comparing coolant temperature to ambient temperature, unless some form of state-change is used. The most basic form of state change is through the use of a standard compressor based refrigeration loop.

Quote:

---

Originally Posted by polar bear

low noise

---

Umm... sure you might not need triple rad with powerful fans, but do you think a compressor is quiet? I know my window air conditioner isn't.

How much does the thing cost?

Quote:

---

Originally Posted by zLynk

Yes, I understand how refrigeration works. No, I don't intend to frivilously spend the money necessary to maintain a 5GPM flow. Besides, racing season is about to open, and $1000 weekends add up quick.

Take a further look into the initial link... you'll find pictures. The device uses a compressor and heater core.

Polizei, I laughed above because it is impossible to maintain an ambient coolant temperature in a traditional water-cooling setup. There will always be a positive differential when comparing coolant temperature to ambient temperature, unless some form of state-change is used. The most basic form of state change is through the use of a standard compressor based refrigeration loop.

---

No kidding. Anyone who knows phase knows that. I dont get what youre trying to prove here. Seems like youre were trying to say before that you can get ambient water temps without using TEC or refrigeration. Theres no way unless your reservoir is just that - a lake or a reservoir.

Quote:

---

Originally Posted by polar bear

Oh, yeah and price listed on this site is €400=$608

---

Well this thread has certainly sparked some discussion! It's been a learning experience, however sadly it's probably too good to be true. Now I've decided to go for a 120.2+120.3+120.3 triple-loop setup for my striker II cpu+gpu x3 + chipset... after all.

But it sure would have been interesting to see how the unit would have performed in a liquid-cooling setup - maybe in addition to the rads? Guess we'll never find out...

well this time i translated it in google and actually read the link.

And it does have a compressor with R134.

However that big black thing is im guessing the heat exchanger, where the cooling takes place. I just dont know how effective this thing would be without a large reservoir.

Also i dont know how your going to get a pump to match that liter per hour mark. Unless you want to go all out aquarium pump. :\

Just noticed this thread, and I must say it's always a bit amusing when the discussion gets personal and people are called stupid, and the consensus is wrong.

So first of all, you're not going to get a correct opinion on a chiller unless you post in the chiller or phase sections where us phase guys hang out.

Secondly (drum roll please) you CAN move more heat with a chiller than it pulls from the wall! In the refrigeration industry we call it COP (Coefficient of performance) and sometimes EER or SEER in the US (which are goofy, so I'm not going to explain).

COP = Power Moved / Power Consumed = (Watts)/(Watts) i.e. dimensionless

So, as an example the COP of a Danfoss Universal Compressor SC10G (ASHRAE conditions):
@-30C = 0.53
@-20C = 1.02
@-10C = 1.57
@ 0C = 2.16
@ 10C = 2.78

(I'm looking at the datasheet for a Danfoss SC10G 50Hz compressor @ danfoss.uk because they use Watts for both cooling power and electrical power)

@0C, this compressor will move 764 Watts, and consume 353 Watts. It's not creating energy, it's moving it like a pump. It flows into the evap (764 Watts of it) from the water, which of course cools the water down in the process. The condenser then has to reject all of the power as heat (minus comp shell cooling) so ~1000 to 1100 Watts.


Hilarious! that you guys have talked/bullied yourselves into believing that a chiller is a scam! I have no idea about the particular chiller in this discussion, but $608USD is much more than the parts would cost even buying single parts (no volume discounts) from a refrigeration supply store.

Quote:

---

Originally Posted by DetroitAC

@0C, this compressor will move 764 Watts, and consume 353 Watts. It's not creating energy, it's moving it like a pump. It flows into the evap (764 Watts of it) from the water, which of course cools the water down in the process. The condenser then has to reject all of the power as heat (minus comp shell cooling) so ~1000 to 1100 Watts.

---

:am: My apologies.

Thats what it was... I had a feeling that the power draw/moved assumption wasnt right. You have to move the heat from the water + heat from the compressor in the condenser. Power draw doesnt have much to do with how much heat you can move. Bong cooling, being an example. No power draw other than a pump.

I blame my mistake on Nyquil. :rolleyes:

Thanks for stepping up and apologizing Polizei. In my opinion all you did was give an incorrect opinion of how something works. No problem there, this is a discussion forum :up:

I think Polar Bear was due an apology (NaeKuh apologized, sort of, in post18), which was an excellent way to welcome him to this forum :up: :welcome: Polar Bear!

