Printable View
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
... 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?Quote:
Originally Posted by NaeKuh
Tell me, how big of a radiator would you need to keep your water temps at ambient?
LOL :rofl:Quote:
Originally Posted by Polizei
I assume you're suggesting a scenario that doesn't include a compressor, right?
Do you know what youre talking about?
Didnt check recient exchange rates, either way this unit is still too expensive.Quote:
Originally Posted by zLynk
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.
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.Quote:
Originally Posted by polar bear
How much does the thing cost?
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 zLynk
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.Quote:
Originally Posted by polar bear
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.
:am: My apologies.Quote:
Originally Posted by DetroitAC
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!
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.Quote:
Originally Posted by DetroitAC
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! ;)
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.Quote:
Originally Posted by DetroitAC
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.
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: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.Quote:
Originally Posted by polar bear
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.
The reason why i say you need a reservoir is because that way you have a headroom.Quote:
Originally Posted by polar bear
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.