KingWin AquaStar AS-3000

Kit Test – Joe

SUMMARY: Good entry-level kit, flexible mounting options.

GW

The good guys at Kingwin sent us the Kingwin AquaStar AS-3000
watercooling kit to try out. It ships in a box with parts for just about any CPU and GPU mounting option – internal or external configuration. In addition, this is the first kit I’ve seen that ships with a noise sensor.

Key Features

  • Compatible with Intel P3 & P4, LGA 775, and AMD K7 & K8 CPUs and all series of Nvidia and ATI GPUs
  • Noise and waterblock heat sensors
  • CPU and VGA waterblocks, copper bases
  • Base unit mounts internally in 2 CD bays or externally
  • Multi function LCD panel displays information in several different colors
  • Tubing with screw-on connectors
  • Adjustable fan speed and temperature controls
  • Alarms for low water and user settable temperatures
  • Waterflow monitor (mechanical)
  • Size: 8″(200 mm)D x 5 7/8″(149 mm)W x 1 5/8″ (40 mm)H; Weight: 6 lbs

The base unit is the size of two CD ROMs:

GW

The display, when powered, shows all the critical data you’ll need:

GW

You can set 7 colors for the LED panel – red, green, dark blue, yellow, light blue, pink and purple; or you can leave it and it will automatically switch among the 7 colors. The waterflow monitor is at the lower right hand side of the case – I did find that at low fan speeds, it made a noticeable ticking noise; the LED behind it flashes all 7 colors.

The Test

KingWin’s AquaStar AS-3000 kit was tested using the CPU Die Simulator; I ran the fans at the lowest and highest settings to give a range of results.

Test
Die Temp

Ambient Temp

C/W

Waterblock Temp¹

Fans On High, 2640 rpm
49.4

25.5

0.33

37.9

Fans On Low, 1140 rpm
59.5

25.2

0.48

49.8

¹Measured by thermistor taped to bottom of waterblock.

KingWin’s AquaStar AS-3000 kit ranks in the middle of all kits (Heatsink Ranking) tested to date with fans on high. The lower the noise, the lower the performance – no surprise here.

CONCLUSIONS

KingWin’s AquaStar AS-3000 is a good entry-level kit – its flexibility is very good and performance is good for a kit this size. The only negative I saw was that the mechanical waterflow monitor makes a ticking noise that’s apparent at low fan speeds.
{mospagebreak}

I decided to test the GPU waterblock that comes with the kit separately – I wanted to see what it could do by itself.

The Test

KingWin’s AquaStar AS-3000 GPU waterblock was tested using the CPU Die Simulator; I ran the fans at the lowest and highest settings to give a range of results.

Test
Die Temp

Ambient Temp

C/W

Waterblock Temp¹

Fans On High, 2640 rpm
51.8

25.2

0.37

44.2

Fans On Low, 1140 rpm
62.1

25.1

0.51

54.1

¹Measured by thermistor taped to bottom of waterblock.

Not much different from the CPU cooler – only about 3-4ºC worse under the same conditions.

CONCLUSIONS

KingWin’s AquaStar AS-3000 GPU waterblock holds its own compared to the CPU waterblock – not substantially worse. For GPU cooling, it will do lots better than the stock coolers.
{mospagebreak}

Summary: KingWin AquaStar AS-3000 can make a very effective laptop watercooling system.

I tested the KingWin’s GPU waterblock as a CPU waterblock to see how it performed – as it turned out, not too far off from the CPU waterblock (GPU Test Results). My ulterior motive was to try to adapt the GPU waterblock for laptop CPU cooling – its size looked just right for the tight confines of a laptop.

I scrounged around the odd-parts box but could not find a clamp that would fit Asus’s Z71A hole pattern – it measured 71 mm across, so I made a clamp using a piece of aluminum. The screws were way too short and luckily I found some I could use that came with an Asetek chipset clamp; it’s not the most elegant setup but will do for a test run (the problem is that the screws are very small metric, and they are a pain to find on short notice).

