DESIGN AND BUILD YOUR OWN AUDIO
AMPLIFIER
Reshed Hussein
Kevin Wade
Final Report
ENEE 417-0105
University of Maryland @ College Park
1
Introduction
The purpose of this project was to design and build your own audio amplifier
from scratch. Good audio amplifiers take a small input signal and magnify it several
times, without distorting it, so that it can be heard at the speakers. In order to design our
audio amplifier we used the scheme provided by professor Yang as a starting point, and
made a few changes in order to meet the required specifications of high input impedance,
low output impedance, and a gain of about 10.
The project was very instructive because it not only helped us solidify our
theoretical knowledge acquired in Microelectronics classes, but it also gave us an insight
into the procedures an engineer must take from a design on paper to a final working
product. Another benefit of this project is the satisfaction that is gained after completing
such a challenging task. Even though our design satisfies all requirements there are many
other ways to design a good audio amplifier; either by using op-amps or JFET instead of
the BJT used in our design.
The design of our audio amplifier was done on PSPICE, where we made sure that
all requirements were met. After completing the design we built the circuit on the bread
board and compared the observed DC voltages and currents with the ones obtained in
PSPICE. We then designed a PCB layout on Microsoft Paint and built the PCB board.
Few changes were made from the initial design in order to improve the performance of
our circuit. This report contains all design schematics and measurements, as well as
observed measurements done on the PCB board.
Circuit design
Our audio amplifier consists of two parts: a differential stage built using 2 BJTs
and an emitter follower. The differential pair is used to amplify the input signal while the
2
emitter follower is used as a power amp stage providing the necessary current to the load.
Figure 1 below shows the circuit built in PSPICE.
Figure 1: Final Audio Amplifier Schematic
3
Figure 2: DC Voltages in PSPICE
Figure 3: DC Currents in PSPICE.
4
Figure 2 and 3 above, show the various DC voltages and currents that were obtained in
PSPICE.
Figure 4 below shows the voltage gain as a function of frequency. From the figure we can
see that the gain is approximately 10 across the audio range (from 10 Hz to 20 KHz).
100
10
1.0
1.0Hz
10Hz
V(Vout)/ V(Vin:+)
100Hz
1.0KHz
10KHz
100KHz
Frequency
Figure 4: Voltage Gain as a function of Frequency
60d
40d
20d
0d
-20d
1.0Hz
10Hz
P(V(Vout)/V(Vin:+))
100Hz
1.0KHz
10KHz
100KHz
1.0MHz
Frequency
Figure 5: Phase as a function of Frequency
5
As it can be seen from Figure 6 below the input impedance in the audible frequency
range is approximately 100KΩ. In the ideal case the input impedance would be infinite,
but since our circuit is far from being ideal 100KΩ is an acceptable value.
1.0M
100K
10K
1.0Hz
10Hz
ABS(V(V5:+)/I(V5))
100Hz
1.0KHz
10KHz
100KHz
1.0MHz
Frequency
Figure 6: Input Impedance vs. Frequency (Amplitude)
180d
170d
160d
150d
140d
1.0Hz
10Hz
P(V(V5:+)/I(V5))
100Hz
1.0KHz
10KHz
100KHz
1.0MHz
Frequency
Figure 7: Input Impedance vs. Frequency (Phase)
As it can be seen from the figure below the output impedance is approximately zero,
which is the desired value.
6
9.0e-30
7.0e-30
5.0e-30
3.0e-30
1.0e-30
1.0Hz
10Hz
ABS(V(V6:+)/ I(V6))
100Hz
1.0KHz
10KHz
100KHz
10KHz
100KHz
Frequency
Figure 8: Output impedance vs. Frequency (Amplitude)
1.0ud
0.5ud
0d
-0.5ud
-1.0ud
1.0Hz
P(V(V6:+)/I(V6))
10Hz
100Hz
1.0KHz
Frequency
Figure 9: Output impedance vs. Frequency (Phase)
Figure 10 shows the input vs. the output as simulated in PSPICE.
7
10V
0V
-10V
0s
V(Vout)
0.5ms
V(V8:+)
1.0ms
1.5ms
2.0ms
2.5ms
3.0ms
3.5ms
4.0ms
Time
Figure 10: Input vs. Output
Figure 11 below shows the PCB layout for our circuit.
Figure 11: PCB layout
8
Although we came up with a scheme for the Tuner we were not able to test it or build it
due time restrictions. The figure below shows the schematic for the Tuner.
0
C3
0
R2
320
0
OS2
OUT
2
LM741
-
OS1
.1uF
5
R8
6
320
1
0
R5
10k
R6
320
OS2
OUT
2
0
LM741
R4
10k
0
R7
160 U6
3
+
C5
.1uF
7
+
7
U5
V+
320
3
-
4
R1
1k
1.5Vac
0Vdc
C2
.1uF
OS1
5
R10
6
1k
1
0
V-
0
R9
V+
.1uFC1
C6
4
V2
R3
160
V-
.1uF
V1
0Vdc
C4
100uF
0
Figure 12: Tuner Schematic
Experiment
The actual built circuit is shown in Figure 13. As it can be seen almost all
voltages are close to the ones simulated in PSPICE. The output voltage is -4.2 mV;
ideally this would be equal to zero but -4.2 mV is small enough not to burn the speaker.
There were several problems associated with the original circuit so we had to change the
biasing resistors in order to achieve the desired output. After obtaining a good result on
the bread board we built our circuit on the PCB board. The most challenging part of our
project was probably troubleshooting the PCB board. We had to build the circuit twice
and make sure that we did not have a cold solder anywhere. In fact after rebuilding the
circuit the second time the circuit was not working because one side of the 47uF
capacitor was not soldered correctly. One of our recommendations for future student is
that they should make sure that every single item is properly soldered.
9
0
12 V
V1
12V
R6
1k
0
R7
1.5k
Q2N3906
11.24 V
940 mV
Q4
1
Q3
R5
510
10.61 V
2
10.85V
TIP29
D1
1N4500
280 mV
R12
1
VOFF = 0
VAMPL = 1V
FREQ = 1000Hz
0
R4
100k
0
Q2
R9
-300 mV
R8
10k
Q2N3904
-950 mV
-1.01 V
R1
1k
R2
1k
100k
1
-280 mV
Q2N3904
2
C1
V32.2uF
Q1
2
1
310 mV
C2
47uF
D2
1N4500
8
-290 mV
R10
1
D3
1N4500
R13
R11
1k
10k
SET = 6.5k
0
-290
mV
TIP30
-920 mV
0
-1.70 V
R14
-4.2 mV
Q5
0
V2
12V
-12 V
Figure 13: Actual DC Voltages.
Figure 14: Gain Amplitude vs. Frequency.
10
Figure 15: Gain Phase vs. Frequency
Figure 14 and 15 represent the Amplitude and Phase of the Gain plotted against
the Frequency. The plots were done using LabView and as it can be seen the log-log plot
of the Amplitude increases smoothly and settles around 21; while the linear-log plot of
the phase has a steep slope, but since the circuit is far from being ideal this is an
acceptable result.
Input and Output Impedance
In order to measure the input impedance we used Kirchoff's current law after
inserting a 10K resistor at the input. The picture shown below shows the scheme that was
used:
V0
1Vac
0
V4
0
R15
Rin
10k
x
V1
11
The procedure we followed was to measure the V0 and V1 shown in the picture and use
the following equation to solve for Rin:
(V0 – V1) / 10K = V0 / Rin
Rin = (V0*10K) / (V0 – V1)
For input amplitude of 200mV and frequency of 1000 Hz the input impedance we
obtained was: Rin = (580 mV * 10K) / (630 mV – 580 mV) = 116 KΩ. This value is
very close to the one obtained using PSPICE which was 100KΩ.
The following table list input impedances measured at different frequencies.
Frequency (Hz)
Input Impedance (KΩ)
15
109
200
112
5400
115
10256
115
40000
116
Table 1: Input Impedance as a function of Frequency
The Output Impedance was measured using a similar scheme as the one used to measure
the input impedance. The pictures shown below show the scheme that was used:
V0
Rout
R1
x
y
V1
V0
Rout
R2
x
y
V2
0
0
By measuring V1 and V2 and using known resistor values R1 and R2, we plugged the
values in the following equations and solved them for Rout:
1. (V0 – V1) / Rout = V1 / R1
2. (V0 – V2) / Rout = V2 / R2
After equating equation 1 and 2, we solved them for Rout. The result we obtained is the
following:
Rout = (V1- V2) / [(V2/R2) – (V1/R1)]
12
Table 2 shows the Output Impedance for several values of input Frequency:
Freq (KHz)
R1(Ω)
V1(mV)
R2(Ω)
V2(mV)
Rout(Ω)
1.040
13.6
59.9
5.6
57.8
0.355
5.295
13.6
56.9
5.6
54.9
0.356
9.927
13.6
58.3
5.6
56.2
0.365
18.684
13.6
58.8
5.6
57.5
0.219
Table 2: Output Impedance as a function of Frequency
As it can be seen from the table above all Output Impedance values are very close to zero
as desired.
The following are pictures of the PCB circuit that we built.
Figure 16: Actual Audio Amplifier (top view)
13
Figure 17: Actual Audio Amplifier (side view)
Discussions
The main obstacle that we encountered with our circuit was not with the design
but with its real world application. We found that PSPICE can really only be a guide or
reference tool used to give the general expected results. When the design gives the
desired results in PSPICE, the challenge is achieving those results on the breadboard.
Once we built our circuit on the breadboard we spent a considerable amount of
time tweaking our design to get the desired gain at the output. Most of our problems
rooted in voltage and current biasing problems. We went through a series of labs testing
a seemingly infinite amount of resistor value configurations until we achieved a
functional output. The only remaining problem lied in an output voltage that was too
large. We remedied this problem by placing a potentiometer before the emitters in our
differential amp stage which made the granular adjustment needed to correct our circuits
biasing currents.
14
With pleasing breadboard results, we moved on to building our PC board. After
constructing a malfunctioning channel in route to our final working board, the main thing
we learned was that the physical quality of the board is what characterizes your results. It
is of optimal concern to draw a layout that allows clean copper paths followed by solid
solder connections in order to attain the breadboard results.
Overall, this lab gave us a real opportunity to experience what it is like to
design and build a real world circuit. It gave us the necessary benefit of realizing the
circuits we have been studying only in textbooks in previous classes at the University.
Future projects may have involved building amplifiers of different classes or selective
tuners like those used in AM/FM radios.
15