Just to pitch in my 2 cents as usual:
I built a silencer cabinet ~1.5 years ago. My PC was already very very quiet,. I would guess 16~18 dBA. Maybe less. About as quiet as a Panaflo at 7~8V, sitting under my desk 3~4 feet away from my ears. But that was not good enough. I was seeking
silence. You know, zero dBA!
Materials were massive medite board / 1" steel tubing frame plus various grades of damping materials such as carpet underlay, acoustic foam, insulation foam strips (around removable panels). The frame was built first, then the medite board screwed/glued to it. Door in front & back, and left side panel removable with large number of securing screws. It was ~ 28” tall, a foot wide and ~28” deep. Weighed closed to a hundred pounds.
The large left/right side panels were composites of 2 outer layers of ½” medite board & carpet underlay, all silicone glued / screwed. Other panels were a combination of ½”, ¾” and 1” medite board. Various types of damping material lined all the inside walls. (Different materials to try and get effective attenuation of noise at all frequencies. )
Carpet underlay covered the entire bottom of the cabinet so it was somewhat decoupled from the floor. Actually, the weight of the cabinet + PC was well over a hundred pounds, so there was probably too much compression of the carpet underlay for any serious decoupling effect.
Had an acoustically lined tunnel the width & length of the cabinet below the chamber where the tower PC sat. Large vent opening in back, large vent opening in front floor of main cabinet, directly under the front intake vent of PC. Outside air comes in from back bottom of case (at floor level) then goes directly where it is needed most, to the intake vent of the PC.
The interior chamber was deliberately sized so the PC would just squeeze in. This was to ensure that the incoming air would be forced to travel through the PC rather than be wasted just going through the chamber but bypassing the PC. The PC case was prepared by taking extreme measures to open up the vent holes front & back, to minimize impedance. The PC had a 5V Panaflo in the Enermax PSU. The Swiftech MC462A had a 5V Panaflo, too, over a 1G T-bird, undervolted / clocked to 933. Two aluminum sandwiched drives on bottom of PC case on foam: Barracuda IV & IBM 2 platter models.
There was a tunnel above the PC much like the one below. The opening was at the top back of the PC chamber. There was a space about 4~5” between back wall and back of PC, for wires. It created a tunnel for hot air to be exhausted from the PSU and the back panel vents, where it would rise and be drawn up into the tunnel.
The exhaust tunnel was where I made my big mistake.
![Crying or Very sad :cry:](./images/smilies/icon_cry.gif)
Instead of finding a way to duct the exhaust air out back in some kind of short acoustically damped baffle, I split the upper tunnel in 2 – running the depth of the cabinet. So the exhaust air had to come to the front of the cabinet then make a 180 degree turn and run back down the depth of the cabinet. This probably made the exhaust duct impedance too high.
A 92mm fan was positioned inside the cabinet at the entrance to the exhaust duct. It was the quietest one I had at the time, fed off an external DC power supply I made, along with a voltage regulator to adjust the fan speed.
It worked: the noise of the PC was greatly attenuated. Maybe by 6 dBA, maybe 8. But the price was 6~9 degrees increase in temps. Turning the tunnel exhaust fan on & playing with its speed helped a few degrees, but then made the noise level much closer to what it had been with the PC outside the case. Although the heat did not created instability, I was not comfortable with everything slowing baking in there, so I took to turning the fan up to full when I left the room & turning it down/off when I was working.
I abandoned the cabinet when I discovered that suspending the Barracuda gave a me a few dB improvement, and swapped the IBM for a second suspended Barracuda. These changes brought the PC noise level down to where it had been inside the cabinet. The cabinet took up too much space, made it hard to hack on the PC, and the background worry about heat was bothersome, so out to the garage it went.
Would I do it again? Only if I could do it fanless without increasing heat. If you look at the specs on the
http://www.plasmic.dk/cocoon/en/datasheet.php datasheet, you’ll see they show a 4-6C rise in temps inside the cabinet too.
Silencing cabinets do have a useful place where:
1) existing PCs are not TOO loud but very expensive to replace and too difficult to modify
2) drop in quiet convenience is desired
3) absolute silence is necessary: the NORENS no fan cabinets make sense in studios & settings where there must be no extraneous noise at all (anechoic chamber??)
Hope that was worth 2 cents.
![Wink ;)](./images/smilies/icon_wink.gif)