EE 420L – Engineering Electronics II Lab – Lab 7
Due: April 3,
2019
Lab Description
·
Design and simulation of an audio amplifier using MOSFETs.
·
Review lab 6.
Lab Tasks
Design an audio amplifier (frequency range from roughly 100 Hz to
20 kHz) assuming that you can use as many resistors,
ZVN3306A transistors, and ZVP3306A transistors as you need along with only one
10 uF capacitor and one 100 uF
capacitor. Assume that the supply voltage is 10 V, the input is an audio signal
from an MP3 player (and so your amplifier should have at least a few kiloohms
input resistance), and the output of your design is connected to an 8-ohm
speaker (so, ideally, the output resistance of your amplifier is less than 1
ohm).
- Your lab report should detail your thoughts on the
design of the amplifier including hand-calculations.
- A good place to start is with the push-pull amplifier
characterized in lab 6.
- Simulate your design. Document the results in your lab
report.
- Build and test your design.
- Document the performance of the design including power
dissipation, output swing, input resistance, output resistance.
----------------------------------------------------------------------------
Starting Point (Given Push-Pull
Amplifier)
·
To begin, we analyze the following push-pull amplifier to see
where modifications should be made
and tradeoffs can be selected.
·
Note that students were given a 22 Ohm speaker for this lab. The
22 Ohm speaker sounds much better
than the 8 Ohm speaker does anyway, so this is the speaker that
was used in the final design.
Simulation Results
Voltage Gain
The voltage gain of the
amplifier is less than 1. A small output voltage is ideal for low power
dissipation,
as power dissipated by a
resistor (the speaker) is directly proportional to the voltage dropped across
the
resistor squared. However,
higher voltage across the speaker will allow the speaker to output a louder
volume.
Current Gain
The current gain of the
amplifier is 333 (or 20mA / 60uA). This current gain is seemingly large.
However,
increasing the size of R1 will
increase the current gain of the circuit. More current through the speaker will
lead to more power dissipation,
as power dissipated by a resistor (the speaker) is directly proportional to the
current through the resistor
squared.
Power Dissipation
Power Dissipated by Speaker
Power Dissipated by MOSFETs
Power dissipation is plotted
above. The power dissipation of this amplifier is higher than the
maximum power dissipation characterizing
the MOSFETs in the datasheet, seen below.
The average power dissipation
of the PMOS is around 1.3 Watts, while the absolute maximum
rating from the datasheet is
625 mW. The average power dissipation of the NMOS is
around
0.92 Watts, while the absolute
maximum rating from the datasheet is 625 mW.
NMOS
PMOS
Ideally, we would want to
minimize power dissipation while maximizing volume and clarity to
the best of our ability.
Breadboard Implementation &
Experimentation
Operation at Frequency Limits Using Function
Generator
100 Hz, Gain of 0.14
20 kHz, Gain of 0.17
Probing Different Points for
Analysis
The starting point results did
not satisfy my partner and I. The voltage gain was
very low, and the volume
coming out of the speaker was
not very loud as a result. We could hardly hear the high-pitched buzzing of the
function generator, and
certainly could not hear music when we connected it.
In Lab 6, we derived the gain
of the push-pull amplifier by hand, realizing that when R1 is large, it can be
estimated to be directly
proportional to the voltage gain of the amplifier. See hand calculations
below.
In order to increase the voltage gain of our amplifier, thereby increasing the volume
out of our speaker,
we knew that we would need to
decrease our input resistance, and increase R1. The
input resistance and the
speaker act like a voltage
divider if there is no amplifier present. The input resistor is in place for
two reasons.
1.
It is used to limit the amount of current flowing into the
amplifier.
2.
It is used to drop the majority of the
input voltage since the gain of the push-pull amp is so large.
Therefore, decreasing the size
of the input resistor will increase the size of our output signal because less
voltage
will be dropped across R1 and
more will be dropped across the speaker. Also, increasing the size of R1 will
increase
our voltage gain.
