Note: Descriptions are shown in the official language in which they were submitted.
'1,64~6
~.~$~:335;~
BACKGROUND OF THE INVENTION
.
This invention relates to electronic devices for receiving a
plurality of radiated electromagnetic signals, filtering a
selectable channel of frequencies from the received signals, and
demodulating the signals of the selected filtered channel. More
particularly, the invention relates to television receivers.
Television receivers of the prior art include a radio frc-
quency (RF) section and an intermediate frequency (IF) scction.
The RF section includes RF filters which are tuned to coarsely
filter a band of channels centered about a manually selected
channel. The output of the RF filter couples to the input of an
RF amplifier. Typically, total gain through the RF scction is at least
20 dB to 30 dB.This gain increases the arnplitude of signals within
the selected channel and additionally makes the noise figure of the
system essentially independent of subsequent elements in the
receiver. The output of the RF amplifier couples to onc input of a
mixer, while a second input of the mixer receives mixing signals
of a selectable frequency. T~e selectable frequency is generated
such that the selected channel is frequency shifted to
approximately 45 MHZ. The output of the mixer couples to a
channel selectioD filter which provides a relatively hiyh
impedance path for frequencies outside of the selected channel,
and a relatively low impedance path for signals inside the
selected channel. Signals at the output of the channel selection
~ilter are therefore primarily comprised of frequencies within
the selected channel.
Each television channel contains audio in~ormation, video
information, and frame synchronizing information. The output
of the channel selection filter couples to an audio demodulator
,: .
~, .
J
. . .
T I - 6 4 o
which separates the audio information from the selected channel;
and the output of the audio demodulator couples to a speaker
which generates audible sounds. Similarly, ~he output of thc
channel selection filter couples to a video processing unit
which separates the video and frame synchronizing information
from the selected channel; and the output of the video processor
couples to a picture tube which converts the video and frame
synchronizing information to pictures.
As described a~ove, a basic function which all telcvision
receivers perform is to frcquency shift the selected channel from
an RF frequency to a predetermined IF frequcncy by a mixing
device. This mixing operation has been perfor~ed in the past by
a variety of nonlinear devices. These devices include vacuum
tu~es, diodes, MOSFET transistors, and bipolar transistors.
However, the prior art mixers also generate o~put terms which
are proportional to their inputs cubcd or raised to higher order
odd powers. In a tele~ision receiver, these odd power terms may
generate interfering signals in the desired channel. ~or example,
such interfering signals are generated when signals are present
in channels on one side of the scl~ctcd ch~el which are onc and
two channels removed from the selected channel. This phenomena
is known as intermodulation distortion. Similarly, the cubic
terms and higher order odd power terms yencrate interf-erin~ sig-
nals in the desired channel when a carrier with amplitude
modulation is present in anyone of the undesired channels. This
phenomena is known as cross modulation distortion. The
frequencies which are generated in a desired channel as a result
of intermodulation distortion or cross modulation distortion
cannot be separated from the information signals lying therein.
.
--2--
~$~
Thus, as the magnitude of the inte~fering frequencies increases,
perceptible picture distortion or sound distortion occurs.
A principal advantage of the television receiver
herein disclosed is that it has greatly reduced cross modulation
and intermodulation distortion. This is accomplished to a large
extent by means of a unique RF m~xer~ The mixer utilizes a
MESFET transistor which has almost perfect square law current-
voltage characteristics. Since the-~ESFET mixer has almost
perfect square law characteristics, the mixer introduces extremely
small distortion into the system. ln particular, the MESFET
mixer handles an interfering signal level of greater than ~6dBm
on its output with less than 1% cross modulation distortion and
-40dB intermodulation distortion products.
The square law current-voltage characteristics of a
MESFET transistor and their application to reduce intermodulation
and cross modulation distortion in a television receiver are not -
taught by the prior art. In the past, MESFET devices were used ;
primarily to achieve high speed operation. For example, they
were utilized in pico-second digital switching circuits. See,
for example, a paper by Cahen et al entitled "A Subnanosecond
Switching Circuit" which was presented at the IEEE International
Solid State Circuits Conference on February 14, 1974 at Orsay,
France. See also, for example, a paper entitled "X and KU Band
Amplifiers with GaAs Schottky Barrier Field Effect Transistors"
by Weiner Baechtold in the IEEE Journal of Solid State Circuits
Volume, SC-8, No. 1, February 1973. High speed digital switching
is combined with high speed linear amplification in a high speed
pulse amplitude modulation device that is described in the paper
entitled "Performance of Dual Gate GaAs MESFETS as Gain Contro]led
~.
!3~
Low Noise Amplifiers and High Speed Modulators" by Liechti in
the IEEE Transactions on ~licrowave Theory and Techniques Vol~
MIT-23, No. 6, June 1975. However, all of the devices in the
above cited references only utilize MESFETS for h;gh speed.
The television receiver herein disclosed is also
novel in that it includes a unique combination of surface wave
device technology with MESFET device technology in the RF-IF
section. A MESFET mixer provides frequencv shifting while
channel selectivity is provided by a single low loss surface
acoustic wave device (i.e. SWD) bandpass filter. This SWD
filter has a sharp passband - stop band transition. For example,
signals 1~5 MHZ above the sound carrier in the passband are
attenuated by greater than 65dB. Conversely, the filter has
a low insertion loss for signals in the passband. The preferred
, - .
embodiment is a unidirectional SWD filter which has an inband
insertion loss of less than 3.5dB,
~ ~:
The SWD filter also has a high center frequency and a
passband which is a small percentage of the center frequency.
Thus, the SWD filter is readily implemented with~piezoelectric
., ~
material having a relatively low coefficient of coupling.
~uartz has such a characteristic, and it is a preferred substrate
material. By comparison, prior art television receivers have
- . ,:
employed SWD filters but only at a much lower IF fre~uency of45
~'Z Accordingly, their passband was a relatively large
percentage of their center fre~uency and thus the devices
needed piezoelectric material having a relatively large co- -
efficient of coupling such as lithium niobate. Lithium niobate,
however, has piezoelectric characteristics which are highly
sensitive to temperature change in the 0-70 C temperature
range. As a result, they require
-4-
.~r
.
temp rature compensation circuitry. By comparison, the
piezoelectric characteristics of quartz are relatively
insensitive to t~mperature change in the 0~70 temperature
range; and thus no temperature compensation circuitry is
required.
Prior art television receivers include RF tuned filters before
the mixer to insure that image frequencies of the selected channel
are sufficien~ly attenuated at the mixer input so as to not
produce interfering mixer output signals. To accomplish this,
the bandwidth of the tuned RF filters are only several channels
wide, and the center frequency is adjusted to align with the
selected channel. RF filters which pass all of the channels of
the VHF and UHF band at one time cannot be utilized because image
frequencies would destroy reception in the selected channel.
To demonstrate the abovepoint, consider the following. VHF
television channels exist from 55.~5 ~HZ to 71.75 MHZ, and
from 177.25 MHZ to 215.75 M~Z; while UHF channels exist from
471.25 MHZ to 889.75 MHZ. Also, as is known in the art, the
image frequencies of a selected channel lie at 2XlF above the
selected channel. Thus, channel 6, for example, which is
.
centered at 87075 MHZ, has an image frequency in a conventional
television receiver at (87.75 +90) MHZ. This equals 177~75 MHZ-
which is within channel 7. Similarly, the picture carrier for u
ch~nnel 14 is 471 MHZ; and thus its image ~requency is 561 MHZ-
which is within channel 29. Accordingly, prior art receivers
need an RF tuner before the mixer to filter image frequencies.
