General > General Technical Chat
Are here some acoustic experts? Chamber concentrating sound to microphone.
<< < (2/2)
coppercone2:
the best you can do is get as close as possible to find noisy components IMO for a engineering precompliance

trying to do it at any sort of distance is cost and room prohibitive. you would get it tested later at a chamber, its like RF, you use near field, even though sound does not have a near field, you pretend it does
Miyuki:
I was thinking about something like this sketch
I have an available underground room with low noise levels. It's in rock outside of the city. Where I can place a test setup covered with some sound absorbent.
coppercone2:
I am not sure that will help you determine how loud it actually sounds when you put the product on a real thing like a table
Nominal Animal:
Standard anechoic chambers cover the walls with uneven foam structures, to disperse sound waves.  The idea is that the soft foam absorbs most of the pressure waves, and the rest that is reflected, is reflected in different directions.  If I've understood correctly, it can be difficult to place the foam structures without causing "resonances", where some frequencies are not dampened as much as others, because of how they happen to bounce between the foam structures in opposing walls; that it is a bit of a hand-crafted art to make anechoic chambers.  Place foam, measure, replace, measure, etc.

I could have understood wrong, though; I've never made any anechoic chambers myself.

Personally, I would make the chamber ellipse-shaped (as viewed from top), with the equipment to be measured placed in one focal point, and a cylinder of fuzz at the other focal point, and cover the walls with standard sound-absorbing pyramid foam.

In an ellipse, all sound generated at one focus will reflect to the other focus.  (There are buildings that use this as an acoustic effect; a whisper in one focus can be heard in the other focus, even if the distance is much greater than one would think a whisper could carry.)

The cylinder of fuzz is basically an omnidirectional sound absorber.  Whatever sound it reflects, will be reflected back to the device, so the cylinder of fuzz would need to be as absorbent as possible.  I'm thinking of something like a 4" diameter foam cylinder covered with a thick shag carpet.

The microphone would be placed at exactly 1m away from the equipment, as per standard.  Actually, I'd use several microphones, so that the directionality of the noise could also be analyzed.  If the equipment is typically used on a table, it would be on a table in the measurement chamber too.  If handheld, then on a thin pedestal at roughly 1m height.

Essentially, instead of trying to make the walls uneven, I'd be assuming the opposite, and focusing the sound reflections to a dedicated dampener.  Whatever covers the walls will hopefully absorb most of the acoustic energy, but some of it will always reflect.  Having those reflections meet at a dedicated dampener, from which they're reflected back to the walls, should – I think and hope! – yield maximum damping, without relying on the small-scale structures in the walls dispersing the sound.  In a way, such an anechoic chamber would have only one "really quiet spot", but that should be fine for measuring noise generated by equipment.
penfold:
The first thing that jumped to mind in measuring sounds with poor SNR: is the noise of interest correlated to any other measurable signal (line frequency, switching frequency etc..) and is it likely to be constant or varying? Or is it related to something which can be varied (such as load current)? There are potentially some ways you could use any correlated events to your advantage. But of course, if you're just interested in total random spurious noise, then it's not useful.
Navigation
Message Index
Previous page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod