Do decibels add up linearly?

[GeneralOreo]

Say you have a tunnel with 5 fans in it, you've got each fan's noise rating. To determine the overall amount of noise - do you just add them up?

An ~8 decibel fan might be so low as to merge with the background, but what about 2 or 3 8 decibel fans?

 

[DrMrLordX]

Last time I saw any generalized answer to a question like this, the answer was that you can add about 3 db to the total fan noise per extra fan.

So if you have 5 8db fans, you're looking at ~20db.

This is just a generalization and is not entirely accurate. It doesn't necessarily hold true when mixing fans of differing noise output levels.

 

[BonzaiDuck]

DrMrLordX -- Thir!!

You must mean an absolute 3 dB, as opposed to a possible implication of a 60% linear ratio.

I've got four 31 dB Sharkoons and two 32 dB Sanyo-Denkis -- the former are intake and the latter exhaust. I'm DEF-initely not deaf, but I don't hear shee-ut coming from my case.

[Of course, the Sharkoons are distributed so that two are facing the carpeted floor and the other two are in the middle of the case.]

 

[Zepper]

Brief answer: No. Each extra identical fan adds a small increment. Someone posted a formula for calculating additive dB, but the value for each of your fans would have to be based on the same system of measurement (and they definitely aren't) to get a reliable end number.

.bh.

 

[herm0016]

the db scale is log. so to double the loudness you add about 3 db i believe.

 

[Billb2]

Yes, 3db = twice as loud. It's logrithmic.

http://www.ieer.org/clssroom/log.html

 

[zagood]

Calculator (4 sources):

http://www.snapfour.com/CombinedNoise_Calculations.aspx

-z

 

[Sheninat0r]

They're right, dB is log, and +3 dB is twice as loud. However, most fan noise measurements are A-weighted, and measured in dBA, which changes things quite a bit. Someone over at SPCR did a test of adding fans, and most of the time it wasn't a simple +3 dBA increase by adding another fan - it varied from nothing up to the expected 3 dBA.

http://www.silentpcreview.com/.../viewtopic.php?t=19297

 

[Zepper]

+3dB is twice the sound pressure, but is about the minimum change that the human ear can detect as a change. +10dB is noticed as a doubling of the sound level to the ear. Good to see someone linked the formulas again. But again, if the mfr's reported dB wasn't measured the same way, you can't use their numbers and get accurate results. Perhaps close enough for horseshoes.

.bh.

 

[DrMrLordX]

Originally posted by: BonzaiDuck
DrMrLordX -- Thir!!

You must mean an absolute 3 dB, as opposed to a possible implication of a 60% linear ratio.

Yes. And it's good that other folks here pasted links to a calculator so the OP can get his answer.

+3 dB per fan is a fair estimate for the overly pessimistic at the very least. I overshot by about 5 dB.

 

[jonmcc33]

Originally posted by: DrMrLordX
Yes. And it's good that other folks here pasted links to a calculator so the OP can get his answer.

+3 dB per fan is a fair estimate for the overly pessimistic at the very least. I overshot by about 5 dB.

More than that I'm afraid when you're talking about something that's near silent to begin with.

 

[BonzaiDuck]

I'm guessing that the rules of thumb persist under some simple, lab-test condtions.

That's why -- if I have fans operating in an "acceptable" range like 32 dBA (according to their individual and possibly inconsistent manufacturer measurement) -- I've spent more time on deploying fans for both maximum airflow and minimum noise. So I put two intake fans near the center of the case, two more on the floor-side/bottom. My internal motherboard, CPU and VGA ducting again reduces noise -- for both choice of materials and coverage averaging 50% of noise sources from either side of the fans.

I should probably get a "noise-meter" so I can measure the effective noise level at so many inches or feet from various sides of the case, but the noise is so low (to my ear) that it just doesn't seem like a worthwhile pursuit -- knowing that my airflow is near-optimal, anyway, and that I can improve airflow even more with the same number of fans, no additional noise, and an afternoon's-worth of Xacto-knife tedium.

Even after those improvements, I can still reduce noise-level further by placing Spire or Akasa padding inside my foamboard-and-Lexan ducts where it cannot be seen. The way the ducts are designed, I can reduce the static volume of air within the ducts and still get better airflow.