- ~ /
35~i
In the prior art, RF tuning is accomplished by varactor
ilters, mechanically variable capacitors, ctc. Ilowcvcr, thesc
are both expensive and difficult to align. By comparison, the
television receiver o~ the disclosed invention has no tuners in
the RF section. In one preferred embodiment, ~ fixcd bandpass
filter is included which pass~s the entire low VHF band, a
second filter is included which passes the entire high V~-lF band,
and a third filter is included which passes the entire UHF
band. These flxed fiLters are simple in design and eliminate
alignment problems. The disclosed in~ention also includes
a mixer having an IF output fre~uency of between 300 MHZ and
400 MHZ. As a result, image frequencies of the selected
channel are placed at least 600 MHZ above the selected channeL
where they are easily rejected by the fixed bandpass RF filters.
The mixer output of the disclosed receiver couples
to a channel selection filter which is implemented by an
improved surface wave device. Prior art television receivers
also used surface wave device channel selection filters. ~owever,
the SWD filter of the disclosed inve~tion is an improvement
from the prior art in that it is constructed on a relatively
small substrate area. The area used by a SWD filter is
proportional to its center frequency. Conventi ~ 1 receivers
have an IF frequency of 45 MHZ, and thus, the SWD's used
therein require substantially more space.
6-
.
Another novel aspect of the disclosed system is that the
IF section includes two mixers at the output of the SWD
filter for ~requency shifti~g the selected channel back to
baseband. The irst IF mixer frequency shifts the selected
channel to approximately 45 MHZ. Most of the gain of the
system is then added to the selected channel signals. The
second IF mixer is utilized to synchronously detect signals in
the selected channel. This architecture permits the described
high IF system advantages to be achieved, withou~ introducing
high frequency feedback problems.
As previously described, prior art television receivers
insert at least 2Q to 30 dB of gain in the RF section to achieve
a low system noise figureO System noise~fi~ure~equals
NFl N3~2-1 ~ NF2-1 + NF2-1 + ...
Gl Gl G2 -- -
wherein NFi and Gi are the nQise figure and~ain of the ith
functional block in the system. Thus, inserting a large gain
in the first functional blocks (i.e. the RF section) lowers
system noise figure by making it independent of noi~e figure
of subsequent circuitry.
The disclosed invention has a unique architecture which
achieves;low noise figure, low in~ermodulation distortion and
low cross-modulation distortion simultaneously. ~ In the
preferred embodiment, the disclosed invention has a maximum
gain in the channel selecting section which is less than 10 dB;
that utilizes circuit elements therein which individually have
low noi~e figures yield a low system noise figure, while the
low gain allows RF and IF circuit elements to operate within their
dynamic range without generating high odd order output terms.
~ J3~
The architecture of the disclosed invention is also
novel in that it has two IF frequencies about which the system
gain is selectively distributed. In one preferred embodiment,
a irst mixer shifts the selected channel to about 330 MHz. The
output o~ this mixer couples to the channel selection filter;
and the output of the channel selection filter couples to a second
mixer which frequency shifts the filtered selected channel to a
second IF frequency of approximately 45 MHz. Most of ~he gain of
the system (approximately 60 dB) is inserted at the lower IF
frequency after the channel selection ilter. Gain at the high
IF is small as pointed out above. As a result, several advantages
of a high IF system - such as simplified filtering of image
frequencies-are~obtained, while feedback inherent in a high
gain-high IF section is avoided.
The disclosed system is also simpler and potentially
less expensive than prior art television receivers in that the
low gain RF section makes feasible integration of a major
portion of the receiver on a single semiconductor chip.
Incorporated in the disclosure are two differently organized
~E-IF sections which are suitable for semiconductor chip
integration.
SUMMARY OF_THE INVENTION c~ s~ I Pc to~
In one aspect o the invention, a ~ n r
is provided which includes a MESEET mixer in combination with
an acoustic surface wa~e device filter. The MESFET mixer
simultaneously receives a plurality of channels and mixing signals
of a selectable frequency, and in response thereto frequency
shifts sele~table channels to a predetermined center frequency.
3~
The ~SFET mixer has almost pe~fect square law characteristics,
and accordingly introduces extremely small distortion into
the system. In one embodiment, the MESFET mixer handles
interfering signal levels of greater than -~6dBm on its output
with less than 1% cross-modulation distortion and -40dB
intermodulation distortion products. The acoustic surface
wave device (SWD) filter has an input coupled to the output
of the MESFET mixer. The SWD filter provides essentially all
of the channel selectivity of the receiver. In a preferred
embodiment, the ST~D filter is a three-phase low-loss unidir-
ectional filter having a finger withdrawal type impulse response.This SWD has an in-band insertlon loss of < 3.5dB, and a stop-
band loss at 1.5MHZ above and below the passband of > 65dB.
.
." ,
" :
, ~, , . :. , ,, :
3~i~
:BRIEF DESCRIPTION OF THE DR~ INGS
I
The novel features believed characteristic of the
invention are set forth in the appended claims; the invention
itself, however, as well as other features and advantages
thereof, may best be understood by referring to the following
detailed description of particular embodiments when read in
reference to the accompanying drawings, wherein:
~,
'
.
:
--10--
35~
FIGURE ]. is a block diagram illustrating a channel
selector constructed according to the invention.
FIGURES 2a-2e are a series of frequency diagrams
illustrating signals at selected points on the channel
selector of FIGURE 1.
FIGURE 3 is a timing diagram illustrating the operation
of a phase locked loop within the channel selector of
FIGURE 1.
FIGURES 4a and 4b are graphs, respectively,
illustrating the amplitude of in-band and out-of-band
signals at various points in the channel selector of
FIGURE 1.
~ FIGURES 5a, 5b and 5c are detailed circuit diagrams of
; RF filters within the channel selector of FIGURE 1. -
FIGURE 6 is a detailed circuit diagram of a two-by-one
switch within the channel selector of FIGURE 1.
FIGURE 7 is a detailed circuit diagram of an RF
amplifier within the channel selector of FIGURE 1~ ~
FIGURE 8 is a detailed circuit diagram of a MESFET ~-
mixer within the channel selector of FIGURE 1.
FIGURE 9a is a plan view of a MESFET transistor
suitable for application in the MESFET mixer of FIGURE 8.
FIGVRE 9b is an enlarged cross-sectional view taken
along the line A-A of FIGURE 9 a.
FIGVRE 9c is a greatly enlarged cross-sectional view
of a dual gate MESFET device showing one source-drain pair
thereof and which comprises another MESFET transistor suit-
able for application in the MESFET amplifier of FIGURE 7.
FIGURES 9d and 9e are graphs illustrating the
operational characteristics of the MESFET transistor shown
in FIGURE 9a.
-- 11 --
.
FIGURE 10 is a circuit diagram of a second two-by-one
switch within the channel selector of FIGURE 1.
FIGURE lla is a circuit diagram of a surface wave
devi.ce acoustic filter within the channel selector of
FIGVRE 1.
; FIGURE llb is a graph illustrating the magnitude-
frequency characteristic of the surface wave device
included in the filter of FIGURE lla.
FIGURE llc is a perspective view illustrating one ~.
embodiment of the surface wave device included in the
filter of FIGURE lla.
~FIGURE lld is a partially perspective diagrammatic
;view illustrating the surface wave device included in the
filter of FIGURE lla wherein the impulse response of the
surface wave device has been shaped by electrode finger
withdrawal
FIGURE lle is a diagram of a phase shifting circuit .-
for use in the input circuitry o the filter of FIGURE lla.
FIGURE llf is a phase diagram illustrating how the
20 phase shif-ting circuit of FIGURE lle can produce a 120 :~
phase shift.