This was our plan moving into the final design implementation.
-------------------------------------------------------------------------
Final Design
·
In designing the amplifier, my partner and I looked
into the tradeoffs associated with varying the input resistor
RS and the internal amplifier resistor R1 for the
majority of our considerations.
Tradeoffs:
o Increasing R1 increases voltage
gain, thereby increasing volume out of speaker.
o Increasing RS decreases current
in, decreases output voltage, decreasing volume and power consumption.
o Decreasing R1 decreases voltage
gain, decreases volume, but also decreases power consumption.
o Decreasing RS increases current
in, increases output voltage, increases volume and power consumption.
·
For optimal volume out, my partner and I decided to both decrease
RS and increase R1, thereby increasing the
power consumption greatly, but also increasing the volume and the
output signal greatly. The final schematic is below.
Associated
Calculations/Characteristics of Amplifier
Theoretical
Input Resistance
Theoretical
Output Resistance
(from spice)
(from
spice)
Theoretical
Power Dissipation of Speaker
Simulation Results
Voltage Gain
The plot above shows that the
theoretical gain of our newly designed amplifier is roughly 2.5. This
voltage gain is exactly what we
were looking for to crank up the volume of the output audio signal.
Current Gain
The current gain above is much
greater than the current gain of the original amplifier we started with
thanks to the modifications
made. Increasing the input current by just 10uA, the output current (in
comparison
with the starting amplifier) is
increasing by a factor of 6! The current gain of the amplifier is now huge
(over 1.7k).
Power Dissipation
Power Dissipated by Speaker
Power Dissipated by MOSFETs
The power dissipated by our speaker
is nearly 150 times greater than the power dissipated by the speaker
in the starting circuit. Power
dissipation is the largest characteristic we had to sacrifice for higher
volume.
Our speaker is dissipating even
more power than it was in the starting circuit, and our MOSFETs are
also dissipating slightly more
power on average than they were in the starting circuit. However, the volume is
much
louder because the gain is much
higher.
In the future, to improve our
design, we would like to optimize gain while also minimizing power dissipation.
Breadboard Implementation &
Experimentation
Click the image above for a
video of the amplifier operating.
iPhone Audio Jack (Music) Used as Input
Operation at Frequency Limits Using Function
Generator
100 Hz, Gain of 0.8
20 kHz, Gain of 1.15
Our theoretical gain is more than
double our experimental gain (likely due to imperfections in the
transistors and power supply at
high voltages), but the amplifier works great overthe
frequency range
100 Hz – 20 kHz. However, we
see some odd thickness in the oscilloscope signals.
This oddity is detailed below.
Noise From Power Supply
2Vpp, 1 kHz Input Signal, 5V
Power Supply Voltage
2Vpp, 1 kHz Input Signal, 10V
Power Supply Voltage
At higher power supply
voltages, there is a large amount of noise on both the input signal and the
output signal. Perhaps the
wires used in the design on the breadboard become better antennae when
the power supply voltage is
larger. Changing the scale on the oscilloscope, we can see that the signals
in the photo above (where power
supply voltage is 10V) are oscillating within the larger oscillation of the
desired signal. This is the
cause of the thickness in our waveforms at power supply voltage of 10.
----------------------------------------------------------------------------------------------
Conclusion
Our final design gave us the voltage
gain that we desired, and our output audio signal was loud and clear.
However, our transistors were
running very hot. We did not realize this until later on,
and we did not have
time to go back and make the
necessary modifications. Our goal was to minimize power while maximizing voltage
gain, but we were only able to meet
our voltage gain goal. In the future, we would improve the design by adding
transistors in series to reduce
the amount of current flow through the devices, thereby lowering the overall
power
dissipation of the devices. This
would also decrease our gain, and we would need to make design decisions to potentially
sacrifice some volume for lower
power consumption by the amplifier to ensure that our transistors can run for
long
periods of time without heating
up substantially.