Quote:

---

Originally Posted by DetroitAC

Thanks for stepping up and apologizing Polizei. In my opinion all you did was give an incorrect opinion of how something works. No problem there, this is a discussion forum :up:

I think Polar Bear was due an apology (NaeKuh apologized, sort of, in post18), which was an excellent way to welcome him to this forum :up: :welcome: Polar Bear!

---

Thanks DetroitAC; actually I have been lurking for awhile as this is a great forum. But I have been reluctant to post as discussions seems to heat up very quickly and people do seem to overreact all the time unnessecarily; as is what happened also in this thread. I have a demanding job; a hectic life with family and all, and I don't really need more tension in my spare time as well, as I consider this a hobby and nothing more. We are all enthusiasts here, with a passion for computers and technology, and probably most of us are here for that same reason.

As for who is right; I don't know yet - I am not the expert here. Hence my reason for posting ;). But I would say as much as this: for the people who have been wrong in this case, it would seem better to have done this without an attitude. NaeKuh - I still think some of your posts have been unnessecarily sharp, and in a way I hope you have been right from the start, as I would be a bit embarrased if I were you if it turns out you have been wrong all the time; as you for instance seemed to warn about "trolls" that believes in the concept of water chilling and that it's in fact possible to move more heat than a unit pulls from the socket. According to you, then both DetroitAC and zLynk have been the trolls in this thread - and perhaps I have been the stupid idiot for posting here initially.

I do enjoy a good discussion, that's the reason I started this thread; hopefully without too much tension though. So thanks both of you (NaeKuh too) for your welcome.

And also - I quickly became aware of the fact that I had misposted; initially I didnt even know there was a chilled water forum. Well, know i know! ;)

Quote:

---

Originally Posted by DetroitAC

Just noticed this thread, and I must say it's always a bit amusing when the discussion gets personal and people are called stupid, and the consensus is wrong.

So first of all, you're not going to get a correct opinion on a chiller unless you post in the chiller or phase sections where us phase guys hang out.

Secondly (drum roll please) you CAN move more heat with a chiller than it pulls from the wall! In the refrigeration industry we call it COP (Coefficient of performance) and sometimes EER or SEER in the US (which are goofy, so I'm not going to explain).

COP = Power Moved / Power Consumed = (Watts)/(Watts) i.e. dimensionless

So, as an example the COP of a Danfoss Universal Compressor SC10G (ASHRAE conditions):
@-30C = 0.53
@-20C = 1.02
@-10C = 1.57
@ 0C = 2.16
@ 10C = 2.78

(I'm looking at the datasheet for a Danfoss SC10G 50Hz compressor @ danfoss.uk because they use Watts for both cooling power and electrical power)

@0C, this compressor will move 764 Watts, and consume 353 Watts. It's not creating energy, it's moving it like a pump. It flows into the evap (764 Watts of it) from the water, which of course cools the water down in the process. The condenser then has to reject all of the power as heat (minus comp shell cooling) so ~1000 to 1100 Watts.


Hilarious! that you guys have talked/bullied yourselves into believing that a chiller is a scam! I have no idea about the particular chiller in this discussion, but $400USD is much more than the parts would cost even buying single parts (no volume discounts) from a refrigeration supply store.

---

Blah i take back what i say and i started to correct myself on post 35, however i still dont see how this would work without a large reservoir. Also i dont see how your going to get the flow requirements for it.

Well, I'm glad that has been smoothed out somewhat, thanks for posting NaeKuh. I don't want to sound too arrogant...but I've been a professional refrigeration engineer for ~10 years now, and I'm pretty sure I'm (mostly) right.

As for this specific chiller, I don't know how it's controlled, why they specify the flowrate they do etc. It seems to me that this is probably an aquarium chiller, and it's occurred to someone that it'll work just fine as a PC chiller. It's such a nicely packaged unit and the sales volume for just PC overclocking I belive are quite small compared to aquariums, so I doubt this much effort would be worth it unless it's primarily for aquariums. I'll look into it a bit...

It IS in fact an aquarium chiller: http://www.hailea.com/E-hailea/product-1.htm

Camouflage has recently posted in the chilled liquid cooling section that he's using one of these units in his setup: http://www.xtremesystems.org/forums/...d.php?t=178594

So we seem to have established that a unit CAN in fact move/dissipate more heat than it itself requires? Well, if so, is it then totally unrealistic that this particular unit can dissipate an energy amount of 790W while it pulls 375W from the socket? Also the required flow rate is listed as 1200-3000 l / h. Is this completely without reach despite with a HUGE pump? Doesnt the D5 / MCP655 have a maximum discharge of 1200LPH? Well, mine does according to it's specifications, unless I can't trust my eyes... ;) Well then what about running a couple of these in series to get sufficient flow?

