The main idea of this project is to input an audio signal with a microphone and depending on the frequency of the input, two rows of LED’s will light up corresponding to the intensity of the bass and treble using the Arduino and a high and low pass filters for the treble and bass respectively.

1. CPE 329‐03 Bret Omsberg
Spring 2015 Evan Kirkbride
Treble and Bass Detector
The main idea of this project is to input an audio signal with a microphone and depending on the
frequency of the input, two rows of LED’s will light up corresponding to the intensity of the bass
and treble using the Arduino and a high and low pass filters for the treble and bass respectively.
System Requirements
Takes audio input and lights up LEDs sequentially based on the low and high frequencies of the
input. Essentially acting as an audio level meter.
System Specifications
Use two bandpass filters to separate treble and bass from a given audio signal that is converted
from a sound wave to an electromagnetic analog signal. This analog signal would then be ran
into three separate amplifiers and then sent into the Arduino’s analog to digital converter to
allow for interpretation. A modified digital signal is then sent to two sets of four LEDs
corresponding to the intensity of the treble and bass of the audio input.
System Architecture
Figure 1 below shows the system diagram of the treble and bass detector. The microphone
converts a sound wave into an electromagnetic analog signal. Since the output of the
microphone is in the millivolts spectrum, the signal must be amplified. The signal is then run into
an amplifier with a gain of eleven. This signal is then ran into two separate bandpass filters. The
bandpass filter corresponding to the bass signal passes frequency between the ranges of 8 to
500 Hertz and the treble filter passes signals from 2,000 to 10,000 Hertz. The outputs of the two
filters are then ran through two separate operational amplifiers with a gain of two to help with
any loss of gain from the filters. The Arduino’s analog to digital converter interprets both of these
signals and outputs on the digital pins zero through seven to determine which LED will turn on
depending on the input to the ADC on pins A0 and A1.

2. CPE 329‐03 Bret Omsberg
Spring 2015 Evan Kirkbride
Figure 1: System Schematic of the Treble and Bass Detector
System Design
The microphone used the configuration in Figure 2 below. The 2.2 kΩ resistor acts as a pull­up
on the output, and its value can vary between 1 kΩ and 5 kΩ. After testing the microphone it
was revealed to have a limited output, with amplitudes around 100 mV. Because of this the
output of the microphone was AC­coupled to an non­inverting amplifier with a gain of 11,
allowing the microphone voltage to vary from 1 V to 2 V.

3. CPE 329‐03 Bret Omsberg
Spring 2015 Evan Kirkbride
Figure 2: Microphone with Amplifier
The bass frequencies were originally designed to range from 0­500 Hz, and the treble
frequencies were designed for frequencies above 2 kHz, based on the sound spectrum. Thus
the filters used to isolate the treble and bass from the microphone were originally a high pass
and low pass filter respectively. However when testing the high­pass filter, high frequency noise
was found to pass through in the MHz range. To counteract this, the high­pass filter was
replaced by a band­pass filter for the treble frequencies, from 2 kHz ­ 10 kHz since the
microphone used only detected up to 10 kHz. This eliminated the noise and allowed the treble
frequencies to come through. This lead to a 10 kΩ resistor and 2.2 nF capacitor for the low pass
10 kHz filter, and a 10 kΩ resistor and 8 nF capacitor for the high pass 2 kHz filter. From here
another non­inverting voltage amplifier with a gain of 2 was used to keep the voltage levels
steady enough for the arduino ADC to read.
The bass low­pass filter had the issue of its time constant holding the voltage output of the filter
for too long. To counteract this, the low pass filter was also changed to a band­pass filter to
reduce its time constant. However in order to keeps its frequency range, the low frequency of
the band­pass was designed to be as low as possible, resulting in a 8 Hz ­ 500 Hz bandpass
filter for the bass frequencies. This filter also had a non­inverting amplifier to make the filter
active and keep the voltage levels steady. This filter used a 10 kΩ resistor and a 0.33 μF
capacitor (three 0.1 μF in series) for the low­pass 500 Hz filter and a 0.1 μF capacitor and 200
kΩ resistor for the high­pass 8 Hz filter.
From here the outputs of each filter were fed into a port on the on­board ADC of the arduino.
The arduino interprets these voltages to sequentially light­up LED’s depending on the amount of
treble and bass voltage. An elaboration on this process can be seen in the software section.
The original idea to light up the LED’s was to use the arduino to drive a DAC, and use a specific
analog voltage value for each ‘step’ of the treble or bass. This value would be sent to a
comparator for each LED, and the voltage value sent from the DAC would be greater than the

