Week 8 was the last week that test were being carried out for the speakers.  During this week the blind study was completed.  In the first sample of 15 people, 9 preferred the MDF cabinet over the Oak cabinet.  These results did not show the significant favor towards the MDF cabinet that was expected, but was fairly consistent with test results of slightly straighter frequency response from the MDF cabinet.  Another sample was taken with 24 different people to attempt to get more conclusive results.  This resulted in only 7 people preferring the MDF compared to the 17 that preferred the Oak.  The second sample shows a 70.8% preference to the Oak, where the first sample shows a 60% preference to the MDF cabinet.  These tests show that listening tests are very subjective and a person's preferences might affect the data.  For example, if a person likes a lot of bass in their songs, they will probably gravitate towards the Oak cabinet because of the slight boost it has in the lower frequencies compared to the MDF cabinet.

Week 7 of the engineering project was the test week for the speakers that were built.  During this week completion of the vibrations test as well as the frequency response test were preformed.  The graphs below show the results from the frequency response tests at the two distances measured.

From these graphs, it can be seen that the MDF did the best job at giving a flat line frequency response.  This is noticeable in both the high and low frequency response graphs.  Another interesting observation from the graphs is that both gave a volume boost to the lower frequencies produced by the speakers, with oak having the highest volume, but a slightly more crooked frequency response line.  Another test that was completed this week was the vibration test to see how much the cabinets vibrated at various frequencies.  This was done using an application and the accelerometers in the iPhone.  The graphs below show the results.

MDF 100Hz

MDF 125Hz

MDF 160 Hz

Oak 100Hz

Oak 125 Hz

Oak 160Hz

These graphs show the acceleration in the x, y and z direction.  The most noticeable direction is along the y axis. For reference, if one was looking straight into the front of the speaker, the vibration would be moving towards and away from them.  This vibration is caused by the movement of the speaker, which is also along this axis.  The most noticeable difference is in the graphs of the 160Hz frequency where the oak has almost twice the amount of vibration that the MDF does.  This vibration is probably what causes the "boomy" bass that the cabinet tends to produce.  This is an unwanted attribute because what the speakers are sitting on will vibrate which can cause damage or unwanted rattling noise.  Another interesting observation is the graphs of the 100 Hz frequencies.  The sinusoidal wave that was shown on the phone suggests that this is the resonant frequency of the speaker cabinet.

During the 6th week of the engineering design project the design team worked on constructing the speaker cabinets, deciding how to collect data from the speakers, and getting preliminary data of the speaker without the enclousure.  The picture below shows the constructed speaker cabinets.

One design change that came up during the construction was to make the baffle, which is the face of the cabinet that the speaker is mounted to, out of plywood.  This decision was made for several reasons.  The first was that the oak cabinet would have gone over the allotted 30 dollars for materials if another piece of it was used, and the plywood was cheaper.  Also, MDF and Oak wood split very easily.  With the mounting screws so close to the edge of the hole drilled for the speaker, something that wouldn't split as easily was needed for the construction.

Another task that was accomplish was to decide which tests will be done on the speakers and how data would be collected.  The three tests that will be done on the speakers are a frequency response test, vibration test, and a subjective test on sound quality.  The frequency response test will be done with an iPhone application.  This application allows the user to play different frequencies and use the microphone on the iPhone to record the volume of sound produced by the speakers.  The application does not give a unit for the measurement of volume, but since the measurements are being used for comparisons, this will not cause a problem.  This will be done 3 in. from the speaker and 7 in. from the speaker.  This will allow the design group to see how the material in the cabinet alters the sound produced by the speaker.  The vibration test will also be done with an iPhone application.  This application uses the accelerometers in the phone to record the acceleration in the X,Y,and Z direction over a period of time.  The frequencies will be selected in the high, middle, and low range to test the vibration of the cabinet.  Since vibration of the cabinet can produce it's own sound, this test will allow the team to see which cabinet does this the least.  Finally,the design team will use test subjects (people) and play a track through one speaker, then the other and have them decide which speaker they like better.  This is a subjective test because different people can have different perspectives on what sounds better when it comes to music.

The final task that was completed was a frequency response test for the speaker without an enclosure. This test will hopefully help see what differences the enclosure makes, not just compared to another cabinet, but in general.  As seen in the graphs below, which are graphs of the volume of sound produced by the speakers at different frequencies, speakers that do not have enclosures have bad frequency responses.  The optimal response is a flat line. This flat line means that the speaker produces every frequency at the same volume.

This is mostly due to the standing waves produced by the back of the speaker.  As explained before, these standing waves can cancel out the waves produced by the front of the cone of the speaker, altering the frequency response.  The hope of the design team is that the cabinet will reduce these standing waves and level out the graph.

The fifth week of the engineering design project was mostly spent waiting for materials to come in so construction on the speaker enclosures could begin.  During this waiting time, it was decided to do more extensive research to help clear up some loose ends and bolster the final report.  To begin, the team looked into the golden ratio for speaker cabinet size and found that the ratio is with respect to the sides, not the size of the speaker.  For example if the front of an enclosure is 6 inches wide, the box would be 9.6 inches tall and 3.6 inches deep, even if the speaker is 3 inches deep.  Also, the team looked into more information about what the golden ratio actually does.  This ratio helps to dampen standing waves in a speaker cabinet.  These standing waves are the opposite of what comes out the front of the speaker.  If positive and negative waves meet, they will cancel each other out, this is destructive interference.  Therefore, being able to eliminate these waves is vital.

Destructive Interference

When two waves interfere, the sum of their amplitudes determines the new wave's shape.  The amplitude of a wave is its maximum amount of displacement from its position at rest to its crest (the uppermost part of a wave).  When two waves are 180 degrees out of phase, phase denotes a particular point in the cycle of a wave,  the crest of one wave is at the same point as the trough (the lowermost part of a wave) of the other.  Since these two amplitudes are opposite of each other, one being positive and the other being negative, when they add together the sum is 0.  Therefore the two waves cancel each other out.

This is not the only measure the team took to create destructive interference between these waves.  More research was also done with respect to the materials.  This research includes how the properties of the materials that were researched will affect the sound produced by the speaker.  The goal of a speaker is to eliminate the standing waves produced by the speaker.  It does this in two ways; through dampening and resisting resonance.  The properties that affect the dampening a material creates are density and stiffness.  Density because the more material that is between one and the speaker, the less one is going to hear.  Stiffness affects the dampening in a different way.  This property is slightly more subjective than density is, but still important.  If a material is very stiff, it will allow the waves inside the cabinet to bounce off of it nearly unchanged, which will then redirect it to another wall of the cabinet.  This becomes a problem because the number of waves inside the cabinet rise quickly as they bounce off the walls instead of being absorbed.  To absorb the sound waves, the enclosure is going to have to be soft.  If the enclosure is soft though it will be susceptible to resonance.  Resonance is when the walls of the cabinet move, making their own sound.  This is caused by the material being flexible.  Unfortunately, being flexible is the opposite of being stiff.  So for the material a good compromise between stiffness and flexibility is needed to try  and dampen, as well as not resonate.  It is the design groups belief that oak wood and MDF are good candidates for these properties.