I am not interested at this stage to spend an enormous amount of time researching the concept of refrigeration and the science behind it, nor spend weeks building a custom made chiller. So hence when I discovered this unit at a german etailer site, for a price that is within reach (albeit expensive; the price is actually less than for, say 3x PA120.3 rads with a total of 9 fans, at about €500=758USD). I was naturally excited about the concept of one unit to cool the whole system. Maybe not to extreme, sub-zero temperatures (as I suppose insulation of the tubes would then be required to avoid condensation and heating of coolant on the way to CPU) but simply as an alternative to several 120.3 radiators and multiple loops which is probably required to dissipate the ~800W from my striker II CPU/chipset/3xGPU setup.

Here's the complete specifications for the particular chiller (courtesy of google translate) and the link: http://www.aquatuning.de/product_inf...leistung-.html

The cooling aggregates of Hailea be for over 15 years in good Durchlaufkühlern consumed by other manufacturers. The new series of ultra heat exchangers are made of titanium. This can be the perfect cooler for water cooling systems. Another positive feature of the device is the low noise level. The smallest unit of the series, the Ultra 300 produced just 20 dB and is quieter than a household refrigerator. Furthermore, the prices of the new series in encouraging regions slipped off.

New revision:
Cool new design
Stable materials for longer life
Lower Geräuchentwicklung
Durability suitable

This run cooler essentially replaced the radiator, as the water on the level you set temperature runtergeregelt. There are temperatures of up to 4 ° C is possible, but note here that the temperature differential condensation can form. Similarly Isolierzubehör (for hose and CPU), we will soon also offer. Alternatively, can the Chiller at a value just below ambient temperature to the problem of condensation is to be avoided.
Due to the low water temperatures are unprecedented opportunities for overclocking. The life of your components is increased by a multiple. Furthermore, this system is also for No Noise interesting applications, as the Chiller in another room can be. We recommend the Chiller as an additional source in a refrigeration system in addition to an existing radiator. How can you demand (About cycles / Benchmark / midsummer etc) off the normal radiator and the Chiller then cooling.

In order to estimate the cooling possibilities here following non-binding instructions. The exact loss of benefits please contact the manufacturer of the hardware.
CPU: 80 (regular) to 150W (overclocked)
GPU: 20 (small cards) to 150W (current models eg 8800)
MB: 10 to 20W per component

The aufsummierte value of the system may be under full load the specified value of the cooling capacity of Chillers not.

Note on coverage for dispatch by Parcel Service (DHL / GLS), we have a special padded carton manufacturing. For the Händling and packaging costs € 15, we need to calculate VAT. This amount is in collection or in pallet shipments, provided the Chiller without the discount will be sent packing.

About compatibility: The Chiller, with connections for 1 inch hose (inner diameter of approximately 19 to 20mm) format. Included are 1 pair of straight and 1 pair of 90 ° angled connections. To Chiller with standard connectors to use, we recommend Art 37018 (G1 / 4) and 37019 (for G3 / 8).

Technical data:

Compressor Performance: 1 / 2 hp
Current consumption: 375 watts
Kältleistung: 790 watts (suitable for extremely About bars and Extreme Cooling)
Recommended water run: 1200-3000l / h.
Weight: 22Kg
Refrigerant R 134 A
Dimensions: 475x360x490mm (LxWxH)
Wärmetausche: Titan
Temperature controller: Digital reading 0.1 ° C setpoint setting in 1 ° C increments

Ah yes also I requested to the moderator (Fugger?) to move this over to the chilled water section, but all I got in reply was an automated message: <fugger@xtremesystems.org>: Recipient address rejected: No such user

I have invited camouflage over to this thread to throw some light on the subject. But be warned; it is likely he will bring a battleaxe and some heavy armor :eek: as I warned him the discussion has been fairly hot so far...

polar bear your discharge is at no restrictions, at 0 elevation.

The moment you drop blocks, you increase restriction, increase flow pressure, and also reduce flow.

Thats why im asking how it would work, because martain calculated somewhere around 1-1.5 gpm was the average flow in a system.

