02b Rfamp PDF
02b Rfamp PDF
02b Rfamp PDF
FM - Radio Receiver
88-108 MHz
Hongwu Tong E98
Fredrik Sverin E99
2
Contents
1 INTRODUCTION ........................................................................................... 4
2 PROJECT DESIGN ........................................................................................ 5
2.1 TRANSISTOR SELECTION ........................................................................... 5
2.2 S-PARAMETERS MEASUREMENT............................................................... 7
2.3 PROPERTY OF THE TRANSISTOR ................................................................ 8
2.4 TRANSISTOR BIASING ............................................................................. 10
2.5 FREQUENCY SELECTION ......................................................................... 11
2.6 COMPLETE DESIGN CIRCUIT ................................................................... 12
3 MEASUREMENT AND RESULT .............................................................. 13
3.1 GAIN ....................................................................................................... 13
3.2 COMPRESSION POINT .............................................................................. 14
3.3 THIRD ORDER INTERCEPT POINT ............................................................ 15
3.4 NOISE FIGURE......................................................................................... 17
4 CONCLUSION.............................................................................................. 18
5 ACKNOWLEDGEMENT ............................................................................ 18
6 REFERENCES .............................................................................................. 18
3
1 Introduction
The input amplifier with RF selection (88-108 MHz) should have low
noise, high gain and frequency selection. The specification of the
amplifier is as follows:
4
2 Project Design
In the project, BFR92A transistor is used. It has high power gain, low
noise figure and low intermodulation distortion. To compromise
between gain and noise, an appropriate operating point should be first
considered. From figure 1 (gain as a function of collector current),
figure 2 (gain as a function of frequency) and figure 3 (minimum noise
figure as a function of frequency), an appropriate operating point was
decided. IC = 10mA, VCE = 10V. The values of Fmin and opt for the
operating point are not available in the datasheet, but from circles of
constant noise figure for other operating points, one can see that Fmin
in the project is between 1.7 dB and 2.4 dB.
5
Figure 2: Gain as a function of frequency
6
2.2 S-Parameters Measurement
After the decision of the operating point, the S-parameters of the
transistor have been measured. The setup of the measurement is shown
in figure 4.
BFR92A
From the Network Analyser, one can see the S-parameters from
frequency 75 MHz to 150 MHz. The S-parameters diagram is shown
in figure 5.
7
From the S-parameters of the transistor, the stability of the transistor at
the operating point can be checked. K = 0.41 and the |∆| = 0.64 at
frequency 100 MHz, where ∆ = S11 · S22 – S12 · S21 and
K = (1 - |S11|2 - |S22|2 + |∆|2) / (2 · |S12 · S21|). The transistor is only
conditionally stable at the operating point, since K < 1. The maximum
stable gain is calculated to ensure that the transistor is capable of
providing the gain specified above. Since GMSG = |S21 / S12| = 30.22 dB,
the gain of the amplifier could be at least |S21|2 = 25.83 dB, so the
specification could be fulfilled.
In order to find the stable areas for output and input, the point Γ = 0 is
chosen. If ΓS = 0, then ΓL = |S22| < 1. If ΓL = 0, then ΓS = |S11| < 1. The
stable areas for input and output can also be found in figure 6.
The input amplifier should have high gain and low noise, but the
maximum gain is seldom obtained for the same source impedance that
gives the minimum noise figure. Compromising between gain and
noise figure should be made. Different gain circles are simulated and
shown in figure 7.
8
Available gain circle at 28 dB
Since Fmin, opt and rn for the operating point are not available in the
technical data for BFR92A, some approximation values are chosen to
investigate the noise property of the transistor. Fmin = 2.30dB,
Γopt = 0.22 + 0.09i, rn = 0.440. Different noise circles are shown in
figure 8.
9
If one choose ΓS = 0.2517 - 0.4340i, ΓL will be in the output stable
area, which is shown in figure 9. ZL = 247 + 113i Ω.
ΓS
From this simulation, one can get some feelings about the transistor.
Some modification has been made, when real circuit is designed.
10
VCC +12V
RC
RB1
RB3
BFR92A
RB2
0
RL
11
2.6 Complete Design Circuit
The complete circuit is shown in figure 12. In order to isolate the
signal design from the bias design, a RFC (radio frequency choke) is
used, which provides a high impedance path at high frequencies.
12
3 Measurement and Result
3.1 Gain
Figure 13 shows a measurement of the amplifier circuit with a network
analyser from 60 MHz to 160 MHz. In the window showing the
parameter S21, there are two markers, one for 88 MHz, which are the
resonance frequency that are chosen, and the other is the mirror
frequency at 109.4 MHz. This resulted in a difference in gain of 9 dB.
In the specification it should have been 20 dB, but it wasn’t possible
for this construction to achieve. There was also a change in gain for S21
between different resonance frequencies, which are to be seen in table
1.
13
3.2 Compression Point
Checking of the design was also made. First there was a measurement
of the compression point at different frequencies, 88 MHz, 100 MHz,
and 108 MHz, which are to be seen in figure 14, 15, and 16
respectively, and the compression points for the different frequencies
can be viewed in table 2. These figures are measured with power
sweep in the network analyser.
14
Figure 16: Compression point for 108 MHz
Two strong signals f1 and f2 with equal magnitude are present in the
input of the amplifier. When the two signals are strong enough, 3rd
order intermodulation products will be appeared at the frequency
2f1 - f2 and 2f2 - f1, which are shown in figure 18.
.
15
f1 f2
2f2-f1
2f1-f2
There were two measurements for the intercept points. One is for
f1 = 90 MHz, f2 = 89 MHz (figure 19) and another one is for
f1 = 102 MHz, f2 = 101 MHz (figure 20), which were made by plotting
the output power verses the input power.
POUT
f1 = 90 MHz, f2 = 89 MHz
f1 or f2
IIP3
2f1 – f2 or 2f2 – f1
PIN
16
POUT
f1 or f2 IIP3
2f1 – f2 or 2f2 – f1
PIN
Figure 20: Third order intercepts point for 101 or 102 MHz
9.00 18.00
8.00 17.00
7.00 16.00
6.00 15.00
5.00 14.00
4.00 13.00
3.00 12.00
2.00 11.00
1.00 10.00
0.00 9.00
88 MHz 2 MHz / DIV 108 MHz
17
The amplifier was also tested when an antenna was connected to its
input. This gave a good result, because it received a lot of stations,
even some stations that are not very strong. There was some noise, but
it was not that easy to cancel.
4 Conclusion
The functionality of the amplifier was quite good for the FM-band. It
was possible to find many different Swedish stations. The gain was as
it should be, but the noise was a bit harder to cancel.
There were many things that could have been better, like the noise, but
the time wasn’t there. The noise parameter for the design was missing,
which did it harder to get the best performance. A way to decrease the
noise is to choose a lower IC.
A problem for the design was the magnitude of the mirror frequency
rejection, which should have been larger. It can be solved by adding
series resonant circuit for mirror frequencies.
5 Acknowledgement
We would like to say thank you to Göran Jönsson, who helped us with
his knowledge, and also the people from Ericsson, who had helped us
with experience from the real life.
6 References
[1] L.Sundström, H.Börjesson and G. Jönsson, ”Radio Electronics”, Lund
2001
[3] L.Sundström and G. Jönsson, ” Radio Electronics Formulas and Tables” , Lund
2001
18