Antec NSK2400 / Fusion Media PC Case

Cases|Damping
Viewing page 6 of 9 pages. Previous 1 2 3 4 5 6 7 8 9 Next

TEST RESULTS

Ambient conditions in the 20' x 10' room were 20 dBA and 22°C throughout testing. Idle measurements were taken 5~15 minutes after boot or reboot, whenever none of the temperatures had changed for 5 minutes. Load measurements were taken after >20 minutes of full load.

System Configuration #1

Baseline Config #1 (one stock 120 fan @ low)
Load
AC
CPU
Board
Hard Drive
SPL
idle
45W
32°C
35°C
34°C
28 dBA@1m
Rthdribl
88W
43°C
36°C
34°C
CPUBurn x2
118W
53°C
37°C
34°C

The first reaction upon hearing the system was, "This is too loud for SPCR!" But, on the other hand, perhaps only when the mute button is pressed.

The noise came from the stock Antec 120mm TriCool fan (even on low), the Zalman 7000 fan (at 5V), the PSU fan, and the hard drive. At close proximity, there was also some low frequency noise that actually went away when the cover was removed. This means it did not come from the case cover vibrating; rather, it's the air resonances in the case subtly excited by the various noise makers.

It's odd, but perhaps as a result of this resonant effect, the TriCool 120 seems to sound a bit louder in this case than in either the P150 or P180. We tried three different TriCool fans to make sure it wasn't a simple case of sample variance.

The cooling was excellent on all counts. The hard drive was very well cooled; it ran >10°C lower than on the desktop in free air. The temperature on the PSU was not monitored, but its exhaust never felt more than moderately warm, and its fan never ramped up from default.

System Configuration #2

The obvious change was to swap out the stock 120mm TriCool fan for a Nexus 120 and run it at 7V.

Config #2 (one Nexus 120 fan @7V)
Load
AC
CPU
Board
Hard Drive
SPL
idle
45W
32°C
35°C
34°C
24 dBA@1m
Rthdribl
88W
45°C
38°C
34°C
CPUBurn x2
118W
57°C
40°C
34°C

The difference in perceived noise was quite dramatic. This was much quieter than Config #1.

There were some small increases in temperatures, but overall, they were quite minor. The low frequency noise mentioned earlier was pretty much gone. If the system was under my desk as a tower-style PC, the noise would be quite acceptable. Seek noise was somewhat more audible than before due to the lower overall noise, but it was very minor. The limitations to further noise reduction was now the power supply.

A quick experiment consisting of turning all the fans off briefly except for the PSU fan revealed that in the NSK2400 case, the noise of the PSU from the front is somewhat lower than when the PSU was on the test bench. The fan points away from the sound level meter. It measured ~23 dBA@1m. Without changes in the PSU, this is the lowest noise level we can expect.

System Configuration #3

This time, we went back to the stock TriCool fan — not one, but both of them. Everything else stayed the same.

Config #3 (both stock 120 fans @ low)
Load
AC
CPU
Board
Hard Drive
SPL
idle
46W
32°C
33°C
28°C
29 dBA@1m
CPUBurn x2
119W
49°C
36°C
28°C

The noise level went up only a bit compared to one fan, in both measured and subjective terms. A decibel is difficult to hear. When you get right up to the box, especially on the right side where the fans are, the increase in noise seems bigger, but from a meter away, it's quite small. Notice how all the temperatures dropped, especially the CPU and the hard drive. The HDD benefits greatly from the intake airflow created by the fan closest to it.

System Configuration #4

How about when a real video card is used instead of the onboard graphics? The Radeon 200 onboard graphics chip does a very good job, but lots of people will want more. We chose the Asus EAX1600XT Silent/TVD/256M fanless card that survived the two hour thermal torture chamber for our recent review. This is an ideal HTPC card with good performance and lots of features, including a breakout I/O box for composite and S-Video signals. We estimated its power demand to be 18W in idle and 46W at full load.


With fanless Asus EAX1600XT video card.

Sharp eyed readers might notice the business card taped to the plastic airflow baffle on the back panel. It was intended to block off the 1.5~2" of space that's open below the baffle so that more of the airflow from the back vent would make it to the HSF before getting blown out the side fans. We're still not sure whether this had any effect.

Config #4 (Asus 1600XT vidcard, both stock 120 fans @ low)
Load
AC
CPU
Board
Hard Drive
SPL
idle
69W
32°C
33°C
28°C
29 dBA@1m
Rthdribl
124W
42°C
36°C
28°C
CPUBurn x2
142W
49°C
36°C
28°C
Rthdribl + CPUBurn x2
161W
53°C
39°C
28°C

Unfortunately, the Asus card does not have a thermal sensor, so we cannot report its temperatures. It never misbehaved in any way, showing no sign of overheating whatsoever throughout the testing, which included >30 minutes of Rthdribl.

The AC power draw of the system jumped by over 20W at idle, and the load with Rithdribl went up 36W. So the system had considerably more heat to deal with. There was no change in PSU fan speed.

Surprisingly, there was also no change at all in any of the temperatures, which suggests that the system could be run quieter without risk — either with one fan or two slower quieter fans. Did we try that? No, we decided to leave that delight to you and upped the ante instead.



Previous 1 2 3 4 5 6 7 8 9 Next

Cases|Damping - Article Index
Help support this site, buy the Antec NSK2480 media case from one of our affiliate retailers!