1. To: S. Scott Moor
From: Abrar Alam, Eric Biesiada, Marcus Speith
Date: December 11th
, 2013
Subject: Final Results Memo
Legos were used to design a device in a project to harvest energy using a simple rotating motor,
and a piezo crystal to be distorted using vibrations. After several different designs were initially
tested, a vibrating platform with a wheel striking the piezo with a lever pulley was chosen as the
main source for energy harvesting. This did not work very effectively. It had occurred that the
vibration of the platform did not distort the crystal enough as originally intended. It was then
clear that stable housing for the motor and piezo was necessary along with direct consistent hits
from the fly-wheel arms for efficient crystal distortion. We were able to perform these tests using
a power supply unit and various other instruments to measure output and record collected data. It
was theses initial tests that we were able to utilize and alter our original design for the desired
output that we were trying to achieve. After several tests were conducted and numerous
adjustments were made, some observations had occurred and are described in the following
sections.
Problem Definition: A button battery needs to be charged and how we do this is energy
harvesting using a vibration system complied by a Lego design. A motor and piezo buzzer is also
used to create this energy in order to create enough vibration to enable a battery charge. The goal
is to somehow harvest this energy using the motor to create enough vibration from the piezo
buzzer in order to distort the crystal for enough energy to be transferred to the battery and
attempt to achieve a 10 percent charge.
Final Design: The project started with three initial designs with two of them not meeting
requirements or any possibility of achieved results. The most realistic design to build and utilize
was a simple platform with housing for the motor and another one for the piezo. Although
construction was successful, with initial testing using MATLAB software and a voltage divider,
we were able to conclude that we were not going to meet desired expectations. This is where
further development and alteration to the design came into place. We moved the piezo close
enough to the fly-wheel to enable full contact strikes from the fly-wheel arms. Fig. 1 below
shows how this design began to take shape. We were then able to fully utilize our design using
this construction; however, there were still problems with the piezo moving out of range from the
strike of the fly-wheel when certain voltage levels reached leaving us unable reach goals from
the max 4.5 volts from the power supply. So, we built stronger reinforcing walls around the
piezo housing, a stopper secured to the base plate, and added another wall around the motor. In
Figure 2, one can see the finished product where we were able to have full stability of our design
without the Legos falling apart and reach the maximum voltage level of 4.5 for continuous
strikes to the piezo without it moving anywhere. Figure 3 gives you a CAD drawing similar to
the finished design.

2. Figure 1 Stripped down design front and side view
Figure 2 Final Design Figure 3 CAD Design Final Product
Methodology: Various methods and tools were used to test and experiment how the harvesting
was to be executed. First, we had a group brainstorm on ideas we thought would work for us
along with researching a few online videos for inspiration. Then, we used MLCad to get visuals
for ideas and to see how they might fit together and to give us an idea of what Lego parts and

3. sizes we were going to need. From there, it was hands on with the Legos by ways of trial and
error. Next, we began to experiment with the motor rotation and perform load and unload tests to
see how it would perform using the power supply at 1.5, 3, and 4.5 volts in all three experiments.
We used a tachometer to measure rotary speed, a voltage divider, and an oscilloscope software
and spectrum analyzer (MATLAB) to measure and record data. Finally, after all of the
modifications to the design were complete and initial testing samples recorded, a breadboard
circuit was constructed with diodes and a capacitor to charge the battery. The battery charge was
tested using a separate charging reader. Of these tests, Figure 4 below shows the motor with no
load where the RMS tops out at 3700 and was measured using a tachometer.
Figure 4
Figure 5 below shows where we tested the motor with the load where it topped out at 2800, also
measured with the tachometer. Once these tests were performed, we ran several initial runs with
the voltage divider and MATLAB to see what kind of waveform and power spectrums we would
get. Getting lower results than we had expected, it was at this time where we were curious to see
what manual strikes to the piezo would rate which resulted in low duty cycles as well, reaching
an average of a 2 percentage rate. It was at this point where we began to make more serious and
advanced alterations to our design before we would begin performing battery charges and have
expectations met.
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
-500
0
500
1000
1500
2000
2500
3000
3500
4000
Volts
RMS
Tachometer Readings - No Load

4. Figure 5
Experimental Results: To test these combined procedures, we had to see what they were all
capable of and how they performed under various working circumstances. Alterations varied
depending on how they performed. The design had to withstand the power of the motor topped
out at 4.5 volts. All of the wires had to stay in place to perform proper running use. We needed
higher duty cycles using MATLAB than our beginning phase low averages of .5-1.2 percentage.
After all of this was done, we ran five minute intervals of up to 30 minutes to test the battery
charge, stopping after five minutes each time to measure a recording.
Motor Speed: The speed of the motor was tested at three different voltages by utilizing a
tachometer. The tachometer gave us readings in rotations measured per second. The speed of the
motor was first tested with absolutely nothing on the load so that the motor performed
impeccably. The table and the graph below represent the results of the three different tests:
Voltage/V 1.5 3 4
Speed/RMS 931.9 2393 3700
Table 1
We can infer from the tabulated information that for every 1.5 increase in the amount of voltage
supplied by the load, the rotational speed of the motor first increases by approximately 2.6 times
and then increases by approximately 1.5 times. Even though the rotational speed of the motor
increases the rate of increase of the speed decreases as the voltage supplied increases. Therefore
a linear relationship does not exist between the two variables. In our second attempt, the motor
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
-500
0
500
1000
1500
2000
2500
3000
3500
Volts
RMS
Tachometer Readings - Full Load

