Note: Descriptions are shown in the official language in which they were submitted.
~2`7~
The Government has rights in this invention pursuant to
Contract ~o. N00024-82-C-5110 awarded by the Department of
the Navy.
Background of the Invention
This invention relates generally to high voltage soli~
state switches and, more particularly, to switches used as
pulse modulators for cathode pulsed tubes used in amplifying
radio frequency signals.
As is known in the art, it is sometimes desirable, as
in radar transmitters, to produce amplified pulses of radio
frequency energy. One such technique includes feeding a
pulse modulating signal to a modulator circuit which elec-
trically couples, or decouples, a power supply to or from
the cathode-anode of a crossed field tube selectively in
accordance with the modulating signal. Such crossed field
tube may be a magnetron, a klystron or a crossed field ampli-
fier (CFA) tube. Typically, the modulator circuit includes
a high power switch tube with the plate electrode thereof
serially connected to the cathode of the radio frequency
tube, the anode of the radio frequency tube being grounded,
and, the cathode of the switch tube being serially coupled
to the negative terminal of a positive grounded high voltage
power supply. Thus, radio frequency energy fed to the input
port of the radio frequency tube is amplified in the radio
.
lZ745~
frequency tube and is coupled to the output port thereof
when the high voltage power supply is electrically coupled
to such radio frequency tube by the modulator; conversely,
the input radio frequency signal is decoupled from the
output port of the radio frequency tube when the modulator
electrically decouples the high voltage power supply from
the radio frequency tube. In this manner, pulsing the
modulator results in pulsed amplified radio frequency (RF)
energy at the output terminal of the radio frequency tube;
such pulsed RF energy having the same pulse width, duty
cycle, and pulse repetition frequency as the modulating
signal fed to the pulse modulator.
While such pulse modulator has been found useful in
some applications, the switching tube used in such circuit
usually has a short operating lifetime when compared to the
radio frequency tube, and thus, such switch tube is a
significant contributor to transmitter maintenance, material
and workload. Further, the heater power required with such
switch tube consumes significant prime power and contributes
to overall transmitter inefficiency since, inter alia, they
require high voltage drops because of high plate resistance,
require a number of high voltage supplies for biasing. Still
further, the ~witch tubes are very susceptible to damage in
a high shock and vibration environment. Thus, overall, the
switch tubes have demonstrated a relatively low mean time
between failures (MTBF).
- 2 -
,.:
,., ~ ,. .
.
- ' : : ', '
-
..
'
.
127'~5~
One technique suggested to eliminate the use of the
switch tube is to use a solid state device, such as a
transistor, in its place. The use of a single transistor~
however, is not practical for high voltage applications
where the transistor will have developed across it the high
voltage of the supply when the transistor is in the non-
conducting state. One technique suggested to remove this
excessive voltage condition across the transistor is to
provide a plurality of serially coupled transistors between
the high voltage power supply and the load. With such
arrangement, however, the drive signals for each of the
transistors must be generally biased to a different
relatively high voltage potential. To provide such drive
signals typically requires the use of a tapped transformer
or series of resistors to provide the properly biased
control signal for each of the serially coupled transistors
thereby reducing the desirability of such an arrangement
because of resonances, time delays, and power loss with the
tapped transformer or series of resistors.
,
s~
2901-678
Summary of the Invention
In accordance with the present invention, there is
provided a switching circuit for electrically coupling, or de-
coupling, a voltage supply to, or from, a load selectively in
accordance with an electrical control signal, such circuit com-
prising: (a) means for converting the electrical control signal
into a radiant energy signal; and, (b) a plurality of switching
modules serially coupled between the load and the voltage supply,
each one of the modules being at a different reference potential,
each one of such modules including: a switching transistor; means,
responsive to the radiant energy signal, for providing a drive
voltage to drive such transistor into either a conducting or a
non-conducting condition selectively in accordance with the
radiant energy; energy storing means, such energy storing means
being coupled to the voltage supply for storing energy therein
when such voltage source is decoupled from the load, such storing
means applying power to such drive voltage providing means when
the voltage supply is coupled to the load; and means, coupled
to such transistor, for regulating such drive voltage to a sub-
stantially constant predetermined voltage with respect to thereference potential of such module to maintain a substantially
. --4--
. . . ~ . .
- .
.- ~ ' .
- . :.
,
8i
2901-678
constant current from such voltage source through each of such
transistors to such load during the conducting condition of such
transistor.
In accordance with a preferred embodiment of the inven-
tion, each one of the switchin~ modules includes means, coupled
across the transistor, for providing a short circuit in parallel
with the transistor when the switchin~ circuit couples the vol-
tage source to the load in the event of a failure in such module
with the voltage normally distributed to such failed module
being distributed among the remaining non-failed modules.
Because of the number of modules used, a failure of one module in
N modules increases the voltage previously distributed to that
module among the remaining modules, thus insuring proper operation
of the transistors in the remaining modules.
~, -5-
,.~
,, ' . ~ ' '' .
. , - ' ' . .
,. ~ , .
~27~5~1
srief Description of the Drawing
For a more complete understanding of the concepts of
this invention, reference is now made of the following
description taken together in conjuction with the accom-
panying drawing in which the single FIGURE is a schematic
diagram of a radar system including a pulse modulator
according to the invention.
:.
, s
':
, ~
~ 25
,!' ~
~ - 6 -
. ~,
, ..... . . .
.
~, . . : - . . .
,~' :' , - ' : ,
.: '. . . . ' -:
' . : ', ' ,, ' ' . :
.
" ' :
. - : ,
.
:, - , ~ .
: ' ~ ~ . -- ' ' ' '
~2745~
Description of the Preferred Embodiment
Referring now to the single FIGURE, a coherent pulsed
Doppler radar system 10 has been selected to illustrate how
the invention might be applied. Thus, the illustrated radar
system 10 includes a radar antenna 12, a duplexer 14, a
radar receiver 16, a radar transmitter 18, a radio frequency
(R.F. ) oscillator 20, a synchronizer 24, and a system trigger
28, all arranged in a conventional manner, as shown, whereby:
(a) during transmit modes, synchronizer 24 sends signals to
system trigger 28 and in response thereto, radio frequency
energy produced by oscillator 20 and coupled to transmitter
18 via conventionl directional coupler 22, is amplified and
pulse modulated by such transmitter 18, such amplified and
pulse modulated radio frequency energy then being coupled to
antenna 12 via duplexer 14 for transmission; and, (b) during
interleaved receive modes, portions of the transmitted energy
re1ected by object~ within the beam of the antenna 12, are
received by such antenna 12 and are passed via duplexer 14,
to the radar receiver 16 where they are heterodyned with
signals produced with the signals produced by oscillator 20
into video signals, such video signals then being resolved
into range bins in response to signals fed to the receiver 20
from synchronizer 24 via bus 26. It is noted that while the
antenna 12, duplexer 14, receiver 16, oscillator 20, synchro-
nizer 24 and system trigger 28 are all of conventional design,
:,, , ~,, .
, - , -
: " . .-, . . - . -
''~.' '' '' :, . -
'
.
: :
1%74~81
the transmitter 18 includes a conventional cathode pulsed
radio frequency amplifier tube, here a conventional crossed
field amplifier (CFA) 30, controlled by a pulse modulator 32
according to the invention.
S AS shown, the crossed field amplifier 30 includes an
anode 34, coupled to ground as shown, a cathode 36 coupled
to the modulator 32, an input port 38 coupled to the oscil-
lator 20 via directional coupler 22, and an output port 40
- coupled to the duple~er 14, as shown. A tailbiter resistor
33 i8 coupled between the anode 34 and cathode 36 in a con-
ventional manner, as shown. The pulse modulator 32 includes
a plurality of, here N, identically constructed switch
modules 421-42N (an exemplary one thereof, here switch module
42N_l being shown in detail) serially coupled between the
cathode 36 of the crossed field amplifier 30 and a high
voltage supply 44. Here, the voltage supply 44 is of any
conventional design and produces a voltage of magnitude V,
the negative potential being at negative terminal 46 and the
positive potential being coupled to ground, as shown. Also
;~ included in the pulse modulator 32 is a plurality of, N,
light emitting diodes 481-48N, the output of each one thereof
providing an input for a corresponding one of the switch
modules 421-42N, respectively, as shown. The input signal
to the light emitting diodes 481-48N is supplied as a common
signal from system trigger 28 via line 50, as shown.