The required flow is 5.3gpm which means you would need 4 D5 pumps, if they scaled perfectly, but by doing so, you'd probably burn each motor out really quickly from the increased rpm speed.

My Iwaki could probably do this, and now its got me thinkn of getting another 1/3hp chiller and seeing if it will work.

Quote:

---

Originally Posted by polar bear

Ah yes also I requested to the moderator (Fugger?) to move this over to the chilled water section, but all I got in reply was an automated message: <fugger@xtremesystems.org>: Recipient address rejected: No such user

---

You need to contact IFMU for forum requests like this. He'll move it after he pistol whips you a couple of times for posting it here in the first place. :rofl:

I'm not giving up on this one yet... Actually the prospect of using one of these units is tantalizing... thinking to perhaps use it to lower temps just below ambient to avoid condensation.

Would it be possible to use one of these to cool BOTH of my CPUs? Ie joining the two loops with Y-piece into chiller then perhaps reservoir then Y-tube to divide into separate circuits. I have a DDC ultra (600LPH head 4,7m) in the Q6600 loop and a D5 (1200LPH head 3,7m) in my E6600 loop.

Problem with this might be different flow rate and also of course different resistance in each circuit meaning water goes route of least resistance meaning mediocre/less flow in one loop. Is it feasible though? Any comments on this (without too many inflammatory statements ;) )? Perhaps running an Eheim 1250 after the Y-joining of the 2 circuits to boost flow rate into the chiller? Mind all this is just on the idea level - need more info first.

And why would I need a reservoir (at all) ?

Quote:

---

Originally Posted by polar bear

I'm not giving up on this one yet... Actually the prospect of using one of these units is tantalizing... thinking to perhaps use it to lower temps just below ambient to avoid condensation.

Would it be possible to use one of these to cool BOTH of my CPUs? Ie joining the two loops with Y-piece into chiller then perhaps reservoir then Y-tube to divide into separate circuits. I have a DDC ultra (600LPH head 4,7m) in the Q6600 loop and a D5 (1200LPH head 3,7m) in my E6600 loop.

Problem with this might be different flow rate and also of course different resistance in each circuit meaning water goes route of least resistance meaning mediocre/less flow in one loop. Is it feasible though? Any comments on this (without too many inflammatory statements ;) )? Perhaps running an Eheim 1250 after the Y-joining of the 2 circuits to boost flow rate into the chiller? Mind all this is just on the idea level - need more info first.

And why would I need a reservoir (at all) ?

---

once again you couldnt get the flow rate required for this system.

1200-3000l/h = 1200/60 = 20 lpm / 3.79lpm/gpm = 5.2gpm

The average h2o system with cpu + gpu + radiator, which should be less restrictive then this unit, will pull about 1-1.5gpm on martin's calculator. There is no way in hell a D5 will pull 1200l/h with blocks and incline. I think you missed me telling you this about 4 times already.

Your blocks will mess up your flow.

You still need an additional 3.5gpm! where you going to get that flow?

There is no single pump minus aquarium fish hobby pumps, or my rediculously expensive RD-30 that could do this. You dont have enough flowrate to meet the requirements of this system.

Well, fair enough. I am actually thinking to use one high-flow pump (not sure which one yet) for a loop that goes straight to chiller then pumps cooled water into an insulated tank. In that way, I can hook up my systems or perhaps several loops to the tank, similar to the way camouflage has done it.

Quote:

---

Originally Posted by polar bear

Well, fair enough. I am actually thinking to use one high-flow pump (not sure which one yet) for a loop that goes straight to chiller then pumps cooled water into an insulated tank. In that way, I can hook up my systems or perhaps several loops to the tank, similar to the way camouflage has done it.

---

The reason why i say you need a reservoir is because that way you have a headroom.

According to thermo, it takes about 350W traveling at 1.5gpm for water temp to change 1C. Thats a lot of heat required. By having this chiller attached to a larger reservoir, you have large volumn acting as headroom. This will always allow your chiller to be used far less, it would pulse instead of be on all the time.

As i said i have one of those chillers, and i already considered the idea of using it, but there isnt a pump out there which doesnt dump a ton of heat back into the water that you could use to pull this off.

Then i discovered the most covetted pump of our hobby, the RD-30. 30feet head on a pump a little bigger then the D5. But its also priced 4x-4.5x if you include the required psu.

So yeah, if your going to spend all that money, why not just phase? A builder could easily build you something much nicer and which will last longer, better tuned for computers,

The cost/peformance ratio on this unit is so not worth it. 250 dollar pump system. 600 dollar chiller, then you'd need blocks, and everything.

Last time i checked, almost any custom builder could whip something for you cheaper on a SS phase unit, and drop the gpu's on water. Still end up being cheaper.

Quote:

---

Originally Posted by NaeKuh

There is no way in hell a D5 will pull 1200l/h with blocks and incline. I think you missed me telling you this about 4 times already.

---

No, I just didnt understand why initially. I thought my D5 would give 1200 LPH under ANY circumstances, well of course now I understand it is under IDEAL circumstances only and it would mean at almost no resistance i.e. head. I have just started looking into Martin's and others flow/head calculations, kinda a lot of info to swallow in just a few days time but I'm getting there. And with a separate circuit for the chiller only you wouldnt have to subtract flow rate for blocks, rads etc. The resistance to flow would be only that of the chiller itself.

Quote:

---

Originally Posted by NaeKuh

The reason why i say you need a reservoir is because that way you have a headroom.

According to thermo, it takes about 350W traveling at 1.5gpm for water temp to change 1C. Thats a lot of heat required. By having this chiller attached to a larger reservoir, you have large volumn acting as headroom. This will always allow your chiller to be used far less, it would pulse instead of be on all the time.

---

Yes now I am starting to get this too... thanks for explaining though. Just can't understand why this unit requires such a massive flow rate...

Quote:

---

Originally Posted by NaeKuh

As i said i have one of those chillers, and i already considered the idea of using it, but there isnt a pump out there which doesnt dump a ton of heat back into the water that you could use to pull this off.

Then i discovered the most covetted pump of our hobby, the RD-30. 30feet head on a pump a little bigger then the D5. But its also priced 4x-4.5x if you include the required psu.

---

Ehh but the RD-30 has the same flow rate as the D5 listed, which I presume is under ideal conditions... 1200 LPH. It differs of course in the head pressure of 10m. It certainly is expensive though.

In his thread camouflage is using an Eheim 1250 (1200LPH, not sure about head).

Quote:

---

Originally Posted by NaeKuh

So yeah, if your going to spend all that money, why not just phase? A builder could easily build you something much nicer and which will last longer, better tuned for computers,

The cost/peformance ratio on this unit is so not worth it. 250 dollar pump system. 600 dollar chiller, then you'd need blocks, and everything.

Last time i checked, almost any custom builder could whip something for you cheaper on a SS phase unit, and drop the gpu's on water. Still end up being cheaper.

---

:shrug: as I said earlier I havent decided on anything just yet. But seeing camouflage's thread got me thinking as to whether it may work. Phase... dont even get me started. Perhaps after a few months of reading/researching yes. Well, I'm going to stick with my new rads (which I haven't even received yet :p: ) then lets see what the next upgrade would be.

Quote:

---

Originally Posted by NaeKuh

Dude... you need to rethink what you just said.

The guy is correct. There is no way a something that consumes 375W can disapate 790W of heat. If that was the case, then you would have created energy. Using 375W to burn off 790W of heat.

Were also talking about compressors here. Compressors superchill a liquid to below freezing. This liquid is then transfered to a heat exchanger where the liquid is cooled. If you look at REAL chillers, you'll see that this is a TOY.

This reminds me of a Salt water tank chiller. These units ARE VERY BAD in the computer world. Why? because when you have over 1.5gl/m flow, it takes roughly 350W of heat to raise or lower the water temp by 1C.

Now how these units work in salt water tanks is this way. Large masss of water, with evap conditions going on, and also adding a chiller, you can bring your tank serveral C's below ambient and actually keep it there for a while.

On a PC system however thats completely changed. There is no evap, and in a sealed loop without a radiator, the water temp has no direction but to climb.

Now if that chiller cant keep up with your heat load, and im sorry, but if you ever seen a real unit that can do 790W, and THERE NOT SMALL, You'll laugh and say nice one.

In short, you'd have to be almost an idiot to pick this unit over a custom builders, and even more stupider to hook it up to a quadcore with tri sli.

---

Buy a huge powerful one, and run some pipes though the walls. Central water cooling! :D

I use a Laing Ultra and a 1250 in my loop, and believe me, the Laing alone have a better flow in that restrictive system then the Eheim! ;)

Think the best way to go is a dual setup of Laings. I use the chiller in a other room and have to go threw about 10 meters of tube. No problem for my one Laing.

