mCubed's HFX mini: Fanless HTPC "heatsink case"

Cases|Damping
Viewing page 4 of 4 pages. Previous 1 2 3 4

TESTING

The following tools were used during testing:

Because the noise of a fanless system is more or less constant — unless you start using your optical drive or push your hard-drive into constant seek mode (e.g. defragmentation) — most of my testing consisted of an examination of power consumption and thermal dissipation. Please note that the thermal diodes were not calibrated, so thermal measurements should only be evaluated relative to each other, not in absolute terms. All of measurements were taken after 30 minutes in the indicated Activity State.

Ambient temperature during testing varied between 20°C and 21°C.

mCubed HFX mini system temperature / power
Activity State
CPU
System
AC Power
Idle
46°C
45°C
71W
Full CPU Load (2x CPUBurn)
53°C
46°C
78W
Full System Load (ATI Tool + 2x CPUBurn)
57°C
51°C
95W

Thermally, the HFX mini worked very well. The maximum temperature of 57°C can be considered to be perfectly safe. However, I would advise against pushing the system very hard in a hot >35°C environment. The temperatures inside the system are relatively high and I'm not sure how good that is for component life. With fanless systems, you usually face the issue of heat buildup inside the enclosure. With the large ventilation holes in the top directly above the motherboard and expansion cards, mCubed has ensured that convection works as well as possible.

It's also interesting to note how the temperatures of the CPU and system changed over time. Within a second of launching CPUBurn, the CPU temperatures jumped up 2~3°C. After that it, slowly increased by about 1°C per 3 minutes to reach its maximum after about 15 minutes. Then it remained within a small margin of +/- 1°C. Once I stopped the load test, the temperature fell by about 3°C within the first minute or so. After that it gradually decreased to reach its Idle value after about 20 minutes. The system temperatures on the other hand took much longer to react to the Load scenarios. After reaching 51°C during the Full Load test it took the better part of 60 minutes in order for the temperature to go back down to the mid-40s.

One final about the heatsinks: The right side heatsink becomes rather warm after extended load periods. It was never too hot to touch, however. The left heatsink remained cold to the touch throughout all of my testing.

Even under full load the system never drew more than 95W from the wall outlet. Considering that this sample included two 3.5" hard-drives, a high-end soundcard and a TV-tuner, this an impressively low value. Of course, actually using the TV-tuner and the hard-drives to record a TV-show might add a couple of watts but then you'll probably not be pushing both CPU cores to 100%. I was quite surprised to see that simply running ATI Tool on top of the two copies of CPUBurn resulted in a 17W increase of AC consumption. This is more power than I thought the onboard graphics card would draw! While running two instances of CPUBurn I was still able to do light browsing and typing with a relatively responsive system. Once I added ATI Tool to the equation things slowed down significantly and the system came to a virtual standstill.

SOUND TESTING

Unfortunately I don't have access to the sophisticated sound level meters and high resolution studio quality digital recording system that Mike and Devon use for their articles. I can provide a detailed account of my subjective experience, which Mike says is the most important part of any acoustic assessment.

First of all I think it's important to mention under which conditions the system was reviewed: A 12sq m home-office with a wooden floor in a very quiet European town. Almost all of the testing was done at night when there's as little ambient noise as you'll ever have in normal household — no cars going by, no radio or TV. Our heating system was turned off at the night... and even the cat was asleep.

The first component that I listened closely to was mCubed's new EF28 power brick which they introduced in our interview with them back in August. It's basically a fanless external AC-DC convertor that provides up to 210W at 12V for an internal DC-DC adapter. I had to move to within 20cm of the unit in order to be able to pick up a faint electronic buzzing.

Next I moved on to carefully listen to the system itself. Having used quite a number of very low-power Mini-ITX systems in the past, some of them fanless, I was keen on seeing how quiet the HFX mini would be. Obviously with a fanless system the only components that are responsible for creating noise are the hard-drives and the optical drive. I therefore decided to divide my listening experience into 4 categories:

  • Idle
  • productivity applications (Firefox with several tabs, OpenOffice Writer, Notepad, Trillian, Winamp5 and Paint.NET) because it reflects how 90% of computing time is spent
  • file search / defragmentation to get an idea of how much noise the hard-drives create
  • watching a DVD to see how much noise the optical drive adds
mCubed HFX mini
Activity
Describtion
Idle
Sitting within 1 meter of the unit you can hear the idle noise of the two hard-drives. Samsung drives are generally known for their very clean and smooth "whoosh" airflow-sound. Due to both drives being mounted in HD silencing boxes you can't hear or feel any vibrations produced by the platters. In a living room setup when you're located more than 3 meters away from the HFX mini you won't be able to hear it when its idle. All in all a very quiet, but not completely silent, experience.
productivity applications
There's hardly any increase in noise-levels compared to the idle mode mentioned above. It's only when the hard-drives go into seek mode - for example upon loading a new playlist in Winamp or opening large OpenOffice douments - that you can clearly hear the system even sitting 3 meters away from the unit.
file search / defragmentation
Often hard-drive seek tends to be the most annoying type of sound. While the seek sounds weren't as sharp and high as the ones coming from my trusted 3-year old Seagate Barracuda IV in one of my other systems, I've become used to the luxury of using 2.5" laptop drives inside most of my systems. These drives produce a much smoother and lower type of seek-sound than the 3.5" desktop drives used inside the HFX mini.
watching a DVD
Not surprisingly the slot-in slimline optical-drive made the biggest contribution to the overall noise level upon loading a new DVD or while copying over data to the hard-drive. Apart from the unavoidable noise created by the moving lens/arm, there was a lot of "whoosh"-airflow which was easily heard 3 meters away from the unit. However once I was watching the movie, the optical-drive became significantly quieter. In fact it became so quiet that it wasn't distinguishable from the rest of the system anymore, no matter how closely I listened to it. The "whoosh" was still there but it didn't seem to add anything to the subjective noise impression. I'd say that in a living-room environment you normally won't hear the optical drive unless you like watching your DVDs with the soundtrack on mute.

All in all, the HFX mini is a very quiet system but I wouldn't call it silent. It may be due to the fact that it's a fanless system with no broadband noise from the fans to help mask it, but I personally didn't expect the hard-drives to be so much of an annoyance. This is a very subjective impression and it is important to realize that I'm testing the system in an environment that's significantly quieter than any normal living room or office. Just to put this into perspective: typing these words on a 2-year old keyboard does make considerably more noise than the HFX mini while it's idle.

CONCLUSIONS

With their HFX mini, mCubed has delivered a very quiet HTPC system with exceptionally good looks. In a living room environment, you are not going to be able to hear anything under most circumstances. In an office environment, the noise produced by the hard-drives will be more noticable as the user tends to sit much closer to the system than in a living room. Seek noise tends to be the most annoying type of sound coming from otherwise silent PCs and the HFX mini is no exception. With productivity applications like the ones mentioned above, these seeks will be present less than 10% of your computing time. If your applications heavily rely on massive I/O from the hard-drives then the seek noise is certainly something you'll be hearing quite a lot.

In the HTPC functions, when watching video or music from the hard drives, the sound is very similar to using productivity applications in that it is very quiet apart from the seeks when new data is loaded from the drives.

There's room for some improvement. I'm sure that using two 3.5" hard-drives is not the best low noise solution. Still, most people will probably use the HFX mini as an HTPC setup where storage demands are endless, and the higher capacity and low cost-per-gigabyte of 3.5" drives are compelling. I wonder what the system would sound like if mCubed had decided to use a 2.5" laptop drive, or maybe even a solid state drive, for the operating system and the most frequently used applications. That would allow the 3.5" drive to spin down when its not being used which would significantly lower the overall noise-level to close to nothing.

In the end, all there's left to do is ask myself whether I would be willing to spend my own hard-earned cash on the HFX mini if I were looking for a very quiet setup. The answer is yes. As an enclosure for a very quiet system, the HFX simply works. With some custom optimisations and a slightly different storage setup than the one inside my review sample, the HFX mini would be very close to the holy grail of combining top-notch performance and cutting-edge features with nearly silent operation.

Many thanks to mCubed for providing us this sample.

*

Articles of Related Interest:

What is a "Silent" Computer?
Shuttle's Smallest Yet: XPC X100
17" Apple iMac - The Official SPCR Review
Fanless Ultra Powerhouse PC by EndPCNoise
The Fanless Hush ATX PC
Mappit A4F: A Truly Silent PC

* * *

Discuss this article in the SPCR Forums.



Previous 1 2 3 4

Cases|Damping - Article Index
Help support this site, buy from one of our affiliate retailers!
Search: