OCZ's Phase Change Cooling Kit!!!

[AKHandyman]

Well, with the news that Intel is going to release a sub $50 water cooing kit (http://forum.pcstats.com/showthread.php?t=41974), I started to get excited ... that is until I saw this ... OCZ's Cryo-Z Phase Change Cooler (http://www.tweaktown.com/reviews/1321/ocz_cryo_z_phase_change_cooler_tested/index.html)!!!

So far, the review is promising and I swear, If I can get my hands on one of those and a E8500, .... I will RULE AND PWN! ...:D

[-=CrAnSwIcK=-]

lol, 8ghz?? (15)

[owen]

What temperatures can phase change go down too?

[Qbical]

What temperatures can phase change go down too?

negative ALOT lol

[AKHandyman]

Easily a -30F ... nothing to sneeze about ... I've had my window open when it has been -50F ... but not for LONG ... :D

[Assimilator87]

Holy crap, I thought OCZ gave up on the Cryo-Z. This is awesome.

[Devils00Champs]

make you own liquid nitrogen phase change cooler :D, seen one few years back someone made, was getting -100C on a P4 full load

[AKHandyman]

make you own liquid nitrogen phase change cooler :D, seen one few years back someone made, was getting -100C on a P4 full loadSounds like an interesting project, but I'd hate to have anything go wrong ... that's dangerous stuff if it should get out of its container ... I've seen movies where people's fingers break off!!! :scared:

I'll stick to something a little more tried and true, thank-you ... ;)

[Devils00Champs]

Look at what i found for you AK
http://www.madshrimps.be/?action=getarticle&number=1&artpage=2422&articID=555

[AKHandyman]

Hey that was awesome ... real "cool" :p

I remember reading about that sometime ago, I forgot that those guys did that ... but it doesn't last long, I mean the LN2 because of the evaporation ... not sure you'd be able to get through Crysis ... :D

[yahooadam]

why are you posting "news" from 2006 :s

[owen]

2003 actually!..........

[Devils00Champs]

2003 actually!..........

which link would be 2003?, dont see any link in this thread dating that far back.

[AlphaChicken]

that liquid nitrogen "pot" was not phase change. it is just simple direct contact of a super cooled heatsink to the processor. nitrogen is poured carefully into the heatsink every few minutes in order to manage the temperature.

also, be careful when buying pre-made phase changers. they can perform ALOT worse than it first appears. with phase changers temperature is not the limiting factor on your overclock. h/o, i wrote a huge long post about it. ill link.

[AlphaChicken]

Here are the links:

Original thread (http://forums.pcstats.com/showthread.php?t=41453&highlight=phase+change)
My Post 1 (http://forums.pcstats.com/showpost.php?p=354738&postcount=6)
My Post 2 (http://forums.pcstats.com/showpost.php?p=354880&postcount=11)

[Dizzious]

Screw liquid nitrogen. I want my pc cooled with liquid helium - it's 60c colder; and much more thermally conductive.

"The thermal conductivity of helium II is greater than that of any other known substance, a million times that of helium I and several hundred times that of copper." (http://en.wikipedia.org/wiki/Helium#Helium_II_state)

[AKHandyman]

Bottom line, everyone wants the best in cooling, whether it's water, phase-change, peltier, liquid nitrogen or helium ... geez :p and they want it as cheap as possible ... :D

[Devils00Champs]

Dont think ud want helium cooling due to that "bug" when the newer intels wont boot under too cold of a temperature.

[owen]

Bottom line, everyone wants the best in cooling, whether it's water, phase-change, peltier, liquid nitrogen or helium ... geez :p and they want it as cheap as possible ... :D

lol well said:thumb:

[Assimilator87]

Here are the links:

Original thread (http://forums.pcstats.com/showthread.php?t=41453&highlight=phase+change)
My Post 1 (http://forums.pcstats.com/showpost.php?p=354738&postcount=6)
My Post 2 (http://forums.pcstats.com/showpost.php?p=354880&postcount=11)

Can you explain how a PCC can run at a cooler temperature, but only handle a small heat load? I don't understand how that's possible.

[AlphaChicken]

I havnt read up on phase change in a long time. I kinda gave up on it when i realized the cost of tools for making custom ones.

If i remember correctly it all has to do with subcooling and superheating. These temperatures are dependant on the refrigerant used. the main thing you want to watch for is that no liquid refrigerant is returning to the compressor, as this will kill it.

also, as to your question. the thing that matters is that the unit is able to hold a certain temperature, not reach a certain one. The proper way to do things is figure out the heat load of your overclocked processor in watts at idle then tune the PCC to that load. That way you know that at idle (your processor's coolest state) no liquid will be returning to the compressor. And then at load, (seeing that at full loads your processor should be around the same watts most of the time) the heat load will increase, but the PCC will be still be negative if tuned right. The main thing is that the unit is tuned correctly as that it doesnt die. Temperature doesnt matter as much, seeing as you tuned for the overclock temp of your processor.

[AKHandyman]

Though I didn't ask, thanks for the 'splaination ... :p But if I buy a kit ... there should be instructions that come with it, I hope ... :rolleyes: When all else fails, read the instructions ... :D

[yahooadam]

I'm not sure how true what alpha chicken is saying is

However, i probably haven't done as much reading on it, but i've never heard of a problem with liquid refrigerant returning to the compressor, this may however be a problem on custom builds due to the construction of the system

Also, theoretically there shouldn't be an liquid returning, because its only liquid when its under pressure, however, once it comes out of the capillary tube at the head, there is more space and it evaporates

[AlphaChicken]

It is liquid when the coolant is under pressure at a certain temperature. The reason pressure plays an important part in that, is that for ANY gas, the higher the pressure, the easier it is to become a liquid. So if a gas at a certain pressure turns to liquid at say, 30 celsius, that same gas at a higher pressure will turn to liquid at a higher temperature, say 40 celsius.

Now, the way tuning works, in simple terms is the ratio of refrigerant being forced into the evaporator from the high pressure capillary tube. Similar to ice, when the refrigerant expands and makes cold, there is only a certain low temp it can reach. Ice never gets below 0 celsius, so why would a refrigerant with set properties ever get below its min. phase change temp? The reason this is important, is that if too much refrigerant is forced into the evaporator, not all of it will change phase, because the evaporator wont be heated fast enough to get the refrigerant above its gasious state temp. for the given pressure of the evap. This is how liquid refrigerant can return to the compressor. The compressor is only meant to operate on gas. Just think about it, water is NOT possible to compress, why should any liquid. It will simply destroy the compressor.

As far as retail units go, they are usually tuned for a wide range of loads. Usually you wont be able to make liquid return to the compressor with a retail unit. I have heard that too high of a load will kill a unit. I am not sure why this is. But as long as the load maximum of the unit you purchase has a fair margin above the overclocked load of your processor at LOAD operations, you should be fine.

The main thing you want with a phase cooler is that it is able to give a CONSTANT and steady negative temperature for an overclock. Since you can overclock first or estimate a super high overclocks range, you can technically obtain "un-obtainable" overclocks because you can just tune the unit with new gas to the heat output of that overclock. This is why custom units are so much better if done correctly.

I am going to do some research and find out why too high of a heat load will kill a unit. Im guessing that its because all the gas cycles though.

in order to explain i need to state how phase works in simple terms:

1: gas is compressed in compressor so that it turns into a liquid at a higher than normal temp. gas is also heated when compressed above the point of condensation.

2: gas is pumped into condenser (a radiator), where it is cooled to the point where it becomes a liquid. the main point of the compressor is to make the as condense at higher than room temps as long as it is at a high pressure.

3: the condensed (liquid) refrigerant is pumped into some sort of metering device. this can be special valves (upwards of $200 costs) that are precisely adjustable, but usually this is done with a VERY small tube called the capillary tubing. the point of the metering device is to keep the refrigerant under high pressure until it reaches the processor and to limit the amount of refrigerant entering the evaporator.