Lap

Since this was a “concept” test, I only secured the waterblock with two bolts – for a more permanent setup, I would use all four.

Test Results

To test this setup, I ran Prime 95 with the waterblock in place and then again using the stock heatsink. Ambient temps averaged 26ºC; I ran the watercooling system with the fans on high. I left the CPU fan attached to avoid any problems with the system not sensing a CPU fan.

Water

Air cooling results were noticeably worse:

Air

A difference of 23ºC – substantial to say the least. I checked the CPU grease pattern after the test:

Lap

Not bad – good coverage.

Conclusions

I did not run the waterblock test too long – both system fans were off during the test, and while there is no problem with the CPU fan, I was not comfortable with the chipset fan off as well – apparently the “controlling” factor is CPU temps for both fans. As I was stressing the system with Prime 95, I did not want to push the chipset too far without adequate cooling (if I were to do this long term, I’d get another GPU cooler for the chipset).

Considering the “cooling overhead” in a watercooling system and the moderate requirement of some laptop CPUs, KingWin’s AquaStar AS-3000 makes a very nice laptop watercooling system; for laptops running desktop CPUs, this may be the only way to keep temps in line. I had to use an external 12 volt power supply to run it – no internal connections possible.

Watercooling the Asus Z71A is a snap, as the CPU is so easy to access. Most other laptops are not so easy and as such watercooling may not be an option. However, in some high temp environments, adapting a good watercooling kit might be the way to solve persistent overheating issues. Not a portable solution by any means, but for a hot DTR, not a bad way to go.

Thanks again to Kingwin for sending this our way.
{mospagebreak}

I tested the KingWin AquaStar AS-3000 both with and without the GPU waterblock in the loop to determine its impace on performance.

The Test

KingWin’s AquaStar AS-3000 CPU and GPU waterblocks were tested using the CPU Die Simulator; I ran the fans at the lowest and highest settings to give a range of results. I first tested the system with the CPU waterblock only and then added the GPU waterblock and restested.

Test
Die Temp

Ambient Temp

C/W

CPU Only, Fans On High
48.7

25.0

0.33

CPU & GPU, Fans On High
49.2

25.1

0.33

CPU Only, Fans On Low
59.6

24.8

0.48

CPU & GPU, Fans On Low
61.7

25.1

0.51

CONCLUSIONS

At the highest fan setting, there was minimal difference; as fan speeds were reduced, the impact was much more pronounced, although the 0.03 C/W difference was at high stress; under idle or low stress conditions, the difference will be minimal.
{mospagebreak}

What’s Inside?

Always curious about how things are put together, I took the AquaStar 3000 apart to see how it’s constructed. Removing the top

Inside

shows the key components:

Inside

A very compact package – I removed the front panel for a better view:

Inside Front

The fans blow from front to back:

Inside Front Off

Typical of entry level watercooling kits, the pump is on the small side – it’s submerged in the reservoir. There is also the water level monitor; note that the fill opening sticks through a hole in the top – if water overflows, it will wind up inside the case, so be careful when adding water.

Res

Lifting the front PCB a bit reveals the waterflow monitor:

Flow Mon

It’s a simple vane in the tubing leading from the waterpump. The radiator has two levels of copper tubing for a total of 16 pipes exposed to the radiator’s cooling fins. Note that there are LOTS of 180º bends – this results in high resistance; with a small waterpump such as this, waterflow through the system is modest at best.

Rad Tubes

Also note the the tubing is horizontal to airflow – this is not an optimal arrangement. The back of the cover contains a second PCB for the LCD and control functions:

Cover

This is a two layer board:

PCB

  1. An A64 3000+? . . .
  2. Athlon 64 3000+ . . .
  3. Athlon 64 3000+: Pimping An Orphan?
  4. Kingwin Arctic Watercooling Kit
  5. Kingwin XT-1264 Heatpipe Heatsink

Tags: