Note: Descriptions are shown in the official language in which they were submitted.
5~6
1 AMPLIFIER ~LFUNCTION DETECTOR
2 Abstract
3 A detection device for sensing malfunctions
4 in a wide band power amplifier, driving the
take-up spool motor of a reel-to-reel tape
6 t~ansport system, is disclosed. A feedback loop
7 is connected rom the output of the amplifier to
8 a malfunction junction positioned at the input of
9 the amplifier. The signal at the malfunction junction
is monitored and compared with a predetermined range
11 of standard reference signals~ Whenever the malfunction
12 junction signal is out of the predetermined range,
13 the power amplifier is turned of thereby preventing
14 damage to the tape. A filter means is positioned
so as to desensitlze the compare means until the
16 ampliEier is operating at steady state.
17 Background of kh _Invention
18 1. F eld of the Invention
19 The in~enti~n relates-generally to circuits
for detecting malfunction in amplifiers and more
21 specifically or an improved circuit for detecting
. . ~
22 amplifie~ malfunckion over a wide range of input
23 signals.
24 2. Prior Art
.
In a general sense, detection circuits
26 may b~ classified into two broad groups. In the
27 first group are circults which protec~ a system
- ~- . . . .
28 rom~transient conditlons or disturbances which
- 29 occur external to the system. Some of the well
. .
-
. '' `"'.',~ '
., . :. . `` : ' '
s~
1 known transient conditions are; over-voltage, under-
2 voltage, over-current and under-current. The general
3 approach in solving the problems in the first group
4 is to sense the abnormal condition at an external
point to the system to be pro~ected and isolating
6 the system once the abnormal condition is sensed.
7 One of the well known procedures is to select an
8 -alternate path to ground once the abnormal condition
9 is sensed.
In the second group are circuits which
11 sense malfunction conditions occurring within the
1~ system to be protected. In other words, the malfunction
13 is generated within the system rather than a mal~unction
14 occurring without the system as in the ~irst group
previously mentioned.
16 The present invention may be useful in
17 ~ystems classified in both the fir.st ~roup and second
18 group, but finds its greatest utility in the systems
19 which are classified ~ithin th~ second group.
In general, the present invention is directed
21 to circuits for sensing ~he malfunction in an amplifier.
~2 More specifically, the invention is directed to
2~ circuits for sensin~ tha malfunction in an ampiifier
24 which drives the take~up spool and supply spool
.
of a reel-to-reel tape transport. An alternate
26 use lS for sensing malfunction occurring in amplifiers
27 which drive the motor, in an X-Y coordinate library
28 system.
29 ` Reel-to-reel tape transport systems are
widely known. In one form a generally cylindrical
_z~
5~
1 mandrel or drum includes a rotating head wheel which
2 carries one or more read/write heads. The magnetic
3 tape engages the mandrel at one point, makes a helical
4 wrap-around at least a portion of the mandrel, and
exits the mandrel at a point which is both axially
6 and circumferentially spaced from the entrance point.
7 The angle of helical tape wrap can vary in accordance
8 wlth design choice, but is usually hetween 180
9 and 360. The head wheel rotates so as to swèep
its magnetLc heads traversely across the tape. The
ll angle at which the head enters and exits the tape
12 may vary in accordance with design choice, from
13 slightly less than 90 to a small angle, su~h as
14 15.
A supply spool and a take up spool is
16 positioned so as to bring the tape in transducing
17 relationship with the rotating head. Each spool
18 is driven by a motor. The motor is in turn driven
l9 by power amplifiers. ~ue to the accuracy which
is required in controlling the tension in ~he tape,
21 any malfunction in the power amplifiers must be
22 instantaneously detected. Failure to detect malfunction
23 will result in snapping or breaking of the tape.
24 ~arious techniques have been used in the
prior art for detecting ampliier malfunction or
26 for detecting errox wi~hin a servo control system.
27 In one scheme a predetermined constant error is
28 generaily introduced by conventional means. For
29 example, if a constant load is belng driven by the
". .
~3~
'"' ~ .' .
1 servo system, a torque motor is used to apply a
2 torque opposite to the direction in which the load
3 is moved. The torque is used to displace a potentio-
4 meter arm and an electrical error signal is introduced
into the system. The error signal is fed back into
6 the system and is monitored. Any deviation from
7 the known error signal is a wa nin~ that the system
8 is malfunctioning.
9 Alfhough the prior art devices emboclying
the above enumerated scheme operates satisfactorily
11 ~or the intended purpose, there are several problems
12 which plague these devices. One of the problems
13 is the high cost and bulkiness of these prior art
14 devices. The high cost stems from the fact that
lS in order to intro~uce the constant error into the
16 system, additional hardware is required; for exampler
17 the torque motor and the circuit for converting
18 the torque into an electrical signal. The aaditional
19 cost of this hardware increases the overall unit
cost. Also the additional hardware adds to the
21 overall size and hence, the bulkinPss of the unit.
22 As is well known to those skilled in the
23 art, the present demand is for low cost miniaturized
24 components. With this restraint, the prior art devices
are not suitabl~ for several applications.
26 Another drawback with the prior art servo
27 monitoring scheme is that the load which is driven
28 by the servo system has to be constant. In a servo
.
-4-
.
1 system where the load is varying, th0 prior art
2 scheme will not function satisfactorily to detect
3 malfunction in the system. There are several applications
4 wherein the malfunction detector must detect malfunction
in a servo system which is driving a variable load.
6 For example, a reel-to-reel servo control system
7 wherein the load (magnetic media) on the take-up
8 spool varies.
g Another prior art scheme which has been
used to detect amplifier malunction in a servo
11 controlled system is the so-called random sampling
12 approach. This scheme requires a test run to be
13 undertaken at a predetermined time during normal
14 operation of the system.
Although the random sampling scheme has
16 solved some of the problems which were created by
17 the scheme in which error is introduced into the
18 system, several new problems are creat~d.
19 In ordex to perform a test run on the
system, the normal operation of the system has to
~1 - be interrupted~ This interference tends to reduce
2i2 the throughput~ i.e~, the amount of work which is
23 outputted from the system. With a random sampling
24 scheme, the system is interrupted at intervals and
a test run is performed. A known signal is introduced
26 into tha system or amplifier; for example, one volt
27 and the output is monitored. If ~he output ~s within
28 a prescribed range, the system is allowed to run.
29 If the output is out of the prescribed range, the
system is malfunctioning and corrective steps are
5--
1 taken.
2 Another problem is that if the system
3 has intermittent failures, the random sampling technique
4 may not detect these failures. In several applications
S it is important that failures or malfunctions be
6 detected instantaneously. For example, in a reel-
7 to-reel servo control system, any failure occurring
8 in the amplif1er means which is driving the take-
9 up spool must be detected at once or else the media
can be snapped.
1~ Still another problem is the fact that
12 the random sampling technique cannot continuously
13 monitor the system for malfunction detection. In
14 several applications, it is imperative that the
malfunction device be capable of continuously monitoring
16 or else intermittent undetected failure will be
17 disasterous.
.
18 Flnally, the prior art malfunction devices
19 and schemes ara operable ovar relatively narrow
input signal xange. However, there axe several
21 applications wherein monitoring has to be performed
22 over relatively wide signal range (e.g., -20 volts,
23 +20 volts) due to the signal range limitation, the
24 prior art devices are not suitable.
Objects of the Invention
..
26 It is, therefore, the object of this invention
27 to detect amplifier malfunction in a more efficient
Z8 and improved manner than was heretofor possible.
6--
'
1 It is another object of the invention
2 to produce a low C05t miniaturized malfunction detector.
3 It is still another object of the invention
4 to produce a malfunction detector which monitors
malfunctions of an amplifier driving a variable load.
6 It is a further object of the invention
7 to continuously monitor to detect malfunctions in
8 an amplifier. -
9 Still a further object of the invention
is to detect amplifier malfunction over a wider
11 signal range than was heretofor possible.
12 5ummary of the Invention
13 T~e present invention overcomes the drawbacks
14 found in the prior art by means of an improved circuitry
which continuously monitors the operation of a wide
16 band amplifier means to detect malfunction. The
17 output of the amplifier means is fed back via an impe-
18 dance circuitry to a summing junction positioned
19 at the input to the amplifier means. Whenever the
summing junction signal falls out of range of acceptable
21 signal, the power supply is shut down~
22 In one embodiment of the invention, a
23 filter means having a time constant equivalent to~
24 the dominant pole of the amplifier is positioned
between the malfunction junction and the error detection
26 means. This embodiment improves the reliability
27 of the error detection scheme in that the error
28 detection means is desensitized unt~l the amplifier
.' ~ .
--7--
' ' ' , ''.
....... ., ., ...... , . ., .. .. , .. . ~.. .. . ............ , . .. ,.~,.", , .
1 is operating at steady state. With this design
2 false alarm due to amplifier slew rate or high
3 frequency triggering is eliminated.
4 In another embodiment of the invention
a gain switch means is positioned between the mal-
6 function junction and the filter means. The switch
7 is in a non-inverting gain configuration. This
8 gain switch means widens the reference band so
9 that the amplifier may operate with an overdrive
siynal in a saturation state without generating
11 a malfunction signal.