FIGURE llg is a diagram of a circuit for producing a
60 phase lag. ..
FIGURE llh is a circuit diagram of the input circuitry ~ -
included in the filter of FIGURE lla.
FIGURES 12a-12b are detailed circuit diagrams of a
linear amplifier within the channel selector of FIGURE 1.
FIGURE 13 is a detailed circuit diagram of a mixer
within the channel selector of FIGURE 1.
30FIGURE 14a is a detailed circuit diagram of a surface
wave device oscillator within the channel selector of
- 12 -
~,
3~
FIGURE 1.
FIGURE 14b is a top plan view illustrating the surface
wave device included in the oscillator of FIGURE 14a. :
FIGURE 15 is a detailed eircuit diagram of a high gain
linear amplifier within the channel selector of FIGURE 1.
FIGURE 16 is a detailed eircuit diagram of a
synchronous detector within the ehannel selector of FIGURE
FIGURE 17 is a detailed circuit diagram of a L-C ~ ~ -
oscillator within the channel selector of FIGURE 1.
FIGURE 18 is a detailed circult diagram of a phase ` :
detector, a ramp-generator, and a loop filter within the
channel selector of FIGURE 1.
FIGURE 19 is a detailed circuit diagram of a voltage .
controlled oscillator within the channel seleetor of .
FIGURE 1.
FIGURE 20 iq a block diagram of a television reeeiver
whieh includes the`ehannel seleetor of FIGURE 1.
FIGURE 21 is a bloek diagram of an alternative
embodiment for the channel seleetor of FIGURE 1.
- 12a -
DETAILED DESCRIPTION OF SPECIFIC EMBODIMENTS
Referring to Figure 1, the channel selecting portion
of a television receiver constructed according to the invention
is illustrated in block diagram form. The selector includes
a VHF antenna 10 having an output coupled via a lead 11 to a
fixed bandpass filter 12 and to another fixed bandpass filter
130 Signals on lead 11 are herein designated Sl(f). Filter
12 has a center freauency of 69 MHZ and a 3dB bandwidth of
34.6 MHZ to thereby permit passage of the low VHF television
channels. Similarly, filter 13 has a center frequency of
193 MHZ and a 3dB bandwidth of 44.5 ~IZ to provide filtering -
of the high VHF television channels. The output of filter 12
is coupled via a lead 14 to~one of the inputs of a 2 x 1 switch
15, while the output of filter 13 is coupled via a lead 16 ~-
to a second input of switch 15. Switch 15 operates to select
either signals from filter 12 or 13. The filter-switch
combination inserts approximately a 1 dB loss on signals in -
the passbands.
The output of switch 15 is coupled to an RF amplifier
17 via lead 18. Amplifier 17 has a varia~le gain which is con- ~;
trolled by signals on an AGC line 19. The m~m gain of ampli~
fier 17 is approximately 4 dB. The noise figure of the
amplifier is approximately 3dB.
The output of amplifier 17 couples to one input of a
MESFET mixer 20 via a lead 21. Mixer 20 has a fixed gain of
approximately 4dB, and a noise figure of approximately 8 dB.
A second input of mixer 20 is coupled to a VHF voltage controlled
oscillator 22 via a lead 23. Oscillator 22 generates local
oscillator (LO) signals on lead 23 of a selectable frequency -
in the range 385-541 M~IZ. In response thereto, mixer 20
frequency shifts the RF signals on lead 21 to a new IF frequency
-13-
~r
. . ~ . .
'3^~
range. The frequency of the LO signals on lead 23 is selected
such that the channel to be received is frequency shifted to
a predetermined high IF of between 300 MHZ to 400 M~IZ. In
one embodiment, this predetermined high IF is approximately
330 MHZ. The frequency shifted signals are generated on a
lead 24 and are designated S2(f).
-13a-
"~ ~,
Lead 24 couples to one input of a 2 x 1 switch 25, while
a second input of switch 25 is coupled via a lead 26 to receive
frequency shifted UHF television channels. Switch 25 is identical
in construction to switch 15. The circuitry for frequency
shifting the UHF television channels, which is labeled as UHF RF
SECTION, is similar in construction to the VHF-RF section, and is
described in detail infra.
The output of switch 25 is coupled to a surface wave
device (SWD) filter 28 via a lead 27~ Filter 28 has a passband
which is shaped to pass only one of the television channels on
lead 27. In particular, filter 28 passes th~ television channel
at the predetermined high IF. The fixed RF filters in
combination wi~h the SWD filter provide essentially all the
filtering in the system. In a preferred embodiment, filter 28
B is a three phase unidirectional filter which has ~ insertion
loss. Typically, the 108s through ilte~ 28 is less than 3.5 dB
in the passband. Conversely, out of band signals are greatly attenuated.
The output of filter 28 is coupled to an IF amplifier
29 via a lead 30. ~he signals on lead 30 are designated herein
as S3 (f) . Amplifier 29 has a variable gain which is controlled
by AGC signals on a lead 31. The maximum gain of amplifier 29
is ~pproximately 30 dB. Thust amplifier 29 is the first high
gain element in the system. To this point, the gain of the system
is characterized as belng no larger than necessary to obtain the
desired system noise figure. Noise figure for amplifier 29 is
approximately 4 dB.
: The output of amplifier 29 couples to a mixer 32 via a
lead 33. A second input of mixer 3~ is coupled via a lead 34 to
the output of an oscillator 35. Osclllator 35 has a surface wave
device resonator 36 as its frequency controllins element. In a
preferred embodiment, resonator 36 and oscillator 35 operate to
generate a mixing signal on lead 34 having a frequency o~ 285 MHZ.
-14-
.'3~6
The 285 MHZ signal on lead 34 is mixed with the si~nal
on lead 33 to thereby generate signals S4(f) on the output of
mixer 32. Signal S4(f) is similar to signal S3(f) except that
it is shifted down in fre~uency by 285 MHZ. Thus it has a picture
c rrier at approximately 45 MHZ.
Additional gain is add~d to the signal of the selected
channel after it has been frequency shifted to the lower IF fre-
quency of 45 M~Z. Mixer 32 adds a fixed ~ain of +10 dB~ }
the output of mixer 32 couples via a lead 37 to an IF amplifier
38 which has a maxLmum gain of +50 dB. The gain of amplifier 38
is ~aried by an AGC signal on a lead 39. By inserting most of
the gain at the lower IF frequency, system stability is increased
since feedback through parasitic capacitances, radiation, etc. is
much less at 45 MHZ than 330 MHZ.
The output of amplifler ~ is coupled via lead 40 to
a tank circuit 41, a synchronous detector 42 and the phase
detector 43. Tank circuit 41 has a center ~requency of approximately
4S MHZ. Synchronous detector 42 has a second input coupled via a.
lead 44 to an oscillator 45~ Oscillator 45 generates clock signals
at a fixed frequency of 45 MHZ on lead 44. The signals on lead 44
are in phase with the 45 MHZ picture carrier on lead 40.
Detector 42 mixes the signals on leads 40 and 44 to generate output
signals S5(f) on a lead 46. Signal S5(f) contains the selected
television channel with the picture carrier at zero Hz and the
sound carrier at 4.5 MHæ. ~ead 46 is then coupled to conventional
televis~on circuitry for sepaxating the sound signal from the
picture signals and for reproducing the sound and the picture from
these si~nals respéctively in a conventional manner.
The synchronous clock signal on lead 44 is kept in
phase with the picture carrier on lead 40 by means of the phase
detector 43 operating in conjunction with oscillator 45. Phase
detector 43 generates phase detection signals PD2 on lead 48 which
indicate the phase difference between the picture carrier on lead
40 and the oscillator signal on lead 47~ Signals PD2 maintain a
90~ phase difference between the signals on leads 40 and 47. This
phase difference is compensated for by oscillator 45 which maintains
a compensating 90 phase difference between the oscillator signals
on leads 44 and 47.