4: the liquid refrigerant enters the evap, which is a low pressure area and immediately expands back into a gas, due to the high temps of the evap (which is the heatsink on the processor) and the lower pressure area. When a gas expands it absorbs energy creating a cold area. this is where the sub-zero cooling comes from.

5: The gas is now pulled out of the evap through a suction line. Generally there is still some evaporating to be done, so the suction line usually gets cold and frosty. (Usually the capillary tube is wrapped around the suction line to keep the liquid refrigerant cold and assure that it stays liquid long enough).

6:The now gasious refrigerant is back in the compressor again.

NOW. The reason I believe that too high of a load can kill the compressor and your processor is that:

The condenser only has a certain amount of heat it can remove. Just like any watercooling radiator. The difference is that the compressor creates ALOT more heat than a watercooling pump PLUS there is the heat from the compression of the refrigerant, PLUS the heat from the actual processor.

If the condenser cant remove enough heat to turn all the compressed and super hot refrigerant into all liquid, then in turn the evaporator wont be cooled as much because there isnt as much phase changing going on. Then in turn, the refrigerant returning to the compressor is even MORE hot. It will just build and build and get hotter and hotter. The condenser wont be able to keep up at all and your compressor will die from overheating as well as probably your computer from not being cooled properly.

Im not sure if that theory is completely correct but it makes alot of logical sense to me. I will look into it more.

Hope this helped you guys. :)

[yahooadam]

I believe your right

If the condenser cant cool the liquid sufficiently then eventually overheated liquid will enter the compressor, which will only be able to handle so much at which point it dies

However, in retail units at least, there are nice 120MM fans over the condenser to make sure its cold enough (same thing in your car with air conditioning)

I'm not sure about custom ones, maybe because they don't have the same electronics or something to make sure the condenser is cold enough?

Anyway, i think i see what your saying, but I'm still not sure, i would still think that the "liquid" coming out the head would have so much room in the suction line, plus the suction line being warm enough that it would evaporate

[AlphaChicken]

but see thats the thing. the suction line wouldnt be warm enough. the suction line almost always frosts over some. that is one thing the prometiea mach II's did to prevent this. they heated the suction line to prevent condensation, but it could have very possibly also helped heat refrigerant so that the suction line didnt return liquid to the compressor.

as far as the condenser in custom units, they also have big air conditioning fans as well. alot of times bigger than 120mm for high load units. it pretty much all comes down to the tuning of the unit. the tuning involves cap tube length, cap tube inside diameter, the amount of refrigerant charged to the system. for example, if it is a high load unit, you can widen the cap tube and shorten it a bit, plus add more refrigerant. This way more refrigerant is pumped at higher volumes faster in order to keep up with the evap heating back up.

the way i see it is that if a unit is tuned correctly to the specific load it was intended for, then there shouldnt be any problems at all. Usually the way retail units are tuned, you can be pretty liberal in what heat loads you put on them.

there is obviously some info i am missing. for instance, how do people (you see pics of it all the time) turn on their units to show people the frosty evap head. in theory according to how the unit works, this would kill the unit as there IS NO heat load. IDK though. I havnt read up on this stuff in a a long time.

[yahooadam]

maybe it doesnt take long to frost over?

I assume the heat buildup issue would take quite a while to be a problem, like after the suction tube has all frosted up and such

[AlphaChicken]

the main thing here is to make sure that the specified heat load of the unit is within the heat load you intend to overclock into. the old mach IIs WILL not at all handle an overclocked quad. the new OCZs probably will. basically what im saying is when buying one of these retail units, look at the maximum heat load handled NOT how cold it gets.

[AKHandyman]

For a minute here, I thought I was back in Mr. Wizard's class ... !!! Whew! Way too much for an old man to comprehend ... :p If the kit works, then maybe I'll check into it ... you guys can keep on discussing it, but I'm outta here ... :D

[AlphaChicken]

you should read up on it AK. its really interesting stuff. I was going to build my own....that would be the way to do it. :D