': ~
~ - 8 -
"; :;,~
.. ~ , - "
, '':' ;' ' '' ' , . ' '' '' ' ~ . ' ' - ' '
.
' ~
:.
.
:, . ~ ' :
7'~5~
In operation, when an amplified pulse of radio frequency
energy is to be transmitted, system trigger 28 pulses "on"
the light emitting diodes 481-48N. The pulses of light pro-
duced by such diodes 481-48N are sensed by the switch modules
421-42N, and in response to such sensed light, the switch
modules electrically couple the negative terminal 46 of the
voltage source 44 to the cathode 36 of the crossed field
amplifier 30 to thereby power such amplifier 30 and enable
it to amplify the radio frequency energy fed thereto from
oscillator 28. Conversely, when the switch modules 421-42N
do not ~ense light from the diodes 481-48N, the modules
421-42N electrically decouple the voltage supply 44 from the
cathode 36 of thé CFA 30 and radio frequency energy from
oscillator 20 is electrically decoupled from the output port
40 of CFA 30. Thus, each time a pulse of RF energy is to be
transmitted, a corresponding pulse of light is produced
simultaneously by each of the diodes 481-48N and, in response
thereto, each one of the modules 421-42N operates electrically
to replicate the light emitted pulse and such modules 421-42N
thus operate simultaneously to pulse modulate the operation
of CFA 30.
Considering now the details of an exemplary one of the
switch modules 421-42N, here switch module 42N_l, it is first
noted that such module 42N_l has a pair of output terminals
52N_1, 54N-1 and that the output terminal 52N_l is connected
., :
1274$81
to the output terminal 54N-2 of module 42N_2 (i.e. the modules
directly serially connected thereto) and the output terminal
54N-1 of module 42N_1 is connected to the output terminal 52N
of module 42N (i.e. the other one of the modules directly
connected thereto). Next, it is noted that output terminal
521 of the first one (i.e. module 421) of the N serially
connected modules 421-42N is connected to the cathode 36 of
CFA 30 and that the output terminal 54N f the last one (i.e.
module 42N) of the N serially coupled modules 421-42N is
connected to the negative terminal 46 of the voltage supply
44, as shown. As will become evident hereinafter, when the
modules 421-42N detect light emitted by the diodes 481-48N,
the output terminals 521, 541 to 52NI 54N of such modules
421 to 42N become electrically coupled together (coupled
through a relatively low impedance) whereas in the absence
of such detected light, the output terminals 521, 541 to 52N,
54N of such modules 421 to 42N become electrically decoupled
(more accurately, coupled through a very high impedance which
is substantially an open-circuit).
Thus, considering module 42N_1, it is first noted that
module 42N_l includes a conventional fiber optic receiver,
hereinafter referred to as optoreceiver 56. ~ere, opto-
receiver 56 is model HF8R 2202 sold by ~ewlett Packard, Palo
Alto, California. Optoreceiver 56 has its input 58 aligned
to receive light from light emitting diode 48N_l and is
-- 10 --
1274~
powered by a voltage (here 10 volts) coupled across ter~inals
60, 62 in a manner to be described. Suffice it to say here,
however, that when a suitable voltage is applied across
terminals 60, 62, a negative going electrical pulse is pro-
duced by optoreceiver 56 on line 64 when such receiver 56
senses a pulse of light emitted by diode 48N_l. The electri-
cal signal on line 64 thus is referenced to the potential at
terminal 62 so that in the absence of a pulse of light, the
signal on line 64 is at a high positive relative to the
potential at terminal 62 and in the presence of a light pulse,
the signal on line 64 goes negative i.e., to a potential near
the potential at terminal 62. The signal on line 64 is
coupled in parallel to a pair of identical invertinq drive
amplifiers, here a~plifiers 66a, 66b. The inverters 66a, 66b
are powered by a voltage coupled across terminals 68a, 70a
for inverter 66a and across terminals 68b, 70b for inverter
66b. The inverters thus invert the negative going pulse
produced by the optoreceiver 56 in response to the pulse of
light by LED 48N_l into a positive going pulse. It is noted
that the signals produced out of inverters 66a, 66b are
referenced to the voltage at terminals 70a, 7nb and thus in
response to the negative going pulse on line 64 go from a
reference potential near the potential at terminals 70a, 7qb
to a potential more positive, that is near the potential at
terminals 68a, 68b. The positive going pulses that are
. ' : . . . : .- :
.~ ' , . ' : '
'' ' ' ' , ~ . ", ~ ' '
lZ74581
produced by each of the inverters 66a, 66b are coupled as
drive signals to the gate electrodes (G) of a pair of n-
channel, enhancement mode, metal oxide semiconductor (MOS)
field effect transistors (FET's) 72a, 72b, respectively, via
resistors 74a, 74b, respectively, as shown. Thus, the swing
in voltage of the control signal fed to the gate electrodes
G is ~ V independent of the voltage at terminal 54N-l . The
source electrodes (S) (and substrates) of the FET's 72a, 72b
are coupled to the output terminal 54N-l via resistors 76a,
76b, as shown, and the drain electrodes (D) of the FET's
72a, 72b are coupled to the output terminal 52N_1.
It is also noted that the gate, or control, electrodes
(G) of the F~T's 72a, 72b are also connected to the output
terminal 52N_1 via resistors 78a, 78b, and capacitors Ca, Cb,
as shown. A zener diode 80 has its anode electrode (A)
connected to output terminal 54N-l and its cathode electrode
(C) connected to output terminal 52N_l. Diode 82 has its
anode electrode connected to output terminal 52N_l and its
cathode electrode connected to the input of a conventional
DC to DC converter 84. The output voltage produced across
terminals 86, 88 of the DC to DC converter 84 is coupled
across terminals 60, 62 of the optoreceiver 56, across
terminals 66a, 70a of inverter 66a, and across terminals
68b, 70b of inverter 66b. Completing the module 42N_l is
a storage capacitor Cs coupled across terminals 86, 88 of
- 12 -
,
. ., ~ ,
,
' : '
1~7~8~
the DC to DC converter 84, as shown.
In operation, when a pulse of light is produced by each
of the LED's 481-48N, a corresponding negative going pulse
is produced by the optoreceiver 56 on line 64. Such negative
pulse is converted to a corresponding positive going pulse
by the pair of inverters 6~a, 66b. The positive going pulse
produced by the inverters 66a, 66b drives the FET's 72a, 72b
to a conducting state (i.e., a relatively low resistance is
produced between the source (S) and drain (D) electrodes).
It is noted that the drive voltages produced by inverters 66a,
66b are self-referenced to the potential at terminal 54N-l-
The drive voltage into the gate electrodes (G) of the FET's
72a, 72b (i.e., the voltage at the gate (G) relative to the
voltage at terminal 42N_l) is regulated (in a manner to be
described) here to within lOmV, in order to generate a
constant current, here 12 amps, through each one of the pair
of transistors 72a, 72b, which is desired to supply 24 amps
for proper operation of the CFA 30. Here, the CFA 30
requires 14,000 volts and 24 amps to provide amplification
of the radio frequency signal fed to input port 38 thereof.
Because each module must pass 24 amps, a pair of parallel
connected FET's 72a, 72b are used in each module, each FET
carrying only 12 amps. Here, voltage supply 44 is an 18,000
volt supply. Thus, 4,000 volts is dropped across the plural-
ity of modules 421-42N. Here, N is eighty so that 50 volts
- 13 -
' ' ' , ' ' '
" :'
'' , :: ~, ,,,:
lZ745~
is dropped across each one of the eighty modules 421-42N.
It is also noted that the terminals 541-54N f each of the
eighty modules 421-42N is at different potential Vs4n, given
by v54(n) = -18,000 + 50(N-n) where N = 80 and n is the
number (i.e., subscript) of the module. Thus, for module
42N_1 (i.e. here n = N-l ) the potential at terminals 54N-l is
Vs4(N_1) = -18,000 + 50(N-(N-1)) = -17,950 volts. AS is noted
above, however, the pulses produced by optoreceiver 56 and
inverters 66a, 66b are referenced to Vs4(N_l) and also the
source (S) of FET's 72a, 72b are referenced to the voltage
V54(N_l). Thus, considering exemplary module 42N_1, 50 volts
is present between terminals 52N_1, 54N-1 with terminal 52N_
being at a positive potential relative to the potential at
terminal 54N-l- When the positive going pulses produced by
the inverters 66a, 66b are removed, the FET's 72a, 72b are
placed in a non-conducting state (i.e. a high resistance is
produced between the source (S) and drain (D) electrodes of
the FET's 72a, 72b) to effectively electrically decouple the
voltage source 44 from the CFA 30. It is noted, however,
that because of the capacitance between the anode and cathode
electrodes 34, 36 of the CFA, such electrodes initially store
14,000 volts when the CFA is removed from supply 44. This
stored 14,000 volts is discharged through the tailbiter
resistor 33 (here 20K ohms) with the result that cathode 36
of the CF~ 30 is at, initially, a negative 14,000 volt
- 14 -
':, , .
:,., - . ' ~ ' '
.' ' ' - ' ' . "
, ~ . . . .
9L274~81
potential relative to ground and discharaes in a short time.
Thus, the full 18,000 volts potential of the supply 44
appears across the eighty modules 421-42N. It is first
noted that the effective resistance across a module (i.e.,
as across terminals 52N_1, 54n-1 of module 42N_l) when such
module is in the non-conducting state may be considered as
substantially equal to the input impedance of the DC/~C
converter 86, represented in the FIGURE by a phantom resistor
87 connected between the cathode of diode 82 and terminal
54N-1 for module 42N_l. Since the input impedance 87 of
DC/DC connector 86 is here approximately 50 K ohms when the
modules are non-conducting, the total resistance between
cathode 36 of CFA 30 and terminal 46 of supply 44 is here
approximately 4 meg ohms when the modules 421~42~ are non-
conducting. Thus, such total 4 meg ohms resistance is 200
times larger than the 20K ohms resistance of the tailbiter
resistor 33, so that substantially all of the 18,000 volts
of the supply 44 is, therefore, distributed, here equally
among the eighty modules 401-40N with the result that each
module has a 250 volt potential between terminals 52N~ 54N
thereof; the potential at terminal 52N being more positive
than the potential at terminal 54N. Thus, here again, each
of the modules is at a different reference potential. That
is, the voltage at terminal V54n may now be represented as
V54(n) = -18,000 + 250(N-n) so that the voltage of terminal
- 15 -
'
:
,
~279~581
54N-lr V54(N-l) is now -18,000 + 250 (N-(N-l)) = -17,750 volts.
However, each of the elements 56, 68a, 68b, 72a, 72b, 80, 84
in the module is referenced to the potential at terminal
54N-l- Thus, the 250 volt potential at terminal 52N_l thus
forward biases diode 82 and such 250 volt potential is thus
electrically coupled to the DC I o DC converter 84. The DC
to DC converter 84 converts the 250 volt potential fed there-
to to, here, 10 volts relative to the voltage at terminal
54N-l- Such 10 volts is coupled to terminals 60, 68a, 68b
to thereby power the optoreceiver 56 and the inverters 66a,
66b. It is also noted that storage capacitor Cs charges to
10 volts relative to terminal 54N-l; thus, when the FET's
72a, 72b of modules 421-42N are conducting, as when the
supply 44 is electrically coupled to the CFA 30, diode 82 is
back-bia~ed and the 10 volt voltage stored in capacitor Cs
is coupled to terminals 60, 68a, 68b to thereby provide the
energy to power the active circuits (i.e., the optoreceiver
56 and the inverters 66a, 66b). Resistors 76a, 76b provide
stability against transconductance variations in the tran-
siRtors 72a, 72b thereby allowing proper current sharing
between such transistors. Resistors 78a, 78b in conjunction
with series resistors 74a, 74b provide the feedback required
to make the the modules have a low dynamic impedance which
is required to maintain an even distribution of voltage
among the modules.
- 16 -
, . .
.~ . - - , . .
";' ' ' ~ .' ` ' ' :
' ' ' ' " ''
;
lZ74581
Considering next the effect of a failure in one of the
modules 42l-42N, as for example, a failure in the one of the
LED's 48l-48N, which drives the module, a failure in the
optoreceiver 56 of such one of the modules, or a failure in
the DC to DC converter 84, when the CFA 30 is driven "on"
~i.e., coupled to supply 44), the zener diode 80 of such
failed module breaks down and short circuits and conducts
the requisite current from supply 44 to the CFA 30 and there-
by prevents a failure of the entire pulse modulator 32. It
is noted, however, that the 250 volts to be dropped across
the failed module is now distributed to the remaining ones
of the modules, here the remaining 79 modules, with the
result that such remaining modules have dropped across them
250 volts plus (250/79) volts; the added (250/79) volts being
only a small fraction of the normal 250 volts for which the
transistors in the modules were nominally designed to operate.
The zener diode 80 also is used to limit the voltage across
the module to, here, 300 volts, in event of arcs in the CFA
30 causing voltage surges throughout the pulse modulator 32
due to unavoidable series inductance in the wire inter-
connecting the modules 421-42N and connecting the modules
to the CFA 30.
As noted above, it is desired to regulate the voltage
at the gate electrodes of the FET's 72a, 72b. This is
accomplished through capacitors Ca, Cb. It is first noted
- 17 -
, ~ . ' ' :
-
, . .
.
, .
1274~1
that when the modules are in a non-conducting mode, there is
250 volts across the terminals 521r 541 to 52Nr 54N. Thus,
there is 250 volts across 52N_1, 54N-l of exemplary module
54N-l The capacitors Car Cb serve as d.c. blocking
capacitors during this condition and thereby prevent power
loss in resistors 74a, 74b, respectively. It is noted,
however, that capacitors Ca, Cb thus charge during the non-
conducting condition. As noted above, during the conducting
mode, storage capacitor Cs provides power to optoreceiver 56
and inverters 66a, 66b. As the energy is depleted from the
storage capacitor Cs, however, the voltage of the gates (G)
of FET's 72a, 72b would absent capacitors Ca, Cb, tend to
"drop". Capacitors Ca, Cb tend to reduce the "drop" by
discharging through the drain-source electrodes of FET's
72a, 72b through resistors 76a, 76b through storage capacitor
Cs, through the terminals 68a, 68b of inverters 66a, 66b, and
through resistors 74a, 74b. It is noted that the discharge
current from capacitors Ca, Cb in passing through resistors
74a, 74b produce a voltage across such resistors 74a, 74b
that is more positive at the outputs of inverters 661, 66b
than at the gates of FET's 72a, 72b. Thu-~, as the capacitors
Ca, Cb discharge, the discharge current produced by such
capacitors Ca, Cb, also decreases and the voltages at the
gates G increase positively. By making (Cs/2)~R74a = R78a Ca
= R7gb.Cb where: Cs is the capacitance of capacitor Cs; R74a
- 18 -
.~' ' '
~LZ7~ 8~
the resistance of resistor 74a; R74b the resistance of
resistor 74b; Ca the capacitance of capacitor Ca; and, Cb
the capacitance of capacitor Cb; the positive increase in
the voltage at the gates G of FET's 72a, 72b from the
discharge of capacitors Ca, Cb, will balance the drop in
voltage at the gates G from the discharge of capacitor Cs
so that the resultant voltage at the gates G will be sub-
stantially constant during the conducting mode of the module,
Having described a preferred embodiment of the invention,
it is now evident that other embodiments incorporating these
concepts may be used. It is felt, therefore, that this
- invention should not be restricted to the disclosed embodiment
but rather should be limited only by the spirit and scope of
the appended claims.
-- 19 --
