Double slit expt using sound?

[bwanaaa]

I want to construct an apparatus to test the double slit experiment using sound waves. Is this a fool's errand? I guess I first have to know whether sound waves can have 'corpuscular' character. That particle like behavior of light is operationally defined by the phenomenon of the photoelectric effect (perhaps I am off base here). But I think the character of a phenomenon is really defined by the type of detector one uses.

One way I imagined to create a very sensitive sound interference detector is to have two very fine conductive meshes very close in a sandwich separated by a microscopic distance. Positive interference of sound waves would create a sufficiently high amplitude at specific locations on the first mesh. Waves in the mesh would push it closer to the second (further) mesh. Measuring the capacitance of wire pairs between meshes should reveal peaks where the wires are close together. Obviously the wires need to be insulated.

Another way might be to measure light scattering off a Mylar sheet. Peaks of interference would be the highest topographical feature on the sheet. These spots should create the most scattering.

 

[intx13]

The double-slit experiment demonstrates the particle nature of light (by the absorption pattern on the front of the slotted wall) and the wave nature of light (by the pattern illuminated on the second wall by light passing through the slits).

An experiment like that is needed because photons are very small and move very fast. Sound is much larger and slower, so it can be examined much easier.

The particle nature of sound can be demonstrated by putting some marbles on an upwards facing speaker projecting a single frequency. The marbles will bounce around and bump into each other forming a wave pattern. The troughs/peaks in the wave pattern, when measured, will correspond to the frequency of the sound. The wave nature can be demonstrated through interference. Grab a pair of noise cancelling headphones.

It would be pretty tough to do something similar to the double-slit experiment with sound. You'd need a material for the wall that reflects sound perfectly (none exists) or at least good enough that you can reliably rule out any sound that passes directly through the wall, as opposed to through the slits. Plus, your wall would have to be really big, since sound waves are relatively low frequency. Additionally, your sound waves would have to be reasonably collimated. This means a speaker with a really massive element so that the surface of the sound wave appears like a plane when it hits the wall.

Then, on the front of the wall, you'd be looking to demonstrate the particle nature of sound. You'd need some way to visualize when a single particle bumps into the wall. Behind the slits, on the back wall, you'd need to show evidence of construction and interference. This part would be simple, just use an array of microphones.

The double-slit experiment is probably not the best way to experimentally demonstrate the particle and wave nature of sound.

 

[Born2bwire]

Sure, sound waves will reflect the Young's Double Slit experiment just as well as light. You would need a rigid wall to act as your screen, concrete is wonderfully reflective of sound, and you would need to adjust the dimensions accordingly. Honestly, I think you could easily get by with a microphone as your detector and a damped plywood wall. The speed of sound is rather slow, 343 m/s, so a 1 KHz signal has a wavelength of around 14 inches if I recall correctly. So get a small microphone and you can easily measure subwavelength variations and thus be able to get a good idea of the interference pattern. Heck, you can experience interference on your home stereo or home theater by simply putting one of your stereo channels out of phase. You can still hear the sound, but the reflections from the walls of the room dominate more since the direct sound waves interfere. So basically, go out to an open field (so you do not get any cavity reflections as interference), set up a loudspeaker at one point, set up a wall with your gaps using a rigid material, then on the other side have your microphone (or just a SPL meter) on a track so that you can move it and measure the sound pressure.

 

[bwanaaa]

Microphones. Much easier. So if I pack them every 1.75 inches, I should get 8 of them per wavelength at a 1 kHz signal. The problem of is sensitivity. To parallel the double slit expt I would need to decrease the amplitude of the sound wave until it is at the excitation threshold of the detector. The goal would be to get only one microphone to detect an impact per sound burst - which would have to be Brief as well(speed of snd= 1100ft/sec , wvelngth=14 in, so a wavelength of sound will be emitted in 1/942 of a second~a millisecond). So a sound packet will require a 1 millisecond burst of sound. to resolve this i think i would need a detector good to half a millisecond ( nyquist limit? )A reflective membrane with coherent light emitter 45degrees off from the detector axis would be much more sensitive than a mic. Isn't something like this used by spies to eavesdrop on conversations? (vibrations on a window). I don't know what kind of transducer can generate 1 kHz for submillisec bursts tho

 

[intx13]

The goal would be to get only one microphone to detect an impact per sound burst

I'm not sure what this means. Why not just have a row of microphones and sample them all simultaneously, repeatedly, and plot the resulting volume recorded at each microphone? Your plot should look just like the interference pattern in the double-slit experiment.

This is how I'd do it:

First, you need a source of collimated sound waves at a particular frequency (say, 1 kHz). These waves will have wavelength 0.343 meters in air at 20 degrees C. If we use a omnidirectional speaker in an anechoic chamber, the wavefront will look like a sphere expanding around the speaker. If we stand far enough away the sphere gets big enough to look like a plane from any direction. That's where we build our wall. To avoid aberrations from the wavefront hitting the edges of the wall, we'll build our wall from side-to-side, floor-to-ceiling. From the back of the wall there are only the two little slits peering through to the other side of the chamber. Far away through these slits we can see the speaker.

Now we need to set up our microphones in a row at the height of the slit. To determine how many microphones we need and how closely they should be spaced (and how far away from the wall they should be) we need to do some math.

Let's say that our slits are d meters apart and we're going to stand up our microphones z meters back from the wall. The peaks of interference - that is, the distance between the loudest points where the waves constructively interfere - will be z * 0.343 * d meters apart. So, if our microphones will be 1 m away from the wall and the slits are 1 m apart, the peaks will be 0.343 m apart.

To resolve volume differences across 0.343 m we need our microphones to be much closer together than that. Let's use 8 microphones and space them 0.05 m apart.

Now we wire our microphones into a receiver that will continuously record the audio. The frequency of the sound wave is 1 kHz, so our receiver needs to sample no slower than 2 kHz. You can get dirt cheap receivers that can sample signals at audio frequencies up to MHz, so that's no problem at all.

Turn on the speaker and start recording on all the microphones. Give it a little bit of time (I'd ballpark 10 seconds or so) and turn it off. Now compute the average volume of the audio recorded for each microphone.

Now make a nice pretty plot of dots in a row. The first dot corresponds to the leftmost microphone, the second to the next microphone, and so on. Color the dots based on the average volume for that microphone. The quieter the average volume, the lighter the dot. The louder the average volume, the darker the dot. That's your interference pattern.

 

[Born2bwire]

Yeah, all you need to do is record the SPL as a function of location. You can do this by an array of microphones/detectors or just have a single microphone/detector on a track that you move and record at different positions. As mentioned above, just take the average or find the amplitude of the recorded wave (which is what a SPL meter would in effect do anyway). No reason to do bursts, the interference pattern arises in the steady state behavior. Just turn on your source which emits a continuous monochromatic sine wave and start taking data. Only other caveat is that you would want to place your source equidistant between the slits so that the incident waves on the slits are phase-coherent. This removes the need of having a large loudspeaker as your source. You can still have an omnidirectional point source if you have subwavelength slits. That is, if you have slits that are small compared to the wavelength, then you can use any speaker as your source as long as the speaker is in between the slits. Or, if you are more competent, you can just calculate the effect that an off-center source would have.

Regarding an anechoic chamber, that is necessary, but doing this outside in say an open grass field should be sufficient since you only have weak diffused ground reflections.

 

[Biftheunderstudy]

The classic ripple tank idea seems like a much clearer example of the double slit experiment. You get nice visual feedback of the interference pattern.

 

[silverpig]

The classic ripple tank idea seems like a much clearer example of the double slit experiment. You get nice visual feedback of the interference pattern.

This. Stick a vibrating stick in a tank of water. Put a partition in the tank with two holes in it at the surface level. Watch the ripples at the end of the tank.