Here is my system prepared for the chiller:

http://www.abload.de/img/imgp36748az.jpg

Under the PSU i can add the chillerloop outside of the system, close the tap on the radiator and i will not loose any coolingpower on the radiator. If i wish to use my system portable i can close the two taps under the PSU and open the one on the radiator, the system runs like it ever did with just the triple radiator. :)

Here is my chiller-area in the other room:

http://www.abload.de/img/imgp3714cg0.jpg

Thanx for your input, just as I thought this thread had died ;)

Wow, that's a neat setup you've got there, very tidy n organized, good work on the cable management. Tubing routing looks good too.

It's nice to see that people here have been able to implement this chiller -I first saw camouflage's setup with one of these, and now your. Just shows it's actually doable. You have a build log of your waterchilled rig? Also, any issues with condensation?

Thanks!

No issues with condensation so far, but i run just 20&#176;C water for 24/7 usage. With that temperature i will have no problems with it, not even with 30&#176;C ambient in summer*. Lowest watertemperature i tested so far was 8&#176;C with an ambient of about 15&#176;C. I have some isulation-stuff lying around here, maybe i will add it soon.

Sorry, no worklog for my system. It is a neverending story.. Every few weeks i change something on it. :D


* http://www.cb-gruber.at/index.php?op...d=29&Itemid=32

Quote:

---

Originally Posted by funkflix

Here is my system prepared for the chiller:

---

Wow... that is the cleanest tube routing I've ever seen in my life. It's almost invisible... you really have to do a doubletake to see that it's there.

@ funkflix,

Your rig has left me speechless... still looking at it in awe => :clap:

How far have you been able to take your E8400 on that rig? Thinking of getting one for my system. Although I probably won't go chilled water yet, I've got a fairly decent WC setup.

EDIT: which case is that?
Also, where is that Laing Ultra hiding? Can't see it in the photo...

Thanks a lot guys! :)

Quote:

---

Originally Posted by polar bear

EDIT: which case is that?
Also, where is that Laing Ultra hiding? Can't see it in the photo...

---

It is a Chieftec Mesh Bigtower. The Laing is above the HDDs. ;)

My E8400 is running at 4200MHz 1,39V 24/7, but i still achieved that with my stock watercooling, but way better temperatures now.

Here is a quick 32M i did 30 minutes ago, with about 12&#176;C water.
http://www.abload.de/img/32m_4700i1z.jpg

Here some more shots of my system:

http://www.abload.de/img/imgp3729ds3.jpg

http://www.abload.de/img/cpub7n.jpg

http://www.abload.de/img/sliz1b.jpg

http://www.abload.de/img/auslassrn8.jpg

If I was living alone without any female to use up the monthly pay check, that one would be cool, got an iwaki MD20 somewhere, that would be perfect for it I think :) But i doubt I would even then, use that much money on one item, I'm kinda cheap :p

Wow, funkflix that is a gorgeous build, and one heck of a restrictive loop.

The eheim just doesn't have the head needed for this application. The high flow rate requirements are probably because it uses a a large plate or a large multi-coil heat exchanger and the system probably measures water temperatures near the outlet so it requires high flowrates to help keep clean pure water from freezing in the HX. You have to remember this is designed for aquariums, you can't put antifreeze in an aquarium so they have to keep the fluid moving to keep it from freezing.

A large body of water for the system to draw from just helps smooth out any ripples from swift load changes. It also will take longer to pull the fluid down from ambient with a fixed thermal capacity. Basically, it will take 5 times longer to pull 5 gallons down to temperature than 1 gallon but, that means it will take 5 times longer to heat back up as well. it's not that simple or that linear but, thats the basic idea.

IMHO this system is a great pre-built system with a nice integrated controller and takes a lot of the guesswork and fiddling out of a chiller but, a lot of the fun as well. Tearing apart an AC and hacking together a chiller is actually really fun, especially once you start doing your first pull downs and seeing -28C unloaded. You have to be careful though or it will give you false illusions of grandeur. :rolleyes:

This certainly won't be an extreme chiller but, it's definately an extreme water system and I think a very nice replacement to a radiator and some fans. Were this system to be used for a 24/7 loop with really consistant loads I would say run it as is with big pump like an RD30 or parallel D5s on a loose loop or parallel DDCs(atleast until I get my D5 tops rolled out) for a restrictive loop. with 10-20&#37; antifreeze depending on how cold you want to run it and monitor your local temperature and humidity levels carefully. At 90% humidity condensation can form a 3C below ambient.