Antec Fusion Remote Max HTPC case

Viewing page 6 of 7 pages. Previous 1 2 3 4 5 6 7 Next

TEST RESULTS

The thermal testing was done in a 20' x 10' room where ambient conditions were were 20 dBA and 22°C throughout testing. Acoustic testing was done in the live room as well as in SPCR's anechoic chamber, where the ambient level was 11 dBA. Idle measurements were taken 5~15 minutes after boot or reboot, after the internal temperature remained unchanged for several minutes. Load measurements were taken after at least 15 minutes of full load — again enough time for the temperature to stabilize. There were no differences greater than about 0.5 dBA in measured SPL between the two test environments. The noise of the system was far enough about the ambient to be measured accurately in either room

1) BASE CONFIGURATION

The overall sound was smooth and subdued, a broadband noise marked mostly by the airflow turbulence of the 140mm exhaust fan. The PSU and video card fans seemed to contribute very little to the overall sound.

At close proximity, there was some low frequency emphasis that went away when the cover was removed. It is not panel vibration; it is the air within. This is common to all cases, and it is the result of cavity resonance: The air resonances in the case amplify any noise inside that happens to be near its resonant frequencies. Applying sound damping on the underside of the top panel could help but it's almost impossible to eliminate entirely. The good thing is that the lower the overall noise of the components, the lower this effect will be. Conversely, the higher the overall noise, the less you'll hear the effect because it'll be drowned out by the more obvious noises. You can hear the effect of the top cover being on or off in one of the recordings.

Base Config (stock 14cm fan @ low)
Load	AC Power	CPU °C	Board °C	GPU °C	SPL dBA@1m
Idle	152~155W	40	48	55	25
CPUBurn x4	225~227W	57	54	56	26
ATITool	222~237W	43	51	83	26
ATITool + CPUBurn	288~295W	57	56	83	26
3DMark06 + CPUBurn	245W~270W 285W peaks	57	56	81	26
Play Blue-ray movie	133W~148W	43	51	62	26

No combination of loads increased the measured SPL by more than 1 dBA, although some increase in tonality could be heard on tests that loaded the GPU. The stock ATI cooler on the HD4870 spun up at least a bit (from the minimum of ~830rpm to a maximum of about 1600rpm), and this lead to a subjective increase in noise. The PSU fan might have sped up a bit in the highest load tests, but this was difficult to confirm or hear. The overall noise was subjectively constant throughout all the tests.

The AMD Cool'n'Quiet dynamic under-clocking / under-volting utility for the CPU was confirmed to be working automatically, as it should in Vista.

The power draw measured on this system was close to the highest ever for any system at SPCR, the one exception being the recent test systems we assembled around the Intel Core i7 processors. Curiously, the power consumed while playing a Blu-ray disk was lower than at idle. This was seen repeatedly; we have no clear idea why it is so. The thermal data while playing Blu-ray conformed better to our expectations — it was slightly higher than at idle.

The cooling was good on all counts. The temperature on the PSU was not monitored; its exhaust felt fairly warm at the highest loads, but its fan never ramped up from default... or if it did, the effect was so subtle that it was not noticed or measured.

Crossfire Configuration

Adding the second video card impinged drastically into the intake of the fan for the first card. The issue is spacing, as the photos below show.

While the second video card has plenty of space for its intake fan, the intake of the first card is drastically reduced. In anticipation of thermal issues, a 120mm fan was added at the front bottom intake. It is a product from a new company, GELID Solutions Wing12 rated at 1500rpm, and 25 dBA@1m. It sounds quieter than the spec suggests, and has a smooth unobtrusive character. The fan mounted in the supplied frame. It was controlled by the motherboard: ~1100 rpm at idle and up to 1600rpm at full load.

It's obvious that a motherboard for dual video cards should have at least two slots between the PCIe-16x slots. Then, two high performance cards — which invariably take up two slots — would have at least one slot between them, or close to an inch. The choice of board here was not ideal... but you can be assured that your smarter motherboard choice will help to achieve lower noise than we did here.

Adding the second HD 4870 card pushed idle power up by some 70W. When the GPUs were loaded, the increase in power over the single GPU system was over 100W. This was the same increase seen when both the CPUs and graphics cards were fully loaded.

Yet, in idle, the noise increase was very subtle, measuring perhaps 1 dBA at most. Subjectively, it seemed a touch more intrusive, probably because the fan on one video card (which have the smallest fans and therefore the highest pitch of any sounds in the system) was always spinning faster.

Crossfire Config (14cm fan @ low, front 12cm fan 1100~1600rpm)
Load	AC Power	CPU °C	Board °C	GPU °C	SPL - dBA@1m
idle	218~225W	41	57	56/75	26
CPUBurn x4	300W	57	57	56/75	31
ATITool	320~333W	45	61	58/85	30
ATITool + CPUBurn	369~376W	57	68	79/86	33
3DMark06 + CPUBurn	310~340W 378W peaks	57	65	76/84	32
Play Blue-ray movie	208~221W	43	57	77/78	27

Almost all the noise output is well below 1kHz. The dominating sonic signature is that of the 140mm stock fan. It is a relatively benign broadband sound. With one 4870 video card, the overall SPL dropped by about a single decibel. Taking the cover off reduces the cavity resonance effects between 100Hz and 300Hz. Although the measured SPL is only about 1 dBA, it is plainly audible in the accompanying sound recording below. It's actually a more significant difference at idle than going from one 4870 card to two.

As the system was subjected to higher load, the effect of the second video card became quite audible, and its thermal effect could also be seen in the GPU temperature. As the table above shows, one of the two GPUs always ran substantially hotter than the other, and thus, its fan ran substantially faster. This was obviously the card whose intake fan was so blocked. Its fan speed never dropped below about 1200rpm, which contributed a bit of tonality to the sound. There were some other increases in temperatures, but overall, they were fairly minor.

As the thermal load increased, the fans in the system sped up to compensate. This effect was quite audible; at the high loads the SPL reached 30 dBA and beyond. All the fans ramped up a bit except the 140mm stock case fan, which was fixed at low speed.

Previous 1 2 3 4 5 6 7 Next

Cases|Damping - Article Index

Help support this site, buy the Antec Fusion Remote Max HTPC case from one of our affiliate retailers!