4. CPE 329‐03 Bret Omsberg
Spring 2015 Evan Kirkbride
voltages compared for certain number of LED’s. The IC’s used for the comparators, however,
did not work, possibly a fault with the IC’s or breadboard used. As a result the LED’s had to be
lit up individually with the pins on the arduino. This sacrificed the number of LED’s that could be
lit, since the arduino had a limited number of pins. This also limited the resolution of the system
overall. Since only a certain number of LED’s could be lit, the voltage ranges sensed would
have to be larger to compensate. With four LED’s for the treble and bass set, the voltage ranges
for each needed to be tuned. The voltage values for the bass component usually varied up to 1
V, while the treble usually varied from 1 V ­ 2 V.
Hardware
Table 1 below shows the bill of materials to create the treble and bass detector system. The
majority of the cost came from the purchasing of the Arduino with the ATmega328p
microcontroller. The total cost of the system was calculated to be $32.54 with the Arduino
making up $24.95 of that cost. The cost of this system could be greatly reduced if the system
was manufactured in bulk so the ATmega328p microcontroller and microphone condenser
would cost significantly less reducing the overall cost of the system.
Table 1: Bill of Materials
Treble and Bass Detector
qty part type designator / id description Cost/unit (U.S. $)
2 Amplifier LM741 Operational Amplifier .50
1 Amplifier LM201AN Operational Amplifier .50
8 Capacitor C104 Capacitor 0.1 uF 100 Volt X7R 10%
Radial
.10
1 Capacitor FC 8828 Capacitor 2.2 nF 0.4
1 Capacitor IIC 103 Capacitor 8 nF 0.2
13 Resistor ­­­­­­­­­­­­­­­­ Resistor, Axial Lead, 1k, 5% .10
4 Resistor ­­­­­­­­­­­­­­­­ Resistor, Axial Lead, 10k, 5% .10
1 Resistor ­­­­­­­­­­­­­­­­ Resistor, Axial Lead, 2.2k, 5% .10
1 Resistor ­­­­­­­­­­­­­­­­ Resistor, Axial Lead, 200k, 5% 0.10
8 LED UT0341R­41­R Red Diffused, T1, 635nm .012
1 Microphone 270­090 Condenser Microphone Element 2.99

5. CPE 329‐03 Bret Omsberg
Spring 2015 Evan Kirkbride
1 Arduino Arduino Uno R3 ATmega328P Microcontroller w/
Arduino Interfacing
24.95
Table 2 below shows the Arduino’s wiring configuration to the rest of the system. This table is
best followed if referenced with the system schematic shown in Figure 1 above. The overall
schematic of the system can be interpreted by referencing Figures 1 and 2 above as well as the
bandpass filter schematic shown in Figure 3 below. The combination of these three schematics
allows for the system to be recreated.
Figure 3: Bandpass Filter Schematic

6. CPE 329‐03 Bret Omsberg
Spring 2015 Evan Kirkbride
Table 2: Arduino Wiring Configuration
Arduino Pin Port Name System Connection Function
GND GND GND Create a common
ground between all
blocks
Power: 5V V​CC AREF Generate the reference
voltage for the analog to
digital converter
AREF AREF 5V Generate the reference
voltage for the analog to
digital converter
Analog: 0 PA0 Output of LM741
Op­Amp for Bass Filter
Input into the Arduino’s
analog to digital
converter
Analog: 1 PA1 Output of LM741
Op­Amp for Treble Filter
Input into the Arduino’s
analog to digital
converter
Digital: 0 PD0 Bass LED1 Determines the state of
Bass LED1
Digital: 1 PD1 Bass LED2 Determines the state of
Bass LED2
Digital: 2 PD2 Bass LED3 Determines the state of
Bass LED3
Digital: 3 PD3 Bass LED4 Determines the state of
Bass LED4
Digital: 4 PD4 Treble LED1 Determines the state of
Treble LED1
Digital: 5 PD5 Treble LED2 Determines the state of
Treble LED2
Digital: 6 PD6 Treble LED3 Determines the state of
Treble LED3
Digital: 7 PD7 Treble LED4 Determines the state of
Treble LED4

7. CPE 329‐03 Bret Omsberg
Spring 2015 Evan Kirkbride
Figure 4 shows the physical representation of the system shown in Figure 1. The Arduino is
powered through the use of a USB cable connected to a computer. The system was wired in
such a way that can be followed by using the Arduino wiring diagram in Table 2 above.
Figure 4: Physical Representation of the Treble and Bass Detector
Software
Atmel Studio version 6.2 development software was used to program the Atmel 328 with C++.
An instructor provided the arvdude.exe, arvdude.conf, and prog.bat files to help program the
arduino with .hex files created by Atmel Studio by placing them in the C:Windows directory.
The program the arduino used began by defining the libraries used as well the functions used to
initialize the components on the microprocessor. The on­board ADC had its initial registers and
states initialized with the Initialize_ADC0() function, and the on­board timer used to generate
interrupts was constructed with the init_Timer0() function. Both of these functions can be seen
below.
Insert ADC and Timer init functs

