Soundproofing materials and their uses for system noise reduction

[mikeymikec]

My latest plan to reduce the noise level of my system had little effect sadly. I changed the stock AMD Ph2 cooler on my 960T to an Arctic Cooling Freezer 13.

It seemed to have a significant benefit when I had the case open and running on its side, but when I returned it to the normal orientation (tower, right of my desk, side panel on, the rear end near a wall), the noise level seemed the same as before, despite the fact that the CPU fan went from about 1700-1900rpm to about 1300. The chassis fan is below 1000rpm at almost all times and the GPU fan is currently 1050rpm, so I doubt they're doing it. The one I don't have a reading on is my Corsair VX450W PSU.

I'm wondering whether I ought to use soundproofing material next (assuming that I can afford it, I haven't really looked into it). If the material is thin enough, I could put it inside the case on both side panels, and I'm also wondering whether putting some of this material on the wall behind the PC as well as on the side of the desk to the left of the PC.

I'm also wondering whether I'm just spoiling myself, and every small reduction in noise levels just makes me fixate on the remaining noise as being 'too much' My hearing is overly sensitive (IMO) to background noises.

 

[BonzaiDuck]

My latest plan to reduce the noise level of my system had little effect sadly. I changed the stock AMD Ph2 cooler on my 960T to an Arctic Cooling Freezer 13.

It seemed to have a significant benefit when I had the case open and running on its side, but when I returned it to the normal orientation (tower, right of my desk, side panel on, the rear end near a wall), the noise level seemed the same as before, despite the fact that the CPU fan went from about 1700-1900rpm to about 1300. The chassis fan is below 1000rpm at almost all times and the GPU fan is currently 1050rpm, so I doubt they're doing it. The one I don't have a reading on is my Corsair VX450W PSU.

I'm wondering whether I ought to use soundproofing material next (assuming that I can afford it, I haven't really looked into it). If the material is thin enough, I could put it inside the case on both side panels, and I'm also wondering whether putting some of this material on the wall behind the PC as well as on the side of the desk to the left of the PC.

I'm also wondering whether I'm just spoiling myself, and every small reduction in noise levels just makes me fixate on the remaining noise as being 'too much' My hearing is overly sensitive (IMO) to background noises.

I still have a pile of SPIRE noise-deadening foam rubber in my parts locker. It comes in handy from time to time -- in small amounts. That being said, if you put it all over your case interior, it is likely to help but a b**** to remove if you ever want to do that.

Before you go out and order the SPIRE, let me ask: How did you mount the case fans? Order rubber "rivet" fan mounts -- don't use screws. Try and isolate all the case fans from any plastic-to-metal or metal-to-metal contact.

The second thing might as well be the first: The Freezer 13, which I just looked at for reference at the Egg. That sucker sports a 92mm fan! It may actually be a quiet fan -- I can't say. But if I'd ordered a cooler myself -- two things -- I would want any fan hanging on the cooler to be at minimum 120mm and better -- 140mm.

If you can simply build a duct for the cooler to a 120mm rear exhaust fan (which you then plug into the CPU fan port), you might get the same effectiveness for the cooler and eliminate a fan in your case. Here, it would help to pressurize the case and thermally-control the other case fans. You can buy foam-art-board at Michaels Arts & Crafts or Target -- many places -- 2'x3' panels of the stuff in black, white, yellow -- whatever -- for $5. By scoring the paper backing on one side, you can bend it to 90-degree angles. Use an Xacto knife to cut it. You can buy special glue for it. Any number of things.

These are just some ideas, and you can pick and choose. But a possible objective could be to reduce the sheer number of fans in a case while actually increasing CFM airflow. And isolate the fans from direct contact with case-metal -- especially that . . .

 

[mikeymikec]

The chassis fan is screw mounted atm.

As an experiment I took some packaging sponge and propped it between the rear of the computer and the wall, with 2-3" gap (the power cable is keeping it away from the back of the PC), and another bit between the nearest corner to me, the desk and the PC. It might be my imagination (I tend to notice the noise level more in the mornings), but I think it has helped.

 

[BonzaiDuck]

The chassis fan is screw mounted atm.

As an experiment I took some packaging sponge and propped it between the rear of the computer and the wall, with 2-3" gap (the power cable is keeping it away from the back of the PC), and another bit between the nearest corner to me, the desk and the PC. It might be my imagination (I tend to notice the noise level more in the mornings), but I think it has helped.

