Optimizing sound energy transfer through plastic medium

[tchains]

Hi everyone, I am looking to improve the transmission of sound from a small 1/2" electro-mechanical transducer through a plastic medium (about 1/4" thick). I understand plastic is not a great medium for this but I do not have the ability to change this. The defacto setup doesn't work too terribly, but at certain times the sound, especially higher frequencies, is muffled. I've been looking into the use of coupling agents (such as gelatin, oils, etc) to help with this. What are the exact properties of the coupling agent that I should be designing around? Should it match the acoustic impedance of the plastic medium? Of the transducer? The average? What should the thickness of the agent be relative to the medium? Any help appreciated, thanks!

 

[nasu]

If the high frequencies are muffled it may be that the absorption in plastic is the cause. The absorption in the medium increases with frequency.
The coupling agent will help if the problem is the reflection of sound at the interface between transducer and the medium (plastic).
If the interface between the transducer and medium includes a thin layer of air a significant amount of sound energy will be reflected. If this is the source of the roblem, any gel or even water will improve significantly the transmission through interface..

 

[tchains]

If the high frequencies are muffled it may be that the absorption in plastic is the cause. The absorption in the medium increases with frequency.
The coupling agent will help if the problem is the reflection of sound at the interface between transducer and the medium (plastic).
If the interface between the transducer and medium includes a thin layer of air a significant amount of sound energy will be reflected. If this is the source of the roblem, any gel or even water will improve significantly the transmission through interface..

So I understand the benefit of the coupling agent to be twofold: the gel like consistency helps to conform to the shape and any surface irregularities (to your point) but it also "matches acoustic impedances". I am not sure though how this latter effect is actually engineered.

 

[nasu]

I don't think as two different effects. I mean, if you have strongly irregular surface the sound will be scattered at the interface, rather than being nicely reflected back.
Adding gel will decrease reflections, no matter is there is scattering or just specular reflection.

The acoustic impedance of air (or other gases) is so much lower (3-4 orders of magnitude) than that of any liquid or solid material than anything that is not a gas will improve the transmission by orders of magnitude.

 

[Nidum]

Some or all of the sound may be being transmitted in a diaphragm mode over a large area rather than in a simple through thickness mode over a small area .

Diaphragm mode is what you have in a solenoid and cone loudspeaker .

The transmission characteristics of a diaphragm depend in a complex way on it's mechanical construction .

 

[Tom.G]

Should it match the acoustic impedance of the plastic medium? Of the transducer? The average?

As is true in any matching situation, the coupling impedance (Z_match) should be the geometric mean of the impedances you are trying to match i.e. the sqrt of the product of the two impedances,
Z_match = sqrt( Z_plastic x Z_transducer)

What should the thickness of the agent be relative to the medium? Any help appreciated, thanks!

Thickness would be just enough to fill any voids between the two, you don't want any bubbles or voids. Any extra thickness just gives a longer path and therefore higher losses.

especially higher frequencies, is muffled.

If possible, pre-compensate the driving source to accentuate the higher frequencies, thus counteracting the low pass filter effects of the plastic. In general, the frequency dependent losses in a medium is directly proportional to the number cycles in the medium. This makes it directly proportional to frequency for a fixed distance.

EDIT: If you are trying to get, say, 10kHz through a few inches of plastic, you may get cancelling interference from the reflections at the far surface, giving discrete nulls in the frequency response. (Just noticed your 1/4" thickness, but I'll leave this edit here for future readers.)

 

[Tom.G]

Afterthought as I was falling asleep:

Some Banks and other businesses have a thick plastic shield between the employees and the public (bullet proof shield). In order to communicate, they use a small, battery operated intercom mounted on both surfaces of the plastic shield. I don't know if the two units communicate wirelessly or are wired. If wired, they can run the wires thru a very small hole in the plastic, which could be sealed if needed.