Micro-ATX Cases - Shoebox Showdown
by Joshua Buss on January 2, 2007 12:40 PM EST- Posted in
- Cases/Cooling/PSUs
Benchmark Comparisons - Sound
Next up, we took our new test bed and re-worked our sound tests a little to better categorize both the levels and characteristics of the noise produced by each case with the accompanying test hardware installed and operational. No case won or lost overall by a clear margin, but there were still distinct victories in smaller areas.
(Sound tests conducted in an ~18dB ambient level environment)
First, for a better frame of reference, we assembled our test bed on the open work bench with our Zalman power supply far enough away that its main fan didn't even turn on. Both of the Zalman coolers - the CNPS-7000 Cu on the CPU and the VF-700 Cu on the GPU - were given 5 volts instead of 12 to keep their levels acceptable for desktop use in a near-silent environment. This open air test bed proved to be nearly silent, registering a record low from the CPU side of the computer of 23 decibels.
An open air computer is hardly practical though, and a good case should actually be able to help bring already quiet levels like these down further with just the right amount of well-designed sound insulation, and much to our surprise, the inexpensive MicroFly did the best job of this! Until we did our loading test, the Ultra case had by far the best characteristics of its sound. Subjectively, it was no different than the open air setup, a very prestigious claim to say the least. When kicked into high-gear unfortunately, the case began to sing a new tune as the included power supply's fan went into its other speed zone, creating a noticeable motor bearing noise along with the whooshing sound of moving air. 3.5 is still a good rating for any desktop PC case (remember that these are 1-10 ratings), but if the environment was quiet enough, a highly-stressed MicroFly might be annoying enough for some people to regret not getting a quieter solution. We did try putting both system fans on 12 volts to see if the power supply would stop going into high-gear, but it still did and then the case was even louder at idle.
The Aria also did a great job of staying nearly as quiet as the open setup, but its power supply fan had just a hint of bearing noise that, thankfully, blended with the air sound from the blower on 5V to get mostly lost in the overall sound signature. For a case as warm as the Aria, it might have helped a lot to allow for a few more active cooling elements, or at least include a power supply that can increase the fan RPMs at the higher temperatures we measured.
PC Design Lab's Qmicra performed admirably with the Zalman supply and single Tri-Cool fan, but when all four fans were installed (and even only running on five volts each) the sound of moving air was certainly noticeable, both to our ears and the dB meter. Strangely enough, the sound-dampening material obviously worked for the dB meter, as the install with the CNPS 7000 cooler registered lower levels than the open-air bench both from in front and the top, but we still noticed the slight hint of extra moving air and motor noise from the single Antec Tri-Cool 80mm fan. We actually speculated that this might be because the good performance of the sound insulation material would make the Tri-Cool (the only active fan immediately next to an open grill) more noticeable without the other fans being heard, and this theory would also explain the almost identical measured levels.
Before wrapping up, it should be reinforced that all of these setups were comfortably quiet - perfectly suitable for use in 90% of computing situations, and a couple even capable of handling the tricky task of working in an audio-sensitive environment, particularly if a high-powered portable machine is desired. Compared to smaller cases we've looked at before and a few famously noisy Shuttle computers, this is a welcome change for the SFF industry.
Next up, we took our new test bed and re-worked our sound tests a little to better categorize both the levels and characteristics of the noise produced by each case with the accompanying test hardware installed and operational. No case won or lost overall by a clear margin, but there were still distinct victories in smaller areas.
(Sound tests conducted in an ~18dB ambient level environment)
First, for a better frame of reference, we assembled our test bed on the open work bench with our Zalman power supply far enough away that its main fan didn't even turn on. Both of the Zalman coolers - the CNPS-7000 Cu on the CPU and the VF-700 Cu on the GPU - were given 5 volts instead of 12 to keep their levels acceptable for desktop use in a near-silent environment. This open air test bed proved to be nearly silent, registering a record low from the CPU side of the computer of 23 decibels.
An open air computer is hardly practical though, and a good case should actually be able to help bring already quiet levels like these down further with just the right amount of well-designed sound insulation, and much to our surprise, the inexpensive MicroFly did the best job of this! Until we did our loading test, the Ultra case had by far the best characteristics of its sound. Subjectively, it was no different than the open air setup, a very prestigious claim to say the least. When kicked into high-gear unfortunately, the case began to sing a new tune as the included power supply's fan went into its other speed zone, creating a noticeable motor bearing noise along with the whooshing sound of moving air. 3.5 is still a good rating for any desktop PC case (remember that these are 1-10 ratings), but if the environment was quiet enough, a highly-stressed MicroFly might be annoying enough for some people to regret not getting a quieter solution. We did try putting both system fans on 12 volts to see if the power supply would stop going into high-gear, but it still did and then the case was even louder at idle.
The Aria also did a great job of staying nearly as quiet as the open setup, but its power supply fan had just a hint of bearing noise that, thankfully, blended with the air sound from the blower on 5V to get mostly lost in the overall sound signature. For a case as warm as the Aria, it might have helped a lot to allow for a few more active cooling elements, or at least include a power supply that can increase the fan RPMs at the higher temperatures we measured.
PC Design Lab's Qmicra performed admirably with the Zalman supply and single Tri-Cool fan, but when all four fans were installed (and even only running on five volts each) the sound of moving air was certainly noticeable, both to our ears and the dB meter. Strangely enough, the sound-dampening material obviously worked for the dB meter, as the install with the CNPS 7000 cooler registered lower levels than the open-air bench both from in front and the top, but we still noticed the slight hint of extra moving air and motor noise from the single Antec Tri-Cool 80mm fan. We actually speculated that this might be because the good performance of the sound insulation material would make the Tri-Cool (the only active fan immediately next to an open grill) more noticeable without the other fans being heard, and this theory would also explain the almost identical measured levels.
Before wrapping up, it should be reinforced that all of these setups were comfortably quiet - perfectly suitable for use in 90% of computing situations, and a couple even capable of handling the tricky task of working in an audio-sensitive environment, particularly if a high-powered portable machine is desired. Compared to smaller cases we've looked at before and a few famously noisy Shuttle computers, this is a welcome change for the SFF industry.
Facebook
Twitter
RSS
37 Comments
View All Comments
warthogism - Thursday, February 1, 2007 - link
I know that you guys were able to fit an CNPS7000 in there and that a 9000 would not fit. What I was wondering is if a CNPS7700 would fit. The difference is that the height of the 7000 is 62mm and the 7700 is 67mm. Also, the 7000 has a 92mm but the 7700 uses a 120mm fan. Does anyone have first-hand knowledge? or at least, does it look like the 7700 will fit? Thanks for the help.gool - Saturday, January 13, 2007 - link
I'm the happy owner of a heavily modded Antec Aria - blow holes, sidefans, PSU-fan replacement and so on. It's based on a MSI Micro-ATX 939 motherboard, AMD 64-3200 (Clawhammer), 1 GB DDR-SDRAM, Liteon DVD-CD combo, a 160GB 7200 Maxtor disc, an external (USB2) Maxtor 250 GB disc and a Nvidia 6600GT card. The latter is the one I wish to replace, but I'm uncertain if I'm able to draw enough juice from the standard Antec 300W psu for this purpose? Seen people with a similar setup who uses a 7800GT card without problem. What are you guys advice - could I go down this road?Cheers,
Svend
Patrese - Sunday, January 7, 2007 - link
Great review, but I'd like to see something about the new (and not no new) uATX motherboards on the market, specially the Asus M2NPV-VM and Abit NF-M2 NView...artifex - Thursday, January 4, 2007 - link
I suggest everyone interested in these smaller form factors to look at sfftech.com and mini-itx.com. Plenty of reviews and some nifty case mods, etc.mino - Thursday, January 4, 2007 - link
I like that you made semi-pro noise tests for the cases.However as changing noisy PSU/fan is the easiest thing to do(I have yet to us a case with included PSU/fan)..
It would be nice to complement out-of-the-box testing with controlled-enviroment testing.
It would be more informative, especially to enthusiasts, to make a reference rig (fans, PSU, all coolers, HDD) and assemble it into different cases.
To know that a case comes with a noisy PSU is nice. However to know its ability to dampen the brum of an HDD, or mute the pinch of a GPU cooler is far better.
Nice work otherwise, keep on.
JoshuaBuss - Thursday, January 4, 2007 - link
This is a very good idea, but to a certain extent we believe it's the manufacturer's responsibility to supply a case with fans that work best for its need, not the user's. At the least a manufacturer should offer recommended buys (like PDCL did) so that a user knows what should work well. If a fans that come with a regular case are so bad that it's truly hurting the overall appeal, we normally do try it with a different configuration to see if we can do better. With these micro ATX cases it's often the case that the included power supply is a custom fit and we have to test it 'as is'.Thanks for the suggestion.
mino - Thursday, January 4, 2007 - link
I suggest you take a look at the ASUS TM-21/23/25 series especially the TM-250.I have yet to see a case of this size with better internal organization, such a good cooling performance. As a bonus the case is very sturdy (especially for such a low weight product).
The ability to put 3 HDD's in a 15.6*6.7*14inch box _and_ cool them without additional fans speaks for itself.
Newegg lists them for $50 with 300W high-quality PSU's.
However we bought 100 a month ago in EU and they came with 350W and _without_ that annoying floppy hole...
mino - Thursday, January 4, 2007 - link
One little problem I missed, those on Newegg(provided they sell the exact model as on image) have ATX 1.3 PSU's.While these Bestec PSU units are pretty good - PFC, very quiet(have 20+ of them deployed) - 180W on 12V rail may not be sufficient for a gaming rig.
I hope the new refresh models(350W,no floppy), we bought recently, will hit the US market soon.
artifex - Thursday, January 4, 2007 - link
I'd like to see a review of the power supplies used in some of these, as well as third-party replacements. Heck, I'd even like to see a new version of the old ATX PS roundup/review Anandtech did years ago -- I bought my Antec TruePower 430 after that, and it's lasted through a second system, but I've no doubt it's getting creaky now.IronChefMoto - Wednesday, January 3, 2007 - link
First off, decent article. I would like to have seen Anandtech test different video cards in the cases, though. One of the issues with nice mATX cases like the Lian Li V300 (???) is that longer, high-level chipset cards like the 7900 series hit pieces of the framework in the case. You have to mod your case or your card for the Lian Li model.While folks won't always be dropping a 7950GT into one of these, having that option is nice if you want to build a small, high-end gaming rig. Please try and make that part of your mATX review setup in the future. I've even asked Newegg.com to start adding card length data to their specs (they're opening the cards anyway for product shots) so you don't end up having to send back a card that doesn't physically fit a case by 1/4".
That said, my friend and I built v1 Aria (when it had more front ports) about 2 years ago. It was the toughest build I've ever had to do. It was cramped, the PSU had to come out for assembly, and the CPU cooler was up next to the PSU. The temps idled 60C with a very hot P4 Prescott (???) CPU. My buddy upgraded the unit with a Zalman Reserator and continues to use it today running at about 35C under load.
I would recommend that anyone looking at an mATX case check to see that the PSU is STANDARD ATX or a replaceable mATX format. The v1 Aria will not accept an ATX replacement without some modification and/or loss of space. This holds true for mATX barebones like Shuttle XPCs (I own/am stuck with two (2)).
IronChefMorimoto