The start of this week we meet at the east campus to discuss the project and prepare for Wednesday's meeting.
On Friday the team met at the Sound and Music facilities on East campus to conduct a proof of concept test. John is allowed access to the recording studios via his employment in the department. The experiment was conducted with the aim of demonstrating the efficacy of attenuating sound transmission through a rigid barrier using a tactile transducer.
Chris Smith advised the team to conduct such a test during last week’s meeting.
John
Following Mr. Smith's suggestions, John devised a procedure for conducting the proof-of-concept testing.
Testing Procedure:
Attach a single transducer to the 5mm plywood and suspend it in the recording studio.
Surround the testing area with modular absorbers to minimize room reflections
Connect transducer, amp, function generator (phone app), oscilloscope, and mic.
Play discrete sinusoids into transducer and measure the SPL dBC on opposite side. Record audio of testing for later analysis.
Measure Vrms of amp output at each frequency
Move mic position and repeat step 4.
Attach a second transducer to opposite side of barrier.
Connect second amp channel to transducer and second oscilloscope channel.
Repeat signal testing at same locations
Record test audio, measure SPL dBC and amp Vrms
After Testing was completed, John compiled the data and produced some tables of the test results.
This week John conducted some independent research to determine an appropriate range of capacitor values and dielectric types for the project.
The Cirrus CODEC datasheet provides schematics for recommended input and output signal conditioning (shown right)
Specified capacitor values are:
Electrolytic:
10uF, 1uF, and 22uF
Ceramic(C0G):
470pF, 0.1uF, 2700pF, 1.5nF, 6.8nF, 2200pF
Through his research, John found that the choice of dielectric material has a significant impact on the distortion performance of an audio circuit.
Capacitors with large signal voltages across them are more prone to introducing distortion and must be selected carefully to minimize signal degradation.
Small Signal Audio Design by Douglas Self provides a good overview of appropriate capacitor types for use in audio circuits.
C0G, Polystyrene, and Polypropylene all have good distortion performance when large signal voltages are applied across them, such as when setting the time-constant/frequency response of a filter.
Distortion introduced by electrolytics can be reduced by simply increasing the capacitance to reduce the signal voltage across the cap.
The InvenSense ICS-40300 datasheet provides some recommended capacitor values when connecting an Op Amp.
The mic output signal contains a DC offset of 0.8V. The 1uF minimum capacitor in combination with the 47k-ohm shunt resistor forms a 1st order high pass filter (fc= 3.38Hz) to block the DC offset.
John ran PSPICE simulation of the suggested op amp circuit using LM4562 from texas instrument and suggested filter values.
Feedback resistors were chosen to give gain of 6dB
AC sweep analysis showed 214.5uS of voltage group delay @50Hz.
Hunter
(Week 8)
This week the CY7C68013A Mini Board (used as the USBi programmer) arrived. This is what we are using to program the DSP, and will save us nearly $80 rather than buying a USBi straight from Analog Devices, which would essentially just come pre-programmed.
The free USBi driver is downloaded and programmed onto the CY7C68013A Mini Board first, so that the SigmaStudio software can then be used to graphically create a program for the DSP, shown in the picture below.
Online resource used to program the USBi. Source:
Additional updates were made to the website. The video links below were helpful in gaining a better understanding of the A-B push-pull amplifier, in reference to this project.
Group
Proof-of-Concept Testing:
The team met on East campus in order to conduct the proof-of-concept testing
Two rounds of testing
1 transducer
2 transducers (opposing sides)
3 mic locations per round
16”
28”
28” and 12” off-center
9 discrete sinusoidal frequencies
50, 75, 100, 150, 200, 300, 400, 500, 1000Hz
Pink Noise
5 second Sinusoidal Sweep 50-1000Hz
Function Generator phone application
Oscilloscope Measurement of Amp Output:
VRMS
Frequency
Comments