Review

PCApex.com, November 22, 2006
Vigor Monsoon II TEC CPU Cooler

Manufacturer: Vigor Gaming

Sponsor: Vigor Gaming

Author: Dominik "Anonymous" Haas

Editor: Gary "GlitterKill" Mullins | Product info link. | Read similar articles. | Available at FrozenCPU.

PAGE  | 1  | 2  | 3  | 4  | 5  

Performance

On to my favorite part, and what you've all been waiting for. Testing was performed with the following:

AMD Athlon64 3200+ Venice core (@ 2300MHz and 1.55 VCore)
DFI LanParty UT nF4 Ultra-D
2 x 512 MB Mushkin BH5
XFX GeForce 6800GT graphics card (OCed to 440 / 1180)

Core and ambient temperatures were taken with a Fluke Series 52 II thermocouple thermometer as the CPU was loaded 100%. This load was achieved using CPUBurn-In. Subtracting these two temperatures yielded the desired dT values (difference between the CPU and ambient air temperature). Only these will be reported, since they're the only data yielding any useful information. For a more in-depth discussion of the testing methodology and equipment, please check here.

The Monsoon II was tested with the stock fan and three 92mm reference fans to characterize the fan performance (only 1 fan at a time):

Scythe DFS922512L ("Low RPM fan")
Panaflo FBA09A12M ("Medium RPM fan")
Vantec Tornado TD9238H ("High RPM fan")

The stock fan was at its highest speed in all the tests. Also, in order to test the reference fans, the wires connecting the stock fan were cut and resoldered with a 3-pin connector, shown below. All tests of the stock fan were done prior to soldering.

Each Monsoon II & fan combination was mounted 5 times. The individual data points are displayed in the graph below. Mountings are fairly consistent.

On a side note, Vantec Tornadoes like the taste of thermocouples almost as much as fingers. Recalibration is always fun and time consuming. Yay.

In the graph below, the Monsoon II is compared against the stock AMD cooler, Icetank CPU cooler, and a custom water cooling setup comprised of:

Swiftech Storm CPU water block
Dangerden DDC (or Swiftech MCP350)
'76 Chevy Caprice heat core + two Panaflo M1A 120mm fans

Inspection of the graph shows just what a well designed cooler aided by a small TEC can do. The Monsoon II with its stock fan alone performed better than a, arguably, top-of-the-line water cooling setup. Quite an amazing feat. The reason is the TEC. TECs are solid state heat pumps and, thermodynamically, can be modeled the same way as a conventional refrigeration setup. Essentially, you're using electrical energy to create an artificial temperature difference in order to remove heat from the CPU more rapidly. Therefore, a heat sink with an active cooling element has the potential to reach lower die temperatures than any passive cooler. The threshold temperature for passive heat sinks is the ambient air temperature ( at 100% efficiency... impossible to achieve), whereas that of a heat sink with an active element is lower.

Anyway, back to the pretty colors. The Monsoon II rocked it. Once I swapped the stock fan for the reference fans I actually managed to get below ambient temperature at load with the Vantec Tornado once. You could never hope to achieve temperatures like that even with a car blower fan attached to a huge radiator and a water block of perfect geometry cut out of solid diamond (for all you cooling nerds out there). Magic? No. As I mentioned, you're using energy to move energy. The energy the TEC is using comes from the power supply and eventually shows up on your electric bill. I wouldn't worry too much, though. For example, assuming the TEC is loaded to 60W and run 24 / 7 (quite a stretch), you'd only be in the hole about $4 / month in Texas.

One final thing I want to point out is the big performance leaps from one reference fan to another. This is also due to the TEC. The efficiency of the TEC, technically called the coefficient of performance, depends on the input current to the TEC and is not directly related to the ambient temperature like passive coolers. This makes characterizing performance more difficult. Add to that the fact that the Monsoon's fan is also dependent on temperature and you have quite a complex equation. Basically, the ambient air temperature and CPU temperature aren't proportional and what may be true with one heat load may not be true for another, as is true when comparing passive coolers. "So what's the point of all this?", you may ask. Well, during testing I was able to achieve a relatively stable TEC temperature and full fan loading so the results are at least consistent. Also, if you have a high heat load CPU and the cooler isn't performing as well as you'd hoped, try swapping the fan out, or attaching another (it can mount 2 fans). This has a big impact and may shift the operating point on the TEC's coefficient of performance graph to a higher value, allowing you to get those near-ambient results. Just like overclocking, there is a lot of room for tinkering and optimization. I don't have the time to test every conceivable scenario so I decided to test it like any normal heat sink. Just keep what I've said in mind when interpreting the results.

The graph is shown below in tabular form. The "95% confidence" refers to the error bars in the graph and "thermal performance number" is something I cooked up to measure performance of a heat sink relative to cost. More details can be found in this forum thread. Once again, the active element complicates things, especially when it's approaching ambient temperature. You can see the Monsoon has a wide range of thermal performance numbers, depending on fan selection. Surprisingly, you actually do get your money's worth with the stock fan, but to really wring the Monsoon for all its worth you'll want to get just above ambient air temperature. That's the g-spot.

	Load Avg. dT(°C)	95% Confidence(°C)	Thermal performance number ($ x dT) [Lower is better]
Stock AMD HSF	15.24	±0.31	N/A
Custom Water Cooling	5.82	±0.08	1455
nMedia Icetank (Stock fan)	12.88	±0.82	502
nMedia Icetank (Low RPM ref. fan)	16.22	±0.70	713
nMedia Icetank (Med. RPM ref. fan)	13.18	±0.20	632
nMedia Icetank (High RPM ref. fan)	11.32	±0.12	600
Vigor Monsoon II (Stock fan)	4.82	±0.18	723
Vigor Monsoon II (Low RPM ref. fan)	9.96	±0.28	1543
Vigor Monsoon II (Med. RPM ref. fan)	5.1	±0.36	811
Vigor Monsoon II (High RPM ref. fan)	0.58	±0.62	95(Outside of normal parameters)

Conclusion

I really had fun reviewing this cooler. Vigor has managed to tackle all the problems associated with a thermoelectric cooling element and integrated it seamlessly into a tidy package. The construction is great and the only problems I foresee users having with it is the annoying stock fan and heating of adjacent components. The TEC gets hot. With the low RPM reference fan on I actually burned my hand when checking for condensation thinking I was touching the cold side, not the hot. I guess it takes the nerves a few seconds to distinguish between them. The cost may be a stumbling block for some, but I wouldn't hesitate to recommend it based on my experience.

The performance is stellar, achieving numbers that even top-of-the-line water cooling kits could never hope to, for the price of a cheapo kit. You may be thinking, "Is water cooling dead?" Certainly not. Water cooling still has the upper hand on the issues it always has, namely, noise, high heat output CPUs, and wow-factor. Other than that they don't have a lot going for them. This, by the way, coming from a guy who has had several water cooling systems over a number of years. I expect to see more hybrid coolers like the Monsoon II hitting the market in the near future. Active cooling elements seem to be the next stepping stone after optimization of heat sink geometry and maximization of thermal conductivity through heat pipes.

The Highs Construction is superb. Excellent performance. No additional power supply needed. Alarm sounds if something bad should happen. A lot of DIY potential here for the tickerers. Same convenience as any other conventional heat sink. Seamless integration of thermoelectric technology.

The Lows

• Expensive.
• Stock fan is annoying.
• TEC heats adjacent components.

I would like to extend a big thank you to Vigor for allowing me to review their product and for their patience. :)

PAGE  | 1  | 2  | 3  | 4  | 5