To complete the phase locked loop, lead 48 couples to
a loop filter 49, and he output of filter 49 couples to a summer
50 via leads 51. Summer 50 has a second input coupled to a con~
troller 52 via a lead 53, and a third input coupled to a ramp
oscillator 54 via lead 55. ControlLer 52 and oscillator 54 operate
to provide a coarse voltage for selecting one channel ~rom another.
Channel selection switches 56 are coupled to controller 52 via
leads 57. Switches 56 generate digital signals on leads 57 indicating
the selected channel. Controller 52 contains 2 digital to analog
converter which generates a coarse channel selection voltage on
lead 53 in response to the digital signals. The signals on lead 53
are summed with the coarse signais on leads 51 and 55 to thereby
pro~ide a loop which is phase locked to the picture carrier of the
selected channel. The output of summer 50 is coupled to VHF voltage
controlled oscillator 22 via leads 58. Oscillator 22 generates LO
signals on lead 23 of a ~requency ranging from 385 MHZ to 541 MHZ
i~ response~to the phase detection signals on leads 58 thereby
completing the loop.
The UHF-RF section of the FIGURE 1 channel selector begins
with a UHF antenna 59. Antenna 59. is coupled to a high pass filter 60
via a-lead 61. Filter 60 has a fixed 3dB cutoff frequency of
approximately 380 MHZ. The output of filter 60 couples to the input
: of an RF amplifier 6Z via a lead 63. Amplifier 62 also has an AGC
input which is connected to receive AGC signals on lead 19. The
-16-
maxLmum gain of amplifier 62 is 4dB, and its noise figure is also
approximately 4dB. A low pass filter 65 has its input coupled
to the output of amplifier 62. Filter ~5 has a fixed 3dB cutoff
frequency of approximately 936 MHZ. Thus, filter 65 in combination
with filter 60, pass the entire UHF band and reject other
frequencies.
A mixer 67 receives the UHF band of signals from filter 65
via a lead 66. Mixer 67 simultaneously receives mixing signals of
a selectable frequency from a UHF voltaye controlled oscillator 68.
These selectable frequencies range from approximately 801 MHZ to
1215 MHZ. The particular frequency at any time instant is
generated in response to channel selection switches 56 so as to
frequency shift the selected channel to-the predetermined high
IF.
The overal1 oper.ation of the above described FIGURE 1
structure is as follows. Antennas lO,and 59 receive radiated
electromagnetic signals which inc:lude the VHF and UHF frequency
spectrum. The signals received by antenna 10 are applied to
non-tuned filters 12 and 13 which respectively pass the entire
low VHF and high ~F frequency spectrum. Similarly, the signals
received by antenna 59 are applied to non-tuned.filbers 60 and 65
5~;t~s
which pass the entire UHF frequency spectrum. Sw~*c~ 15 and 25
select one of these three frequency spectrums in response to logic
signals from the channel selection switches 56.
~ ESFET mixers 20 or 67 then frequency shift the selec~ed
frequency spec~rum ~o a predetermined IF in the 300 MHZ-400 MHZ
range. Mixer 20 frequency shi-fts VHF signals, whereas mixer 67
frequency shifts UHF signals. Both mixers have a large dynamic
range o~er which their output is an almost perfect product of
-17-
their inputs. As a re~ult, the system has improved performance.
For example, mixers 20 and 67 can handle >+6dBm interfering
signal levels on their output with less than 1% cross-modulation
distortion.
. Switch 25 then couples one of the mixers to the input of
SWD filter 28. Filter 28 ~reatly attenuates signals outside the
selected channel. In particular, the stopband of filter 28 is
notched such that the lower adjacent sound carrier and upper
adjacent picture carrier are attenuated by more than 6SdB. All
other out of band signals are attenuated by at least 55dB.
Conversely, in band insertion loss of ilter 28 is onIy 3.5dB.
Output signals from filter 28 are sent to amplifier 29-which
is the first high gain component of the system. Amplifier 29 has
a maximum gain of 30dB. In comparison, the total gain of the
system before amplifier ~9 is less than lQdB. As a result of the
low RF-IF gain, intermodulation distortion and cross modulation
distoxtion is decreased. Further, system stability is improved
since high requency feedback due to parasitic capacitance,
radiation, etc. is avoided. At the same time,~low system noise
igure is achleved ~y circuit elements which indlvidually have a
low noise flgure in combination with a sufficiently high gain.
- The low gain RF-IF section aLso makes feasible integration
of all of the channel selectors, except the fixed filters, on a
single semiconductor chip. An outline of such a chip t S
designated by the dashed line 69 in FIGU~E 1. The process f~r
fabricating the chip includes a combination of steps presently
well known for constructing MESFET devices and bipolar devices. A
second chip is utilized to construct SWD filter 28 and S~D
3~
resonator 36. The fixed filters 12, 13, 60 and 65 are construc.ted
of discrete components.
Signals in the selected channel at the output of amplifier
29 are frequency shifted to 45 MHZ by mixer 32. Then they are
further amplified to -lOdBm by amplifier 38. The selected channel
video signals are then reduced to baseband by synchronous
detector 42; while the sound carrier of the selected channel is
shifted to 4.5 MHZ. A 4.5 MHZ trap 70 removes the sound from the
video signals, and a low pass filter 71 removes all signals
except the video of the selected channel~ The video signals at
the output of filter 71 are sent via lead 72 to video processing
circuitry, while the audio ignals at ~.5 MHZ on lead 46 are
sent to audio processing circuitry. This audio-video circuitry
is described infra in conjunction with FIGURE 20.
Referring now to FIGURES 2a-2e, there is illustrated a
set of frequency diagrams of signals Sl(f) - S5(f). In FIGURE
2a, the low VHF band is shown generally at 75a,the high VHF band
is shown generally at 75brand the UHF band is shown at 75c. Each
of the bands is comprised of a plurality of channels; and each
channel has a ~requency spectrum which is assigned as illustrated
in detail ~t 76~ The frequency allocation and type of modulation
of signals within each channel is a well known standard that is
fixed by the FCC. FIGURæ 2b is an exemplary frequency diagram of
signal S2(f)- In the example illustrated, signàl S2(f) contains
channels in the low VHF spectrum. The selected channel is ne.ar
330 MHZ. The frequency of the LO signal on lead 23 minus the
frequency of the picture carxier from the selected channel equals
330 MHZ. Since the mixing frequency is higher than the picture
carrierl the frequency spectrum at the output of mixer 20 is
inverted from the input frequency spectrum as shown at 77.
FIGURE 2c illustrates the frequency spectrum of signal
---- 19
_ .
s3(f). Signal S3(f) is the output of the channel selecting
filter 280 Thus it basically contains only frequencies within
the selected channel. Signal S3(f) is then amplified and frequency
shifted down by 285 MHZ. The result is signal s4(f) as illustrated
in FIGURE 2d.
Signal S4~f) is further amplified, and then synchronously
. detected by detector 42. These operations producP signal S5(f)
as illustrated in FIGU~E 2e. Note that the mlxing action of
synchronous detector 42 again inverts the frequency spectrum of
the seIected channel~ Thus, the picture carrier of the selected
channel is at 0 HZ, and the sound carrier of the selected channel
is at 4.5 M~Z as shown at 78.
FIGURE 3 is a timing diagram illustrating the operation
of the channel selecting phase locked loop of FIGUR~ 1. As shown
therein, signal PD4 on lead 58 is comprised of components PDl-PD3.
Signal PD3 constitutes a coarse channel selection voltage which
is produced by controller 52. Signa:Ls PD2 and PDl provide the
fine tuning for the phase locked loop. In the example of FIGURE
3, one particular channel'is selected duri'ng a first time inter-
valQTl, while another channel is selected duxing a'time interval
T2. Signal PD3 provides the coarse voltage for channel selection.
Signal PDl compensates for any DC offset between signal PD3 and
,a~ .
the desired voltage level.,~ signal PD2 provides a dynamic
correction voltage to eompensate for instantaneous phase or
frequency differences between the signals on leads 40 and 44.
F IG U~e~s 2,a ~,e!.
The magnitude of,the signals of ~ are illus~
trated in FIGURES.4a and 4b. FIGURE 4a illustrates the magnitude
of signals in the desired channel, wherPas FIGURE 4b illustrates
~he magnitude of interfering signals in a channel that is two
.channels removed. Curves 81-84 iIlustrate the magnitude of
signals in the ~esired channel at various points in the system
when the incoming signal strength is 0 dBM, -35 dBM, -55dBM,
and -85 dBM respectively. As these curves illustrate, gain is
first added by amplifier 38, then by amplifier 29, and finally
i35~
by amplifier 17 as the input signal strength in the selected
channel decreases. In particular, the RF section has no gain
unless the input signal strength in the selected channel is less
than -55 dBm. And the RF section reaches its maximum gain of +3
dBm when the input signal strength of the selected channel is
between -55 d~m and -85 dBm
As previously pointed out, the low gain RF section in
combination with the MESFET RF amplifiers and MESFET RF mixer
provides a system having superior channel discrimation capability.
This is exemplified in FIGURE 4b by the relative signal strength
`, of the selected and unselected channels. For example, curve 85
of FIGURE 4b illustrates the case where the input signal strength
of the selected channel is~-55 dE~ and input signal strength two
- channels removed is +1 dBm. Similarly, curve 86 illustrates
B the case where the input,signal strength in the selected channel is
-35'd~m and the input signal strengt:h two channels removed is
+2 dBm. The most stringent re~uirement for the receiver is
when the ~irst RF amplifier,xequires gain (i.e., when the signal
strength of the ~esired channel i5 less than -55 dBm). This is
because the added gain increases nonlinearities in the RF section
and aggravates the cross modulation and intermodulation dis-
tortion~ Thus, curve 85 illustrates the most stringent condition
for the system. Under the conditions of curve 85, RF amplifier
17 and mixer 20 must have cross modulation distortion and inter-
Th,s
modulation distortion of less than 1%. ~ c~i~ requirement is
' met, since the MESFET designs herein described have less than 1%
a~l
cross modulation distortion~intermodulation distortion when
their output signal level is less than +6 dgm.
After the selected channel has been shifted to 330
, MHZ~signals two, channels removed rom the selected channel are
greatly-attenuated. Mixer 20 has a tuned output which attenuates
~21-
3~;~
two channel removed signals by -~ DB. SWD filter 28 atten
uates all out of band signals by at least -53 DB. At the
same time, mixer 20 adds 4 DB of gain to the desired signal,
and filter 28 inserts only 3.5 DB of loss to signals in the
selected channel.
The details of the various block of FIGURE 1 will
now be descri~ed in conjunction with FIGURES 5-19. Referring
first to FIGURE 5a, a circuit diagram of bandpass filter 12
is therein illustratedO Basically, filter 12 is comprised
of two series resonant LC circuits 91 and 92 and one parallel
resonant LC cixcuit 93. Filter 13 is similarly structured.
A primary function of RF filters 12 and 13 is to
pass one band of channels while rejecting image frequencies
by an amount sufficient to eliminate perceptible picture
interferenceD Since this system uses a 330 MHZ IF, all image
frequencies are 660 MHZ above the frequency of the desired
channel. Thus, the low VHF images are rejected by the three
poll bandpass filter of FIGURE 5a by greater than 80 DB. In
the system of FIGURE 1, just perceptible picture interference
occurs when image frequency signals are less than 36 DB below
the level of the desired signal at the filter output. Thus,
for example, the filter of FIGURE 5a provides adequate image
frequency rejection when the desired picture signal level is ;~
at -55 DBM and the image frequency level is at -11 DBM. Image
frequencies for the low VHF and the high VHF spectrum lie
within the UHF spectrum, and the level of UHF signal at the
input of the V~IF antenna can normally be expected to be less
than -11 DBM.
FIGURES 5b and Sc are circuit diagrams of a high
pass filter and a low pass filter suitable for use by the UHF
radio frequency section of FIGURE 1. All of the UHF images fall
outside of the TV band. The first image falls at approximately
-22-
.~, .
..
3~
1130 MHZ and the last image falls at approximately 1545 MHZ. Thesefrequencies are allocated for aircraft navigation, with the TACAN
system having the highest power output. But TACAN power output is
only 5 KW with a 1.~ x 10 to 1 duty cycle. And thus the TACAN
signal is 50 DB lower than that transmitted by a 1 MW TV trans-
mitter. Therefore, the filters of FIGURES 5b and 5c provide
adequate rejection of image signals for the UHF band.
FIGURE 6 is a detailed circuit diagram of switch 15.
one portion is basically comprise~ of a diode 101 connecting
ieads 14 and 18, and an RLC bi~s network 102 having a control
input 103. A DC voltage control signal SELLOVHF is applied to
lead 103 for selectively turning diode 101 on or off to thereby
select or deselect the low VHF signals on lead 14. Another
portion identical to the one descxibed above couples lead 18 to
lead 16, and is utilized to select and deselect the high VHF
channels.
Referring next to FIGURES 7 and 8, there is illustrated
a detailed circui~ diagram of RF ampli~ier 17 and mixer 20,
respectively. Basically, amplifier 17 is comprised of a dual
gate MESFET transistor 1~1 having a source coupled to a bias resistor
and coupling capacitor 112, and a drain coupled to an L-C bandpass
circuit 113. The gain of transistor 111 is varied about the levels
indicated in FIGURE 4a by an automatic gain control signal AGCRF.
Signal AG~RF couples to a gate of transistor 111 through a voltage
dividing network 114.
SLmilarly, mixer 20 is comprised of a single gate
MESFET transistor 121 having a source coupled to a bias resistor and
coupling capacitor 122, and a drain coupled to an L-C bandpass circuit
123. Circuit 123 has a center frequency of 330 MHZ. The mixing
signals from VHF VCO 22 are coupled to the gate of transistor 121
via an ~C circuit 124.
;hl~,3~
It should b~ emphasized that the Schottky har~ier
gate structure of the MESFET transistor of amplifier 17 and
mixer 20 yield significant performance improvements over
other known devices. Narrow gate depletion mode MOSFET devices
have a high frequency response, but they approximate square-law
transfer characteristics over only a very narrow range of gate
bias. Since departure from square-law opera*ion results in
cross modulation and intermodulation distortion, the devices
are restricted to a small dynamic range. In comparison, the
MESFET devices 111 and 121 have nearly ideal square-law
characteristics over the entire operating range indicated in
EIGURE 4a. JFET devices also have a good square-law transfer
characteristic, but their high frequency performance is
greatly reduced because of parasitic capacitance and process
difficultites that limit their usable geometries. ~
Referring now to FIGURE 9a, there is illustrated a `
complete MESFET device suitable for use as mixer 20. The
MESFET of FIGURE 9a has a closed gate 131. Gate 131 is
approximately 80 mils in length. The gate metal is approximately
0.3 mils wide, while the Schottky barrier gate in contact with
the gate metal i5 approximately 0.15 mils wide. The source
consists of five fingers 132-136 having lengths of 32.7 mils,
- 4.1 mils, 4.1 mils, 7.7 mils, and 12.1 mils respectively.
The width of these fingers is approximately 0.3 mils. The
drain consists of fingers 137-140.
FIGURE 9b is an enlarged cross-sectional view
taken alon~ the line A-A of FIGURE 9a. The MESFET is
constructed on a silicon substrate 140~having a P-type
impurity. The resistivity of substrate 140 is approximately
50 ohm-centimeters. Each of the source electrodes 132-L36
,;
,
3~
couples to an N~ doped region 142~146 respectively. These
doped regions have a resistivity of approximately 0.005
ohm-centimeters. The doped regions extend beyond their
cor=esponding electrodes by
~,' '', .
~,
: ,. .
, '` ,:
:`
',
-24a~ -
35i~
approximately 0.3 mils and are separated from the spaced apart
; gate electrode by approximately 0.15 mils. Similarly, the metal
electrodes 137 140 which form the drain electrodes are coupled -~
to underlying N+ doped regions 147-150, respectively. The ~e~-
~e 5'~ v~
and geometry of the drain doped regions ~ similar to
that of the source doped regions.
FIGURE 9c is a greatly enlarged cross sectional view
of one source drain pair within a dual gate MESFET device suitable
for use with amplifier 17. The entire device is constructed
similar ~o that of FIGURE 9a, with the modification that two gates
are interleaved between the source and drain electrodes. In
FIGU~E 9c, source electrode 132a and drain electrode 137a
correspond to electrodes 132 and 137 of FIGURE 9a. Gate
electrodes 131a and 131b occupy the space of electrode 131 in
FIGURE 9a.
FIGU~E 9d illustrates the I-V characteristics of the
devices of FIGURE 9a, while FIGURE 9e illustrates their trans-
conductance as a function of gate voltage. In an ideal square-
law device, the drain current is proportional to the square of
the gate voltage. And since transcon~uctance equals the partial
derivative of drain current with respect to gate voltage, the
transconductance is directly proportional to gate voltage for
ideal square-law operation. FIGURE 9e demonstrates such a
Iinear relation between the transconductance and gate voltage
for the MESFET device of FIGURE 9a. !''
Referring now to FIGURE 10, a circuit diagram of switch
25 is illustrated. Switch 25 is constructed identical to the
previously described switch 15 of FIGURE 6. The signals on leads
24 and 26 are selectively coupled to the output lead 27 via DC control
S E L L~
signals select ~HF (SELVHF) and select UHF (~VHF) respectively.
FIGURES lla-llh illustrate the details of SWD filter 28.
Filter 28 is comprised of a surface wave device chip 160 having
an L-C input circuit 170, and an L-C output circui~ 190 for
f~3~
impedance matching and phase shifting siynals to and from SWD
de~ice 160. ~eads'172-174 and 192-194 couple circuits 170 and
190 to SWD 160 as illustrated in FIGURE lla.
D,evice 160 has a magnitude-frequency characteristic as
illustrated in FIGURE llb. In particular, device 160 attenuates
signals which are 1.5 MHZ above the picture carrier of the desired
channel by at least 65 db. At that frequency, the sound carrier
of the channel adjacent to the desired channel is present. It is
important to greatly attenuate this sound carrier because it is
translated into the video signals of the selected channel by
synchronous detector 42. That is, the sound carrier at 46.5 MHZ
is'translated to 1.5 MHZ by detector 42. In the receiver of
FIGURE 1, ju~t perceptable picture lnterference occurs when the
sound carrier at 46.5 MHæ is passed- through SWD filter 28 with
a magnitude that is within 36 dB of the picture carrier at 45 MHZ.
Thus, SWD filter 28 with its greater than 65 dB adjacent channel
rejection enables the receiver of FIGURE 1 to have ~ood picture
reception even though signals at the filter input in an adjacent
channel are much larger than signals in the desired channel. For
example, if the signal level of the desired channel is -55 DBM at
the input o~ filter 2~, then the 46.5 MHZ sound carrier at the
input oE filter 28 can be as high as -26 dBm and ~he receiver of
FIGURE 1 will meet the requirement of a -36dB difference between
the in band and out band signals at the output of filter 28. By
comparison, prior art television receivers typically have per-
cepkable picture interference when the adjacent channel sound
carrier is -40 dBm.
FIGURES llc and lld illustrate the physical structure
of one embodiment of SWD 160. Basically, device 160 is comprised
of a piezoelectric substrate 161 which in a preferred embodiment
is made of quartz. Quartæ has a desirable feature in that the
velocity of surface waves through quartz is practically indepen-
dent of temperature. The dependence of velocity on temperature
is of considerable importance since velocity effects the center
frequency of the filter as described below.
Three electrically independent conductors 162~164 are
,.ti3r5~
disposed on substrate 161. Each conductor has corresponding
finger electrodes 162a-164a which axe disposed on substrate 161
in a comb like fashion. Fingers 162a-164a are equally spaced.
The distance between two consecutive fingers on the same con-
ductor is one wavelength of the center frequency of the filter.
The velocity of a surface wave in quartz is approximately 3,300
meters per second, and the center frequency of device 160 is
approximately 330 MHZ. Thus, the distance between two consecu- -
tive 162a fingers for example is approximately 10 x 10 6 meters.
Conductors 162-164 are coupled to the input circuitry
170 of FIGURE lla via leads 172-174 respectively. Input cir-
cuitry 170 generates voltages on electrodes 162-164 which are
120 out of phase with each other. This phase relationship
generates a unidirectional surface wave on substrate 161. That
is, waves in the forward direction add constructively, while
waves in the reverse direction add destructively. As a-result,
device 160 has a low insertion loss for signals in the passband.
In particular, the loss through filter 160 is no more than 3.5
Pa~
~ DB. United States ~a~e*~ 3,686,518 issued August ~l 1972 to
~artmann et al and assigned to Texas Instruments Incorporated
includes additional structuraI details of unidirectional surface
wave filter 160.
FIGURE lld illustrates a method for shaping the impulse
response of surface wave filter 160. The method therein illus-
trated is known as the ~inger withdrawal method. It involved
removal of groups of fingers from selected portions of substraté
161. Basically, the amplitude o the impulse response in thos~
regions from which the fingers are removed is reduced below the
value which it would ha~e if the fingers had not been removed.
The method thus provides the capability to control the relative
~2-7--
b3~`~
amplitude of the impulse response along the length of substrate
~61. The desired impulse response is obtained by taking the
inverse Fourier transfer of the requency response of FIGURE llb;
and then fingers 162a-16:4a are selectively removed in accordance
with the desired impulse response. Further details of the finger
withdrawal method are contained in U. S. Patent No. 3,946,342
issued March 23, 1976 to Hartmann and assigned to Texas Instruments
; Incorporated.
FIGURES lle-llh illustrate a method for determining
the value of inductors 175-177 and capacitor 178 comprising input
circuitry 170. As was previously described, one of the functions
of circuit 170 is to generate voltages on leads 172-174 which are
~: shifted in phase from each other by 120. FIGURE lle illustrates
that this 120 phase shift can be achieved by a circuit 180 which
produces a 60 phase lag between the voltace on lead 172 and the
~oltage on lead 173, in combination with a grounding of lead 174.
B This point is:~urther illustrated by the phaseS diagram of
FIGURE llf.
; A circuit for producing a 60 phase lag is illustrated
in FIGURE llg. The circuit consists of a pi shaped R-L-C network
consisting of a capacitor 181,and inductor 182, a capacitor 183,
and a resis~or 184. Included in FIGURE llg are two equations
relating the angle of phase lag between leads 172 and 173 in
terms of components 181-184. In the case at hand, the angle of
phase is 60~, and resistor 184 is the resistance between the
electrodes 162 and 163 divided by 2. Thus, utilizing the equa-
tions of FIGURE llg, values ~or ¦Xl¦ and ¦Xcl can be calculated.
Impedance 183 is physically implemented by coupling
. inductor 175 across leads 173 and 174; and impedance 182 is
physically implemented ky coupling inductor 176 across leads 172
-2~-
3~
and 173 as illustrated in FIGURE llh~ The parallel combination
of inductor 175 and the capacitance between leads 173 and
174 due to the SWD electrodes is chosen to equal impadance 183.
Similarly, the parallel combination of inductor 176 and the
capacitance across leads 172 and 173 due to the SWD electrodes
is chosen to equal impedance 182. Inductor 175 may typically
equal 30 nanohenries, while inductor ~ may typically equal
35 nanohenries as an example.
Inductor 177 is then added between leads 172 and 174
while capacitor 178 is added between leads 172 and 29 so as to
match the impedance between leads 172 and 174. Typically,
l7
inductor 177 is approximately 25 nanohenries, and capacitor
is 10 picofarads.
To this point, the discussion with reference to FIGURES
lla-llh has concentrated primarily on the structure of input
circuitry 170 and the input transducer of SWD device 160. It will
be understood ~owever, that surface wave device 160 also has an
output transducer on substrate 161 of a construction similar to
that of the input transducer~ Also, output circuitry 190 of
F~GURE lla is constructed similar to lnput circuitry 170.
Referring next to FIGURES 12a and 12b, a circuit dia-
gram of RF amplifier 29 is therein illustrated. Amplifier 29
is comprised of stages 200 and 201. Stage 200 includes a single
bipolar transistor 202 as i~s amplifying element, whereas
stage 201 includes several DC coupled bipolar transistors in an
integrated circuit 203 as its ampiifying element. Circuit 203 is
indicated in FIGURE 12a as a single circuit element, and is shown in
detail in FIGURE 12b.
Amplifier 29 provides a maximum gain of 30 DB to
signals on its input lead 30. A small portion of the gain is
-29-
3~ ~
provided by stage Z00 which has a relatively good noise figure,
whereas the remainder of the gain is provided by stage 201. AGC
control signal AGCIF couples to stages 200 and 201 as indicated
in FIGURE 12a through bias circuits 216 and 217. The magnitude of
signal AGCIF varies so as to keep the output of amplifier 29 at
approximately -26~5 DBM. The variation of gain versus input
signal strength for amplifier 29 was previously indicated in
FIGURE 4a.
The output of amplifier 29 couples to mixer 32 which
is illustrated in the circuit diagram of FIGURE 13. The basic
mixLng circuit element utilized therein is a bipolar transistor
220. Transistor 220 has an emitter 221 which is coupled to the
output of amplifier 29. Similarly, transistor 220 has a base 223
which is coupled.to SWD oscillator 35, Oscillator 35 generates
mixing signals of 285 MHZ on the base 223 of transistor 220. As
a result, sum and difference fxequencies are generated on the
co~lector 225 of transistor 220. Collector 225 is coupled to an
LC tank circuit 226 having a resonant frequency of about 45 MHZ.
The ou~put of tank circuit 226 is coupled via a tapped transformer
227 to lead 37 and slgnals S4(f) are generated thereon.
Referring next to FIGURES 14a and 14b, a circuit diagram
of oscillator 35 and a schematic diagram of acoustic surface wave
resonator 36 are illustrated. Oscillator 35 includes a bipolar
transistor 231 as the amplifying element. Transistor 231 has a,
grounded base 232, an emitter coupled via lead 233 to the input of
resonator. 36, and a collector coupled through a capacitor 234 to
the output of resonator 36. An LC circuit 236 couples a DC supply
. r ~
. E ; voltage V~ to the collector of transistor 231 and also provides
an output signal on lead 34.
.,
....
-30-
.
Surface wave resonator 36 has a relatively high
resonant frequency of 285 MHZ. Thus it is relatively small in
size. Device size decreases as the resonant frequency increases.
Typically, the SWD 36 of FIGURE 14~ is only approximately 0.10
inches in length. SWD 36 also has good long term frequency
stability. This is because the resonator has a large Q. Typically
its Q is greater than 15,Q00. Q is the ratio of energy stored to
energy dissipated per cycle within the device.
SWD ~esonator 36 is comprised of a piezoelectric sub-
strate 240 which in the preferred embodiment is made of quartz.
Reflective grating structures 241 and 242 are disposed at
opposite ends of substrate 240. These grating structures form
e s ~ 6 s ~L~ 7~e
discontinuities in the surface of~240 which reflect surface waves
thereon. Gratings 241 and 242 may be comprised of grooves t or
alternatively bars of gold or copp~r for an example. The bars
are~spaced apart by one half wavelength of the resonant frequency.
Typically, 250 to 400 bars are contained within each of the
grating structures 241 and 242; The Q of the resonator increases
a~ the number of bars increases. Also, as previously pointed
out, the velocity o~ the surface waves in quartz is relatively
insensitive to temperature change. Thus, the resonant frequency
of xesonator 36 has low temperature drift. Typical~y, the
resonant fre~uency varies less than 20 KHZ over~temperature range
Cl f. ,o ~ QC
.
An input transducer 243 and an output transducer ,244
are disposed on substrate 240 in the space between grating
structures 241 and 242. Lead 233 couples to input transducer 243,
and lead 235 couples to output transducer 244. Transducers 243
and 244 are comprised of a number of interleaved fingers which
are placediat the peaks of the resonating standing wave that is
set up by reflecti~e grating structures 241 and 242
-31-
"ii3l5~
Typically, 60 fingers are on each txansducer. Further details
of resonator 36 are included in U.S. Patent 3,886,504 issued
to Hartmann et al on May 27, 1975 and assigned to Texas
Instruments Incorporated.
FI~URE 15 is a circuit diagram of amplifier 38. Ampli~
fier 38 is basically comprised of a circuit 250 which was
previously illustrated in detail in FIGURE 12b. Nodes A-G
of circuit 250 as illustrated in FIGU~E 15 correspond to the
node A-5 as illustrated in FIGURE 12b. Signals from mixer
32 couple to the input of circuit 250 through a capacitor
; 251. The gain of circuit 250 is automatically ad~usted by a
gain control signal AGCIF. Signal A5CIF is coupled to node
C through an RLC circuit 252. The output of circuit 250 is
coupled to lead 40 through an LC tank circuit 253.
FI5URES 16-19 are detailed circuit diagrams of the
remaining portions of the channel selector of FIGURE 1. The
circuits utilize conventional components and are generally ;~
self-explanatory to those with ordinary skill in the art.
Synchronous detector 42 is illustrated in FIGURE 16. Detector
42 is essentially comprised of a commercially available chip
MC1596. Chip MC1596 is~described in the Linear Integrated
Circuits Catalogue on pages 8-404 to 8-414 of Motorola's
Semiconductor Data Library, Volume 6, Series, 1975. Details ;
o.f the circuit are given on page 8-411 in FIGURE 23 of the :
cited reference. Chip MC1596 is appropriately biased at each
of its inputs and ouputs by ~LC circuits 261-269 as illustrated
in FIGUP~E 16. The biasing required by the component is also
described in the cited reference.
Oscillator 45 is illustrated in FIGURE 17. Oscillator
45 has a MESFET transistor 270 as the amplifying element, and
frequency determining L-C feedback networks 271 and 272. Two
-32-
. 315~
separate output signals are provided by oscillator 45. One of
the output signals is generated on lead 47 and the other signal
i5 generated on lead 44. Leads 44 and 47 are separated by an RLC
phase shifting network 273 which generates approximately a 90
phase difference between the two signals. Phase shifting network
273 insures that the picture carrier on lead 40 is in phase with
the oscillator signal on lead 44. The phase detecting circuit 43,
as illustrated in FIGURE lB, generates phase detection signals
PDl and PD2 which lock the oscillator signal on lead 47 90 out
of phase with the picture carrier on lead 40. Circuit 273 rein-
serts this 90 phase shift.
The phase detector 43 of FIGURE la also utilizes chip
MC1596 as was utilized ln sync detector 42. Bias networks 281-287
are applied to chip MC1596 to achieve the phase detecting function.
Ramp generator 5~ and loop filter 49 couple to outputs of chip
MC1596 as illustrated in FIGURE 18. The combination generates
pHase detecting signals PDl and PD2 on leads 51.
FIGURE 19 illustrates a circuit diagram of VHF-VCO 22.
V~F-VCO 22 utilizes phase signals PDl and PD2 in combination with
a third signal PD3 to genexate the selectable frequency L0
signals on lead 23.. Signal PD3 is a multilevel analog signal
which provides a coarse voltage level of a unique value far each
channel to be selected. Signal PD3 is generated by controller
52 in response to manually operated channel selection switches 56
as is previously described in conjunction with FIGURE 1. Signals
PDl-PD3 are coupled across a varactor diode 290 through summer
50. Diode 290 in combination with. capacitors 291 and a micro
strip 292 form the ~re~uency determining circuit of VHF VCO 22.
Signal amplification within VCO 22 is provided by a bipolar
transi.stor 293.. Transistor 293 is selectively enabled or
-33-
d~ 3~i~
disabled ~y a control signal "enable VHF" (ENVHF). Signal ENVHF
couples to the base of transistor 293 through a resistor 294 and
to the collector of transistor 293 through an inductor 295. It
is selectively ~ in response to the position of the
channel selection switches 56.
A complete television receiver which incorporates the
channel selector of FIGURE 1 is illustrated in block diagram form
in FIGURE 20. The channel sel~ctor has input leads 11 and 61 for
receiving VHF and UHF television signals as previously described.
The channel selector also has input leads 53 for receiving coarse
analog voltages indicating the selected channel. The RF
section of the channel selector frequency shifts the selected
channel to approximately 330 MHZ, while the IF section of the
channel selector filters the channel at 330 MHZ and frequency
shifts the filtered channel to baseband.
Lead 46 is the audio output of the channel selector.
As previously described, the sound carrier of the selected
channel i's generated on lead 46- at 4.5 MHZ. Lead 45 is coupled
to the input of an audio demodulator 301. Demodulator 301
generates sLgnals on a lead 302 with an amplitude that is
proportional to the requency of the fre~uency modulated signals
on lead 46. This frequency demodulation process may be
implemented by a variety of circuits that are well known in the
art. The demodulated signalc on lead 302 couple to the input
of a speaker 303 where they are electromechanically conver~ëd to
audible sounds.
Lead 72 is the composlte video output of the channel
selector. That is, s'ignals on lead 72 include frame synchronizing
information and vid,eo information of the selected channel. Lead
,
-34-
~ ~ ~ .
~' 3~ ~ .
72 couples to the input of a video processing unit 304, Video
processor 304 separates the picture signals from the frame
synchronizing signals. The picture signals are generated on a
lead 305 which couples to the electron gun input of a picture
tube 306. The frame synchronizing signals are generated on a
lead 307 which couples to the input of a drive circuit 308.
Drive circuit 308 generates horizontal and vertical synchronizing
signals on a lead 309 which couples to electron beam deflection
circuitry 310 of picture tube 306. Drive circuit 308 also
generates horizontal synchronizing signals on a lead 311 which
couples to an input o picture tube high voltage generator 312.
Additionally, drive circuit 308 generates synchronizing signals
on lead 73 which couples to th~ AGC circuitry 74 of the channel
selector.
Tele~ision receiver components 301-312 have been
described in detail in many prior art publications. See for
example the Fundamentals o~ Display System Design by Sol Sherr,
1970, published by Wiley-Interscience. A bibliography on pages
445-46g of the cited reference also includes many additional
re~erences.
FIGURE 21 is a block diagram of a second embodiment
o f a channel selector constructed according to the inyention.~ A
significant portion of this second embodiment is similar in
co~struction to that of the FIGURE 1 embodiment. The similar
portions are indicated by the identifying reference numeral's.
One structural difference of the second embodiment is
that it has only one VHF filter. That is, signals on VH~ antenna
10 are coupled to a local oscillator mixer 320 through a fixed
filter 321 which passes the entire VHF band of frequencies. A
second difference is that the embodiment of FIGURE 21 contains
no RF amplifier. As a result, the system has improved inter-
a~
modulation distortion and cross modulation distortion but has~
,
3~6
increased noise figure. Still another difference is that the
FIGURE 21 embodiment contains only one RF MESFET mixer. A two
by one switch 322 is provided having one input coupled to
receive VHF signals on a lead 323 and a second input coupled to
receive UHF signals on a lead 324. The output of switch 322
couples via a lead 325 to MESFET mixer 320.
The embodiment of FIGURE 21 also includes a different
means for generating the coarse channel selection voltages on
leads 53. As FIGURE 21 illustrates, controller 52 is comprised
of a phase locked loop. The loop receives reference signals of
a fixed frequency from a circuit 331, and simultaneously
receives feedback signals from the LO ~CO. The LO VCO signals
are sent through.a variable counter 332. Counter 332 divides
by anumber which is selectable via logic signals on leads 333.
The signals on .leads 333 are generated by a logic circuit 33~
in response to logic signals received from the channel selection
switches 56. A~phase detector :335 compares the output signals
of variable counter 332 to the reference signals, and generates
phase detection signals for the ~O VCO~ Thus, a relatively
high local oscillator frequency is generated when counter 332 is
5elected to divide by a relatively large n~mber and vice versa.
Various embodiments of the invention have now been
described in detail. However, many changes and modifications
can be made to the above details without departing from the
nature and spirit of the invention. For example, the IF
frequency of the mixer output is not restricted
to 330 MHZ. Other IF frequencies in the range of 300-400 MHZ
may be employed. A~ another example, a bipolar RF amplifier may
-36-
be substituted for the MES~T ,R~ amplifier. This is because
the mixer introduces cross modulation distortion and inter- ,
modulation distortion into the rece~ver to a much smaller
degree than does the linear RF amplifier. Thus, utilization
of a MESFET mixer yields a receiver having greatly reduced
third order distortion even though the RF amplifier is bipolar.
As another example, the surface wave device resonator utilized
to generate 285 MH~ mixiny signals may be comprised of a
single transducer as opposed to a dual transducer. In a
single transducer resonator, lead 233 couples to one set of
electrodes on the transducer while lead 235 couples to the
other set of electrodes. The single transducer is configured
similar to transducer 243. As still another example, the
channel selector of FI~URE 1 or FIGURE 21 may be readily
adapted for use in systems other than television receivers.
The channel selector has application wherever one channel of
frequencies is desired to be selected from a plurality of
nonoverlapping frequency channels. The information contained
in the channels need not be television signals. Therefore, it
will be understood that many changes and modifications can be
made in the above details without departing from the nature
and spirit of the invention. It ~s understood that the
invention is not limited to said detail except as set forth
in the appended claims.
-37-
, ~
,
,,
,
: ,
,: ,