Yeah! You can do that, too! Since I'm "air-cooled" and have a side-panel fan, I'm quite aware what an advantage it can be to have that panel facing my couch! On the other hand, it's a powerful 200mm fan spinning at 1,300 rpm, and I'm surprised how four 1/2"x3" strips of the SPIRE wrapped around the fan "feet" eliminated all vibration from the side-panel. If I couldn't use rubber-rivets, I used nylon zip wire-ties to secure the fan. Even that seems "crude" and kludge-y, but if they're the same color as the case, you can't see them.

Given your fan rpm's you cite, I was a bit puzzled at your irritation. But the rubber rivets help -- always. Excluding the use of a decibel meter, there's an element of personal tolerance and subjectivity in the noise factor, but always -- bad fans with motor-whine are universally noisy to every listener.

 

[mikeymikec]

Given your fan rpm's you cite, I was a bit puzzled at your irritation.

I think the location of the case makes an enormous difference in terms of how the sound reaches the user. For customers, I've been building the same PCs repeatedly, fan speeds the same, yet in different settings the sound is somewhat different. For example, my HTPC/server is the same spec as most of the other computer builds I've done, yet because it's next to a couch and curtains, it sounds virtually silent. In contrast, the door to my PC room is also near the PC, and on the occasion that I close the door, the noise from the PC is louder, presumably because the sound has another hard surface to be reflected back to me.

Unfortunately I can't think of a better location for my computer in general.

I might save up the packaging sponge I sometimes get and pad my PC out completely

I think I am unusually fussy in this department, I haven't had a single customer say (of any of my PCs in say the last 8 years), "it would be nice if it was quieter", but having said that, most of those are quieter because they're all using integrated graphics whereas mine has a 5770 that likes to occasionally go full-whack in terms of RPM - yes it is free of dust!

 

[BonzaiDuck]

I think the location of the case makes an enormous difference in terms of how the sound reaches the user. For customers, I've been building the same PCs repeatedly, fan speeds the same, yet in different settings the sound is somewhat different. For example, my HTPC/server is the same spec as most of the other computer builds I've done, yet because it's next to a couch and curtains, it sounds virtually silent. In contrast, the door to my PC room is also near the PC, and on the occasion that I close the door, the noise from the PC is louder, presumably because the sound has another hard surface to be reflected back to me.

Unfortunately I can't think of a better location for my computer in general.

I might save up the packaging sponge I sometimes get and pad my PC out completely

I think I am unusually fussy in this department, I haven't had a single customer say (of any of my PCs in say the last 8 years), "it would be nice if it was quieter", but having said that, most of those are quieter because they're all using integrated graphics whereas mine has a 5770 that likes to occasionally go full-whack in terms of RPM - yes it is free of dust!

I can't disagree . . . the current discussion seems to be "water versus air," and people complain about pump noise and fans -- for water-cooling!

I could've had a case that only provides for 140mm fans, but I have HAFs with the 200mm options. I replaced one with an NZXT equivalent -- probably the details can be found among the threads here, but may not be relevant to you.

I can only remember when I built a P4-533FSB system and used an early heatpipe cooler with a 92mm fan (I tried several). That was "loud." I use my current beast as "multi-purpose" and it provides me HTPC functions. There are three computers in this room -- all with HAF midtower cases and the 200mm intake fans. You can "hear" them -- a little air-turbulence, not much of anything in motor-noise. But they don't affect my listening pleasure. . .

 

[Vinwiesel]

Some little things can make a big difference. I used to have my PC on a board to keep it up off the carpet, but the board seemed to amplify noise, so I now have the case sitting on a couple of bricks.

Walls seem to increase noise too, sometimes just angling the computer a little can help, or rotate it 90 degrees so a different side is facing you.

One system only had a single 120mm intake fan, which either had to run fast (noisy) or run slower, and internal fans then had to run fast (noisy). I fixed this by installing a second 120mm fan in the front, using 3 of the 5.25 bays. With 2 slow intake fans, it is a lot quieter.

This system also had a tiny fan on the chipset, and after getting everything else quiet, this little guy seemed ten times louder. I soldered up a fan extension with several diodes in series to drop the voltage to this fan, where it is quiet but still keeps the chipset at around 50C.

Finally, I have my cases lined with Peel & Seal. This is an asphalt roof patch material with an aluminum backing. It is heavy, but it is the mass that reduces noise transmission, and I don't move my cases much. It won't work miracles, but it took my cases from "pretty quiet" to "quiet". There is zero chance of removing it from your case once applied. It is basically tar. It is surprisingly non-stinky due to the aluminum backing.

One final note, If all else is quiet but the video card screams during gaming, disable some options you don't care about, like fancy shadows, specular lighting, bloom, AA, etc. V-sync works well in some games, (or terrible in some) but can reduce the gpu workload a lot.

 

[doyll49]

My experience is the best way to have a quiet case is quiet components and good vibration control of mechanical components like HDDs, fans. Cases are full of vents for sound to come out through.

A quiet CPU cooler, quiet GPU and all fans on automatic temperature/load controlled speed is paramount. With case fans to component fan air ratio of 1.5-2.0 : 1.0 and airflow structured to expel heated component exhaust out of case without any of it mixing with cool intake air going to components a system is virtually silent below approx. 900rpm (can't hear it a meter away), very quiet to approx 1100rpm (must concentrate to hear it) and quiet to approx 1300rpm (know it running).

My i7 980 on X58A-UD5 under TC14PE (putting CryOrig R1 Ultimate on today), Asus ENGTX580 DCUII, in Define R2 on 40mm castor base w/ back, bottom & front vent grills removed, front vents modified to accept 140mm fans, 3x TY-140 fans w/Silverstone filter/grills and no exhaust fans at idle CPU is 24-29c@650-750rpm cooler, 600-700rpm case; 100% load CPU is 43-49c@900-1050rpm in 22c room. GPU is about 10c warmer and idles silently at 750-850rpm, full load is 1400-1700rpm... quiet, but loudest component in system.

Quiet enough that in a room you can hear wind rustling the leave on tree outside I have to concetrate to hear it under load.. have to look at case power light to tell it's running during normal use.

 

[coffeejunkee]

Motherboard? Case? According to Arctic site Freezer 13 should go down to 600rpm. A 92mm fan at 600rpm is pretty quiet. 1000rpm for the gpu sounds right too, PSU is Seasonic made, usually very quiet.

Leaves the casefans, I run mine at 300/400rpm and can still hear them. So how much below 1000rpm exactly?

 

[doyll49]

92mm fan at 600rpm will not be moving much air. I would keep the idle speed at least 800-900rpm, maybe 1000rpm. Dropping my TY-140 CPU fans from 800rpm to 650rpm increases mobo temps by 5-8c easily. And TY-140 moves more air at 650rpm than a 92mm fan.

 

[coffeejunkee]

No it won't, but it's not really necessary. A 960T doesn't powergate as well as newer cpu's but with CnQ enabled it should stay relatively cool during idle. So what if it idles 5C higher?

My point is: if the motherboard is running the fan at 1000rpm instead of the normal 600rpm the cpu fan curve isn't set optimally for silence.

 

[doyll49]

No problem for CPU.
Fan spec is 600-2000rpm but I would be surprised if motherboard default fan setting has it idling at 600rpm.
Without hearing the fan it is impossible to know how much noise it makes, but would guess is in the 900-1000rpm range it isn't much.

 

[mikeymikec]

Asus M4A89GTD PRO/USB3, Qfan standard profile (I'm considering switching it to silent, though I would be interested to see what the load temps are like then). 12cm chassis fan that comes with the case as well, set to standard profile, normally spins at about 800rpm.

Coolermaster Elite 330 (The air inlets on the side panel are covered up with sponge, not the two bits of sponge I mentioned earlier)

Coming back to the computer after half an hour of idle, the CPU fan is at 1100pm, and CPU temp at 31C.

 

[doyll49]

How does it sound now?

 

[mikeymikec]

TBH I'm pretty sure that the silent profile increases fan RPM rather than decreases it! I'll keep an eye on it for 24hr.

 

[coffeejunkee]

Well, I would suggest to use Speedfan to competely turn down cpu fan and casefan to see which one creates most noise. Turning off psu fan or gpu fan is not possible so that will be your baseline noise.

Maybe adding a front fan will work too, better airflow could allow all fans to spin slower (except psu and gpu again, I'm pretty sure they are already at their minimal rpm).

If still noisy you could start looking at sounddampening material. I have no experience but my thinking is, you can't block in or outtake fans so you will always have air moving which by definition creates sound. I just wouldn't expect wonders.

 

[mikeymikec]

However, I suspect that a lot of 'fan noise' sound is actually echoing/reflected sound. Reduce the reflection, reduce the sound perceived? Dunno.

I haven't used SpeedFan for fan management before, admittedly I'd probably need to switch off QFan first, but does SF has the sort of fan management that MSI AfterBurner has?

 

[BonzaiDuck]

However, I suspect that a lot of 'fan noise' sound is actually echoing/reflected sound. Reduce the reflection, reduce the sound perceived? Dunno.

I haven't used SpeedFan for fan management before, admittedly I'd probably need to switch off QFan first, but does SF has the sort of fan management that MSI AfterBurner has?

Just my two-cents worth. SpeedFAn is a great tool, and I used it to control the CPU fan on an old Springdale ASUS motherboard. But if the board manufacturer provides BIOS features and software that do essentially the same thing, I'd stick with those unless they give you trouble. You might still use Speedfan as a diagnostic tool, even so . .. Ultimately, if you're satisfied with the control provided by SpeedFan, go for it . .

 

[coffeejunkee]

However, I suspect that a lot of 'fan noise' sound is actually echoing/reflected sound. Reduce the reflection, reduce the sound perceived? Dunno.

I haven't used SpeedFan for fan management before, admittedly I'd probably need to switch off QFan first, but does SF has the sort of fan management that MSI AfterBurner has?

You can set a fancurve in Speedfan but it's a bit complicated. I just set min and max fanspeed and desired temps and the program varies fanspeed automatically according to those values. Works good enough for me. The main reason I use it is because my mobo limits fancontrol to a rather narrow range. I can't set fanspeed lower than 60% in bios, with Speedfan I can set it to 20%. 60% is noisy, 20% is quiet. No need to disable QFan either, Speedfan overrides bios control.

Just my two-cents worth. SpeedFAn is a great tool, and I used it to control the CPU fan on an old Springdale ASUS motherboard. But if the board manufacturer provides BIOS features and software that do essentially the same thing, I'd stick with those unless they give you trouble. You might still use Speedfan as a diagnostic tool, even so . .. Ultimately, if you're satisfied with the control provided by SpeedFan, go for it . .

Well yes, in an ideal world bios fancontrol would let us have a quiet pc during idle and light usage while gently ramping up fanspeeds under load. In the real world however...

Also, Speedfan lets you control fanspeeds based on any components temperature while bios looks at cpu temp only.

 

[BonzaiDuck]

You can set a fancurve in Speedfan but it's a bit complicated. I just set min and max fanspeed and desired temps and the program varies fanspeed automatically according to those values. Works good enough for me. The main reason I use it is because my mobo limits fancontrol to a rather narrow range. I can't set fanspeed lower than 60% in bios, with Speedfan I can set it to 20%. 60% is noisy, 20% is quiet. No need to disable QFan either, Speedfan overrides bios control.



Well yes, in an ideal world bios fancontrol would let us have a quiet pc during idle and light usage while gently ramping up fanspeeds under load. In the real world however...

Also, Speedfan lets you control fanspeeds based on any components temperature while bios looks at cpu temp only.

No disagreement in principle there. There was probably a time when other components deserved more attention. Not so much now. But it still points up Comparetti's attention to the details of his software, you can do more with it than you can with what you get in the mobo package.

My own obsession is to reduce the potentially-complex to the elegantly simple. Modders and DIY'ers here may not spare the expense of fan-controllers with their own processor and USB communication with the motherboard. Indeed, these things are like the extras you might get buying a top-end luxury car. But it's also nice if you can avoid it altogether and get the same essential thermal control, cooling and noise reduction.

IF you dump the software provided by the mobo, and IF Comparetti has added your board and chipset to his vast accommodations in Speedfan (or the various sensors used in mobos across the market), you have something like your Victorinox camping knife. You won't use all of it, but you'll always have the parts you want to use in this or that situation.

On the other hand, if the mobo software fills the bill for a simpler configuration and you only "need" the CPU temperature to do it, nothing lost there either.