8. CPE 329‐03 Bret Omsberg
Spring 2015 Evan Kirkbride
These functions are called in the beginning of the main function. The beginning of the main
function also sets up the PORTD pins on the arduino as the outputs that drive the LED’s,
enables interrupts that will be generated by the timer, and initializes the outputH and outputL
variables. These variables act as the upper and lower 4 bits of the PORTD register that
determines the logic of each output to the 4 LED’s for the bass and treble. This initial steps of
the main loop can be seen below.
Insert Main Loop Start
The ADC was used to interpret the bass and treble inputs with the 5 V from the arduino as
reference. Each while loop first reads the bass input, and then the treble. For each reading, an
if­statement determines the appropriate voltage level the bass or treble is at, and how many
LED’s to light up. The higher the voltage, the more LED’s light up. This output is reflected in the
outputL and outputH variables. This process can be seen in the code below.
Insert for loops code
The timer was used to generate interrupts, and during each ISR the arduino would update the
output to the LED’s. Without this the LEDs would blink too fast to be noticeable if they were
driven in the while loop. So having the timer with a clock running at 1/1024th of the arduino
clock helps the visual feedback remain smooth. However, this means that the arduino reads
data faster than it outputs it. This can be changed by the prescaler on the timer, however the
visual feedback on the LED’s might suffer as a result. The ISR can be seen in the code below.
Insert ISR code
Test Results
The microphone used was fairly cheap, and could not pick up much sound. The easiest way to
test for sound was the tap the mic directly. This was quick and allowed for small pulses that
contained high enough frequencies to be seen on the filters. Blowing or whistling in the
microphone worked as well, however these frequencies could be filtered out by the filters used.
The acquisition of the treble and bass frequencies was fairly successful. The act of filtering and
amplifying the sound took some time in order to consistently get values that the arduino could
read. This testing resulted in the need for band­pass filters rather than low­pass and high­pass,
as well as the need for active filters.
The resolution determined by the arduino was limited by the amount of LED’s that could be
driven. Originally ten LED’s per treble and bass was planned, however this had to be reduced to
four LED’s due to hardware limitations.
Personnel

9. CPE 329‐03 Bret Omsberg
Spring 2015 Evan Kirkbride
Bret Omsberg handled most of the hardware setup and schematic creation. He also contributed
to the design of the filters, starting with a simple RC low­pass and high­pass that would later
become two active bandpass filters that provided a much sharper signal with the removal of
high and low frequency noise. The troubleshooting aspect of the project as well as the
powerpoint creation was done in conjunction with Evan.
Evan Kirkbride handled most of the software design, and adapting the software with the outputs
from the hardware. Troubleshooting with the hardware and software as well as the presentation
preparation was done along with Bret.
Conclusion
Bret Omsberg
This project offered many learning experiences along the way. The original design of this
system included the use of digital to analog converters, converters, and transmission gates. As
the implementation process continued, it was realized that the DAC was no longer necessary
which saved money from a design perspective. The original design also included ten LEDs for
each Bass and Treble output, but this was reduced to four per line due to the comparators not
operating properly. A large portion of the problems occurred in setting up the filters. The filters
were first implemented using simple RC low­pass and high­pass circuits but this presented a
new issue, noise. This was corrected using a bandpass filter for the high­pass filter and was
made even sharper with the introduction of an op amp that made the filter an active bandpass
filter. The low­pass was eventually redesigned into an active bandpass filter to provide a signal
that had very little noise. This project could be improved with the inclusion of twenty
comparators that would allow ten LEDs per bass and treble line that would increase the
resolution of the measurement. A higher quality microphone would also allow for a broader
spectrum of frequencies to be analyzed due to the microphone used in this project cutting off at
10 kHz which is only half of the hearable frequency for humans. The project could also be
adapted so that LEDs were arranged in a geometrical shape to make it more aesthetically
pleasing to the user.
Evan Kirkbride
This project allowed for us to not only use the arduino in they ways we have learned about
throughout the quarter, but also allowed us to use circuits we learned about in previous classes.
Namely filters and amplifiers, and this project allowed for a synthesis of those elements. Since
these components had to work with an audio signal, a signal that is more noisy and sporadic
than typical signals used to test these circuits, the components needed to be high­quality and
consistent. This lead to the testing of this hardware to be the most intensive part of the project.
Figuring out the appropriate frequency ranges, using amplifiers to keep the gain of the signal
consistent, and eventually changing each filter to band­pass allowed for some theory practice
and familiarity with these systems. By the time we got to the arduino and LED light­up process,
there was not much time within our deadline to ensure their function, and once it was clear that

10. CPE 329‐03 Bret Omsberg
Spring 2015 Evan Kirkbride
the hardware needed was not working, the arduino had to be used directly to drive the LED’s.
This was the biggest let­down of the project, and if more time was allotted to improve, getting
more LED’s driven for the treble and bass would be the first step, allowing for more resolution.
The quality of the microphone would probably be the next step, allowing for better quality sound
over a larger frequency range. A container for this system could also be useful to make this
system more marketable.
References
ATmega328 Datasheet
ElectronicsTutorial Active Band­pass Filter Design Webpage
Texas Instruments LM741 Operational Amplifier Datasheet
Texas Instruments LM201AN Operational Amplifier Datasheet
Appendix A – Marketing Document

11. CPE 329‐03 Bret Omsberg
Spring 2015 Evan Kirkbride
Appendix B – Source Code