5. was fully loaded and the rotational speed was measured with the piezo devise being struck in full
motion usage. The results are tabulated and graphed below:
Voltage/V 1.5 3 4.5
Speed/RMS 840 1800 2800
Table 2
We can now deduce that due to an increase in the weight of the motor, the maximum rotational
speed of the motor has decreased. The same voltages were used in our second attempt. Apart
from the difference in speed, all the previous observations regarding the rate of increase of the
rotational speed are the same.
Piezo Output: The output of the piezo devise is measured by using an oscilloscope software and
a spectrum analyzer on MATLAB. With the help of such software we were able to find the
waveform, power spectrum, frequency, duty cycle and Volt-RMS readings of the piezo devise in
two different cases. The results are displayed by the figures below:
Example A MATLAB Output Graph at 3 Volts

6. In Example A, the function of the piezo devise was recorded at 3 volts. The results displayed a
VRMS of 4.1283 per volt cycle with a frequency of 3125 hz and a duty cycle of 7.1 percent. In
example B, the function of the piezo devise was recorded at 4.5 volts and resulting information
gave us a VRMS of 4.204 per volt cycle, a frequency of 2993 hz and a duty cycle of 7.98
percent. The tests were conducted over a period of 30-60 seconds during which the model
maintained design durability. It is observed that as the voltage supply is increased from 1.5V to
4.5v, the voltage produced increases by .8V.
Example B MATLAB Output Graph at 4.5 Volts
Charging Circuit Output:
Battery Charging: Several tests were made during battery charging. We started to see what the
battery was at the beginning of our test as well as ran several tests to see if the design would stay
intact at full voltage of 4.5. The battery charge at initial testing started at 1.806 and ended with
1.152 after 35 minutes of continuous tests ranging from five minutes per charging cycle. Figure 6
below shows the testing results. A battery charging reader was used to record this data. The
results seemed to decrease in number, then increase half way through, and then decrease again
ending at 1.217 after 30 minutes of charging use
.

7. Battery Charge
Conclusions: Although our final results leave us feeling unsure if this was the best possible
design to harvest energy, there are several conclusions we have come to agree upon:
1. Our design stayed intact for a full 30 minutes and some additional time just for
cautionary testing purposes.
2. There is no damage to any of the Legos do to the vibration, nor did we use any rubber
bands, tapes, ties, or any other tools to construct our design. It was nothing but a motor,
piezo, and Legos.
3. Our design is very simple in construction and set up. The motor rests in a housing and a
wheel hits the piezo which also sits in a sort of harness.
4. During our first test recording with MATLAB, we physically dangled the piezo and kept
hitting it with our fingers or tapping it on a table with a pencil to see what kind of
readings we should expect to get and they were still disappointing. Our max duty cycle
was 3.1 percent.
5. During our battery charging and voltage dividing test, it seemed as if the power supplier
was topping out between 3.5-4 volts. Even when the power was increased, the needle did
not seem to move or fluctuate too much higher than those gains.
6. We think it would be possible for our design to keep charging for an extended period,
minding that it does not vibrate its way off of the table or the clips become loose.
7. Our biggest flaw design, setback, and time consumption was the motor. The wires would
not stay in and they are very sensitive. Very careful hands and wire placement is
essential. Making sure there was not too much slack in the equipment was very beneficial
as well. When it is neat and organized (i.e., no cables resting or pulling on each other)
things flowed more smoothly and stress free.
0
5
10
15
20
25
30
35
0 2 4 6 8
MinutesCharged
Charge Level
Battery Charge Charge
Level
Battery Charge Minutes
Charged
Linear (Battery Charge
Charge Level)
Linear (Battery Charge
Minutes Charged)

8. 8. Our design was successful at keeping a consistent waveform and duty cycle at topped out
voltage from our secure design of the motor and piezo placement. Duty cycles usually
consisted of 5.6-7.9 percentages at 4.5 volts.
9. We believe that with our design, the voltage setting, the power of the strikes, or duty
cycle, that we could not have increased it or improved it much more in the design without
breaking the piezo, shaking the Legos apart, or knocking the piezo so hard that the wires
ripped out. More powered would be needed and the piezo stay intact and take the
punishment. This is what we have concluded.
Parts List: