Hardware Hailea HC-500A Water Chiller Review

I was always under the impression that going below ambient opened up a can of condensation or two ?

Was there no possibility of condensation anywhere ?
I would have thought that with liquid entering the CPU block at 14°C odd, it would have droplets of water forming on the tube, and maybe on the cpu block ?

 

Looing good, I would've considered one if it was noise-free (or at least like a few case fans) and if it was in the region of +/- £300.
Always thought of building something similar myself. The basterds beat me to it. I guess I can forget that patent now ;p

 

The noise kills it for me. My PC is all air cooled and it's practically silent. The PSU is completely silent, and the CPU and case fans are Sharkoon golf ball 1000's so they are barely audible. It's also using a SSD.

The only real noise is the graphics card when it kicks up, but that only happens in the middle of gaming when I have my speakers turned up anyway.

So if I was to 'upgrade' to a water cooling setup, although I could get everything cooler and overclock higher, I would definitely be looking to at least match this kind of volume.

 

Condensation wil occur yes, but this isn't a problem if it stays on the tubing/waterblock and doesn't bead. Beading is when the condensation reaches a level where it gets so large it drops off the tubing. At that point, you're in a world of hurt, unless you have protected your PC.

 

95C for WC setup WTFFFFFFF is wrong???????????????????????????????????????

 

from the review....

"You can avoid beading one of two ways: either protect your components with a moisture absorbing material such as neoprene pads, or set the coolant temperature to be above the dew point (the temperature at which condensation starts to form) of the room. You can calculate the dew point in your room using this calculator. Fortunately, the HC-500A allows you to control the coolant temperature in 1°C steps, via the small screen and the three buttons on the front panel."

 

was an high clock > Blastflow quad 120mm-fan radiator fitted with eight Scythe Gentle Typhoon fans

must of been some high voltage going into the CPU to hit 95c

 

oops, missed that bit.
Must have skipped a page somewhere... (at work)

Still, you have a temperature limit to how low you can go.

I gave up watercooling (twice) due to the fact I couldn't get much more out of my system (noise no prob).
Also, when it came to changing GPU's, it was a right royal pain in the arse.

back to high end aircooling...

 

wow. i really need to win the lottery. would wrapping your hoses in a anti-dew material (neoprene?) let you run this at 4 degrees?

This should be in the premium player rig for august.

 

First of all TEC's aren't any less efficient than a compressor based system. I should know I have 2 completely self contained medical chillers which use TEC's to maintain a water temperature within +-.1C. It does this through PWM and is very efficient as long as the TEC's are properly cooled, in the case of these chillers they are air cooled by a rather large aluminum heatsink (~250x130x200mm) with forced air. My entire unit with 3 high power fans(20W), pump, controls, TEC, and psu draws ~3a when running a 100% duty cycle and can maintain a water temperature of 12c with a C1 Q9550@4ghz 1.5V running LinX, and prime95. That means this whole unit draws no more than 360VA when pushed to the absolute peak. It can maintain 9C water temp when just running prime95 alone. So it is roughly 80% efficient since the TDP of the CPU is ~210w. Also rapidly cycling a compressor doesn't sound like a good idea to me it is fine for an aquarium which this unit was made for since the load doesn't rapidly change. But a PC type load and the small amount of water in the loop will play hell on that compressor. My guess is it will cycle on and off several times a minute at low loads. The TEC's being solid state have no problem with this since they don't have moving parts and are not inductive loads. Go ahead and check the inrush current on that unit when it is cycling on and off like that. Also another thing is does is what type of metering unit is used for refrigerant, that alone can VASTLY affect the efficiency and load handling characteristics. If the unit is cap tube based this would be the worst possible configuration for this type of load, where a TXV or EEV would be much more appropriate yet expensive.

TLDR; compressor based water chillers are a bad idea for PC type loads unless they are set up to continuously run.

 

It would be a good start, but I'd also recommend protecting the waterblocks/components they are cooling with some moisure-protecting material too.

 

Also you guys didn't mention what temp the chiller was set to maintain in the article (I might have missed it) which is pretty important to know. Also neoprene isn't designed to absorb water, it is an insulator, like polystyrene (stryofoam). Also depending how low you go more insulating measures are needed to ensure condensation won't form, like greasing the socket and board or sealing it with putty, liquid tape, lacquer etc.

Also, I considered a aquarium water chiller for my computer but it isn't really a good idea, at least not in the manner you guys applied it.

 

Are you looking to review any of the lower models in the future? They look a little more affordable but with plenty of cooling for CPU and a single GFX card. *fingers crossed*

 

This has enough cooling potential for multiple PCs. It would be interesting to see a multiroom version of this, or maybe use this for a small supercomputer cooling setup.

 

This hailea is just a 'chiller', normally used with aquariums (commonly, reef aquariums generate more heat and overheat because the lighting is many times hotter and intense to grow the corals). I have a unit many times larger than this one on my reef aquarium.

The only problem I see is that this was designed to keep a body of water (say, a hundred gallons or so) a few degrees cooler (at worst, take 85 degrees down to 80F). So its a bit like a refrigerator, and the same precautions/problems that happen with using one to cool a PC. Chillers are not to be on 24/7, they have thermostats and are used to much larger volumes but not the constant power influx. Granted, a 150 gallon reef may be getting about 1000 watts of heat added to it (large wave pumps pumps, lighting/IR), but thats usually where a much larger chiller comes in as well. The haileas are known as some of the cheapest made units... I wouldn't trust my computer to it. To improve the performance and reduce the cycle wear on any chiller in this sort of setup, I would suggest a large rez so the heat capacity of water can be exploited and keep the thermostat from triggering all day long... preserving the compressor as well.

Check em out...
http://www.marinedepot.com/chillers__index-ap.html

 

large insulated tank

 

that thing is so overkill... I want 10 !
Watercooling is so sexy.. I've been reading about it for 8 years but never got the balls to try.

 

Undertheradar is exactly correct these are designed to cool large bodies of water which have high specific heat therefore take a long time to heat or cool. Even if 1000w of heat is being added to a 150gal of water it will take a long time to warm up. Then you have this chiller say it can produce 1500w of cooling at 25C then it would be able to cool albeit slowly to the desired temperature with a cycle of say +-2C. In this serious the chiller runs for extended periods of time then shuts off for an extended period of time while maintain the water temp. This PC setup has a very small amount of water therefor a much lower specific heat and even if the computer only applies say 400W of heating to the water this is MUCH more stressful on the unit. This is because the water temps will fluctuate quickly and erratically constantly cycling the compressor on and off also if it doesn't have a dynamically adjustable refrigerant metering device this is even more of a problem. As there will be large fluctuations in refrigerant pressures and evap temperatures. The only safe/effective way to use a compressor type cooler in this situation is a always running state or a cycled state with a large reservoir to act as a heat capacitor if you will and with both of these I recommend a dynamic metering device although the reservoir style doesn't really require it.

You mentioned the direct die coolers and these are always running, and can be statically metered even tho a TXV is recommended since it is more efficient and stable. Basically this unit is not designed for this type of use and I wouldn't trust it being left alone whereas my TEC based chiller is a different story. TEC may only be slightly less efficient but the way you are using this unit that hardly makes a difference. Also a TEC unit is silent if it is fanless unlike a compressor based unit. I suppose we will find out how reliable this is in the near future tho, I suspect it will probably last a couple months if the compressor is decent but it will be like constantly starting and turning off your car engine kinda wear.

 

Adding a radiator to your loop won't necessarily reduce efficiency. If the water temperature after the waterblocks is higher than ambient then you will benefit by putting a radiator before the inlet of the chiller. Obviously you will take a hit in terms of flow rate but if you're looking at buying chillers for your system what's another pump to compensate?
Lower inlet temperature = lower outlet temperature = more mhz?

 

I want that Antec case!!!!!