Transcript Chap07--XTR & RCV Ci..

Transmitter and Receiver Circuits
MODULATOR CIRCUITS
 Solid-state AM modulators are usually biased class C, and
the output RF signal amplitude is modulated by causing
the collector supply voltage to vary with the modulation
signal.

As illustrated in Figure 6-19, the collector voltage is varied
by transformer-coupling (TA) the audio modulation signal
in series with Vcc.

The RF choke (RFC) and RF bypass capacitor (RF-BP)
isolate the carrier from the audio section and power supply
line.
1
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
Transmitter and Receiver Circuits
Figure 6-19. Transistor AM modulators.
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
2
Transmitter and Receiver Circuits
 If the RF amplifier of Figure 6-19 had a constant supply
voltage (nomodulation), the collector vote e would be Vcc.
 With RF input, the collector voltage can swing almost to
ground (less Vcc(sat)) on the negative swing and to 2V,
on the positive swing, thereby maintaining an average Vcc.
 For 100% modulation, however, the peak supply voltage
reaches 2Vcc, and therefore the positive peak RF swing can
reach 4Vcc, as seen in the collector waveform of Fig. 6-19.
 As a result, the choice of transistor for the AM modulator
must include high breakdown voltage. In particular,
BV > 4Vcc
(6-29)
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
3
Transmitter and Receiver Circuits
 Since an amplitude-modulated signal must maintain
amplitude linearity, any amplifier following the modulator
must be class A or, as is usually the case in transmitters,
class B push-pull.
 Figure 6-20 shows an all-solid-state transmitter output
with a tuned push-pull amplifier final. R1 can be adjusted
to bias Q3 and Q4 for class A or class B (actually, almost
class AB) operation.
 Sometimes R1 is replaced by a diode that will help trackout temperature variations of VBE for Q3 and Q4.
 Diodes D1 and D2 form a network to allow the driver Q1 to
be modulated on positive modulation peaks when Vc
exceeds Vcc.
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
4
Transmitter and Receiver Circuits
Figure 6-20. All-solid-state AM modulator/transmitter.
5
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
Nonlinearities in RF Mixers
 An RF mixer requires a nonlinear circuit component.
 Two broad categories cover the range of nonlinearities:
1). the general nonlinearity that is expressed
mathematically by a power series;
2). the nonlinearity produced in switching or sampling
mixers.
6
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
Nonlinearities in RF Mixers
 The output versus input nonlinearity of any device can be
expressed mathematically by a power series such as
io = Io +avi +bvi2 +cvi3 +… +nvin
(7-1)
illustrated in Figure 7-6.

Part (a) shows a mixer circuit with the LO and RF voltages
transformer-coupled to a high-frequency diode.

The currents produced in the diode produce an output
voltage vo a ioZ in the tuned-output circuit impedance Z.
7
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
Nonlinearities in RF Mixers
 Figure 7-6b shows the nonlinear input-output V-I curve
of the diode. Id is the bias current developed for the
purpose of placing the diode at the best operating point
for optimum-mixer performance.
 Curves 1, 2, and 3 show the linear (avi), 2nd-order (bvi2),
and 3rd-order (cvi3) nonlinearities at the bias point Id.
The constant a is the slope of the io curve at the bias
point, and similarly, constants b and c are scaling factors
for the first two nonlinearities.
 For a simple 2nd-order nonlinear device such as a FET,
io = Id + avi + bvi2
(7-2)
8
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
Nonlinearities in RF Mixers
Figure 7-6. Mixer circuit (a) and diode characteristic
curve (b) showing nonlinear input-output relationship.
9
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
Nonlinearities in RF Mixers
 The inputs vLO and vRF are linearly added by transformer T1
to produce
vi = vLO + vRF = coswLOt + AcoswRFt.
 Substituting for vi in Equation 7-2 gives
(7-3)
 The fourth term, which is a result of the 2nd-order (square
law) nonlinearity, can be expanded by a trigonometric
identity as
bcos2(wLOt) = (b/2)(1+cos(2wLOt))
= (b/2) + (b/2)cos(2wLOt)
 This shows that some dc and LO second-harmonic currents
are produced by distortion.
 Likewise the last term of Equ. (7-3) will yield dc and
RF second-harmonic currents.
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
10
Nonlinearities in RF Mixers
 These, together with the scaled input current (terms 2
and 3 of Equ. (7-3)), yield
io = Id + (input signal and 2nd harmonic current)
+ 2bA.cos(wLOt).cos(wRFt) ,
(7-4)
where Id = (Io +b/2 +bA2/2).
 The last term of Equation (7-4) is the mixer product of
interest,
2bA.cos(wLOt) x cos(wRFt).
11
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
Nonlinearities in RF Mixers
 By trigonometric identity,
2bA.cos(wLOt) x cos(wRFt)
= bA.cos(wLO + wRF)t + bA.cos(wLO - wRF)t

which shows that the product of two input sinusoidal
signals will yield sum and difference frequency currents
with amplitudes of one-half of the product term.

Replacing the product in Equ. (7-4) with this result gives
the total circuit current io.
 If the parallel LC circuit of Figure 7-6 is tuned to the
difference frequency wLO - wRF = 2p(fLO - fRF), then the
output voltage developed across the high-impedance tank
will be proportional to the RF input voltage but will have
an intermediate frequency of fIF = fLO - fRF.
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
12
RF Mixer Circuits
Figure 7-8. Mixer circuits (a)&(b).
13
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
RF Mixer Circuits
 In the simple diode mixer of (1), the LO is capacitively
coupled. C1 is always small in value because the LO
frequency is almost always higher than he RF;
 more importantly, the high impedance of the small C1
allows for isolation of the oscillator circuit from the mixer,
while T1 provides for impedance matching of the mixer to
the RF circuit.
 Mixer (2) is an active mixer in which a net conversion gain
is achieved. The base-emitter junction is driven into
nonlinear or switching operation by the large LO signal.
 Mixing occurs in the input junction, and the transistor
current gain and tuned-collector circuit produce power
gain at IF. The overall circuit gain is 6 dB less than the
circuit would produce with an IF signal as an input.
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
14
RF Mixer Circuits
Figure 7-8. Mixer circuits (c).
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
15
RF Mixer Circuits
 In Mixer (3) makes use of a dual-gate MOSFET to
achieve good isolation between the LO and RF input
circuits.
 Also illustrated in this VHF mixer for a TV tuner (RF
front-end) is a circuit connection for automatic gain
control (AGC).
16
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
RF Mixer Circuits
Figure 7-8. Mixer circuits (d).
17
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
RF Mixer Circuits
 Integrated circuits, such as the CA (or LM) 3028A
which operates to 120 MHz, also give good isolation
in this mixer configuration.
 As seen schematically in mixer (4), the 3028A has a
differential amplifier with a (normally) constant
current source Q1 which is modulated by the LO.
 Thus, sum and difference frequencies are produced.
This IC has numerous applications, including AM
modulator and cascade amplifier with AGC.
18
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
RF Mixer Circuits
Figure 7-8 Mixer circuits (e)&(f).
19
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
RF Mixer Circuits
 Circuit (5) is the double-balanced mixer which
achieves isolation between each of the three ports.
 Circuit (6) illustrates the frequency converter. The IF
tuned circuit has very low impedance at the much
higher LO frequency (due to C1), and C2 provides a
feedback path for the local oscillator.
 Also, the oscillator tuned circuit is essentially a shortcircuit to the IF due to L1; therefore the emitter circuit
has low impedance for good gain.
20
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
Balanced Mixers Application
21
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
Balanced Mixer
 A primary application of a mixer is to shift the spectrum of a
signal, usually down in frequency. This application is typically
encountered in equipment such as receivers, spectrum
analyzers, and radio repeaters.
 The careful balancing of the Schottky barrier diodes in the
10514 and 10534 Mixers provides excellent suppression of the
local oscillator and input frequencies at the output port. This
usually eliminates the need for a BPF following the mixer.
 Figure 1 shows typical mixer performance with -3 dBm input
at the RF (or "R") port and +7 dBm input at the local
oscillator (or "L") port.
 Conversion loss is typically less than 6 dB. Excess noise
contributed by the diodes is negligible above 50 kHz, so noise
figure is essentially equal to conversion loss.
22
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
Balanced Modulator
 The balanced modulator application is an extension of
the frequency conversion mode just described. Here,
however, the signal of interest is shifted up from base
band and centered on a suppressed carrier, typically at
radio frequencies.
 Both time and frequency domain presentations of the
balanced modulation output are shown in Figure 2.
 Balanced modulator operation is useful, along with the
proper filters, for generating the proper frequency
components for SSB-SC systems.
 It is also an inexpensive way to synthesize a two-tone RF
signal with stable frequency spacing from an audio and
an RF signal source.
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
23
Balanced Modulator
Figure 2. Frequency and time
domain presentations of balanced
modulator output. Note suppression
of carrier and absence of higher
order sidebands.
24
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
Amplitude Modulator
 The balanced modulator output can easily be converted
to a conventional amplitude modulated format by
reinserting the carrier at the output of the mixer. This
is illustrated in Figure 3.
 Another, although less desirable, way of re-establishing
the carrier is to unbalance the unit by applying a direct
current bias along with the modulation signal to the
"X" port.
 Generally, however, lower distortion and better control
of modulation index result if the carrier is reinserted
after the mixer.
25
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
Amplitude Modulator
Fig. 3. AM signal constructed by
reinserting carrier on balanced
modulator output. Time domain
display compared modulation
signal with AM signal envelope.
Block diagram illustrated method
for synthesizing such a signal.
26
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
Pulse Modulator
 The broad bandwidth of the 10514 and 10534 Mixers
allows the current controlled attenuation mode to be
extended to pulse modulation.
 Figure 4 shows both time and frequency domain
presentations of a 100 ns wide burst of a 50 MHz
carrier. Here power during burst is +4 dBm.
 Clean bursts with only minimal video feedthrough are
provided for RF levels between 0 and +10 dBm by
excellent bandwidth and balance.
 Generally envelope rise time and shape are preserved
if frequency spectrum of desired RF output is within
bandwidth of "L" and "R" ports (500 MHz for HP10514, and 150 MHz for HP-10534).
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
27
Pulse Modulator
Figure 4. Frequency and time domain
presentations of pulse modulated 50
MHz carrier. Pulse burst is 0.1 ms wide;
repetition rate is 1 MHz.
Note fast, clean turn-on and turn-off
characteristics provided by broad
bandwidth of control, or “X” port.
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
28
Phase Detector
 With the same frequency present at “R” and “L” ports,
the dc output from the “X” port indicates the phase
difference between the two input signals.
 As this phase difference varies from 0° to 90° to 180°, the
output will vary from a maximum negative value to zero
to a maximum positive value, following a cosine function.
A typical plot of this response is shown in Figure 5.
 The specified low frequency (1/f) noise, not usually given
for mixers, makes the HP-10514 and 10534 particularly
useful as phase detectors in phase lock loops.
 They are also valuable in determining the short term,
relative stability of high quality signal sources.
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
29
Phase Detector
Figure 5. Plot shows typical dc output
from “X” port when HP-10514/534 is
used as phase detector.
Low 1/f noise qualifies these mixers for
use in phase locked sources requiring
low fm or phase noise.
The accompanying block diagram
represents such a stabilized source.
30
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
Current Controlled Attenuator
 The HP-10514 and 10534 Mixers can also be used as
variable attenuators under the control of a direct
current input.
 The transfer function between “L” and “R” ports
depends on the current entering or leaving the normal
difference frequency (or "x") port.
 As shown in Figure 6, a control current of 10 mA will
give minimum attenuation, while greater than 40 dB
isolation occurs at zero current. Note the wide range of
linear attenuation and the low minimum loss.
31
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
Current Controlled Attenuator
Figure 6. Typical attenuation L to R vs. dc "X"
control current; fL = 30 MHz at 0 dBm.
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
32
Mixing Modulated Signals
 When a modulated signal is up or down converted in
a properly designed mixer the output will contain the
same modulation as at the input.
 For example, suppose a carrier fR is amplitudemodulated by a sinusoid of frequency fm.
 The time domain and frequency spectrum for this
AM signal are sketched in Figure 7-10 as the input to
a mixer.
 The (ideal) LO spectrum is also sketched with fL > fR.
This is high-side injection.
33
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
Mixing Modulated Signals
 The mixer output spectrum may be sketched by computing
the sum and difference frequencies between the LO and each
of the RF spectral components.
 Assigning numerical values to fR, fm, and fL will make this easy,
and if you want to portray accurate amplitudes, make the
amplitudes of each sum and difference output component
one-half that of the RF input spectrum.
 The results for the IF spectrum are shown at the filter output.
Please notice that the time sketch has the same AM index and
envelope rate (fm) and shape.
 Only the average frequency (fIF) and amplitudes (1/2) are
different from the RF input. The one-half amplitude (-6 dB)
is ideal for a second-order mixer nonlinearity; actual mixer
efficiencies will vary.
34
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
Mixing Modulated Signals
Figure 7-10. Down-conversion of AM signal (high-side
LO injection to mixer).
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
35
Tuning the Broadcast Receiver
 Figure 7-11 is the schematic of a simple but
representative AM superheterodyne receiver available
in kit form from Graymark Incorporated.
 The separate circuits do not track correctly the RF
resonant frequency and the LO frequency to produce
exactly 455 kHz.
 While the RF is being tuned from 540 to 1600 kHz
(3:1 frequency range), the LO must tune from 995 to
2055 kHz (2:1 frequency range).
36
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
Tuning the Broadcast Receiver
 A compromise front-end alignment scheme is used in order
to avoid sideband (therefore audio) distortion.
One procedure used for receiver alignment is as follows
(refer to Figure 7-11):
1). Disable the LO with a short to ground from the negative
lead of C2.
2). AC-couple a very low-level 455-kHz signal to the base of Q2
and tune T2 through T4 for a maximum level at the output
of T4. This aligns the narrowband IF amplifiers (use a sweep
generator if available).
3). Alternatively, AM the input generator and tune for the
maximum at the detector output. Keep the output as low as
possible by reducing the signal generator power to avoid
saturation.
37
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
Tuning the Broadcast Receiver
4). Remove the LO short; set the signal generator to 540
kHz and the radio tuning dial (C1 to the low end of its
range. Tune T1 for 995 kHz using a counter, or
maximize the receiver output at T4 or the detector
output (with AM).
5). Set the generator to 1600 kHz and C1 to the high end
of its range. Since C1 will have a minimum capacitance,
the RF and LO trimmers C1A and C1B can be tuned for
fLO = 2055 kHz and maximum receiver output. Keep
the signal generator power low to avoid saturation.
38
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
Tuning the Broadcast Receiver
Fig. 7-11. AM superheterodyne receiver.
39
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
AGC CIRCUITS
 AGC system is shown in our prototype AM receiver in
Figure 7-11, where the detector diode is reversed. The
controlled amplifier is an NPN transistor so that, as the
AGC voltage increases (more negative voltages), the Q3
bias current decreases.
 When the IF amplifier gain is controlled by a reduction
of current, the system is known as reverse-AGC.
Reverse-AGC is used in satellite and space transponders
or battery-operated receivers where minimum power
consumption is important.
 Otherwise, in ground stations and home broadcast
systems (AM, FM, and TV), forward-AGC is used.
40
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
AGC CIRCUITS
 With forward-AGC the amplifier bias point is set at IF,
as shown in Figure 7-26. As the received signal strength
increases the AGC voltage increases.
 The fed-back AGC voltage increases the already hightransistor bias current (Ip ～ 10 to 20 mA) so that the
base region is flooded with current, thereby spoiling the
transistor current gain b.
 Forward-AGC has the advantage of increasing the
amplifier power-handling capability as gain is reduced
in order to maintain the best overload characteristic on
strong signals when it is most needed.
41
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
AGC CIRCUITS
Figure 7-26. Transistor gain versus bias current.
Bias points for forward-and reverse-AGC.
42
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
AGC CIRCUITS
 Figure 7-27 shows an example of a circuit arrangement
that would have to operate as forward-AGC. With no
input signal the detector diode will be reverse-biased
due to the positive voltage established by the R1-R2
voltage divider.
 These two resistors, along with R3, set Q1 for high gain
with collector current in the range of IF in Figure 7-26.
As the positive peaks of the received signal at the
detector increase, the base bias voltage on Q1 increases.
 Hence, the bias must be must be set near IF for the gain
to be reduced for increasing input signal.
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
43
AGC CIRCUITS
FIGURE 7-27. AGC network.
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
44
AGC CIRCUITS
 Assume that the R1-R2 voltage divider sets the bias point
and AGC line to
R1Vcc/(R1+R2) = V1
(= VAGC)
 Not only does this establish the Q1 base bias voltage,
but it also sets AGC threshold.
 This is because the IF input to the detector must reach V1
+ Vd on peaks in order for the diode to start conducting
and the AGC voltage to begin rising.
 Vd is the diode cut-in voltage drop, which we can assume
to be approximately 0.2V for germanium diodes.
45
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
AGC CIRCUITS
 If V1 is already established by the choice of R1, R2 and Vcc,
then the detector input power required for the onset of
AGC (threshold) must be
Pi (threshold) = Vpk2/2Zi
= (Vi + 0.2)2/Rdc

(7-44)
Rdc is the dc resistance from the diode cathode to ground,
which is, for this circuit,
(7-45)
where (Ri)Q1 = (b+1)R3 is the dc input resistance of Q1.
46
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan
AGC CIRCUITS
 AGC of a transistor is a very nonlinear function. For
reverse-AGC the reduction in base voltage reduces emitter
current and, therefore, voltage gain due to the increase in re.
 This is approximately linear until VBE goes below about
0.55V. However, IF amplifiers are designed to amplify the
input power.
 At the same time that voltage gain is reduced due to
current-starving of the transistor, the current gain is also
decreasing---this is in itself a nonlinear function.
 Forward AGC relies on spoiling the b by flooding the
transistor base region with current. Therefore, putting a
value on AGC threshold is mostly empirical, which often
includes, a potentiometer in the circuit.
47
Prof. J.F. Huang, Fiber-Optic Communication Lab.
National Cheng Kung University, Taiwan