12 The foregoing and other objects, features
13 and advanta~es of the invention will be apparent
14 from the following more particuli~r description
o~ preferred embodiments of the invention, as illustrated
16 in the accompanying drawings.
17 Brief Description of the Drawin~s
18 FIGURE 1 is a perspective view of a reel-
19 to-reel tape control system with the malfunction
d~tector of ~he present invention in block diagram
21 form. ~
22 FIGURE 2 shows a more detailed view of
23 a motor power amplifier and the malfunction detector
24 with filtering means.
FIGQRE 2A shows an alternate embodiment
26 of a filter means.
.
27 FIGU~E 2B shows one type of voltage-follower ~
28 filter means~ `
29 FIGURE 3 shows ~he malfunction detector
with gain switch means.
--8--
... ..
1 DescrIption of the Preferred Embodlment
2 FIGURE 1 illustrates the simplest form
3 of the malfunction detector as it is incorporated
4 in a reel-to-reel tape transport system. In the
drawings similar elements are identified by the same
6 numbers~ In its simplest form, the tape transport
7 unit comprises magnetic media 10. One end' of magnetic
8 media 10 is wound onto take-up spool 12. Driving
9 means 14 can rotate spool 12 in a clockwise or counterclock-
wise direction. Whenever spool 12 is rotated in
11 the clockwise direction, the amount of maynetic
12 media 10 on the spool increases. Likewise, if spool
13 12 is rotated in khe counterclockwise direction,
14 the amount of magnetic media 10 on the spool decreases.
With this configuration, the combination of spool
16 12 and magnetic media 10 represents a variable load
17 means for driving means 14. The other end of magnetic '
18 media 10 is wound onto a supply spool (not shown).
~9 A more detailed description of a reel-to-reel system
in which the malfunction detector can be utilized
21 is'found in U. S. Patent 3,809,335 entitled ''Web ~ -
22' Movement Controlled in a Reel-to~Reel Web Transport"
23 issuad to J. P. Mantey and assigned to the assignee
24 of the present invention. Also, publication entitled
"Tachometer Feedback in a Reel-to-Reel Control System"
26 by N. H. Hanse~, Jr., published in IBM TechnicaI
27 Disclosure Bulletin, Volume 17, No. 11, April, 1975,
28 page 3194, discloses a similax reel-to-reel'system.
_9_
,
1 Although the invention is described in
2 relation-ship with a reel-to-reel tape transport
3. system, this should be construed as illustrative
4 and not a limitation on the scope of the in;ention;
since the malfunction detector finds use in any :
6 application where a pow~r amplifier has to be monitored
7 or malfunctioning~
8 Still referring to FIGURE 1, the malfunction
9 detection,is comprised of power amplifier 160 The
output of power amplifier 16 is connected to driving
11 means 14 and output signals which control driving
12 means 14. Power amplifier 16 is comprised of operational
13 amplifier 18 connected in series with power booster
14 20. Operational amplifier 18 is comprised of a
negative input terminal 22 and a positive input
.16 terminal 24. Positive input terminal 24 is tied ~ .
17 to ground. . : .
18 Malfunction junction 26 hereinafter
19 called malfunction detection means 26 is connected
to negative input terminal 22. Basically, malfunction
21 detection means 26 is a summing junction. It accepts
22 a feedback signal from terminal 28 which ties output
23 of powsr amplifier 16 to malfunction detection means
24 26 via .feedback resistive means 30. -Also, malfunction
detection means 26 accepts a signal from terminal
26 32. The signal on terminal 32 is derived from an .
27 external driving source, for example, Ein tsee FIGURE
28 2). The external driving source is connected by input
- : ~
-10-
l resistiv~ ~leans 34 to terminal 32. Malfunction detection
2 means 26 sums the two signals, one on terminal 28
3 and the other on terminal 32 and outputs an error
4 signal. The error signal is used for driving power
amplifier 16. The error signal is also fed into
6 malfunction detector 36 via terminal 38. As will
7 be explained subsequently, the error signal on terminal
8 38 lS compared with a reference range of signal
9 in malfunction detector 36 and a malfunction signal
is outputted on terminal 40 whenever the error signal
ll is not within a predetermined range. The malfunction
12 signal is used tv either shut down the power supply
13 or inform a microprocessor that power amplifier
14 16 is defective.
As will be explained suhsequently, the
16 input signal is picked off by terminal 42 and fed
17 into malfunction detector 36. Mal~unction detector
18 36 has circuitry which will monitor and control
19 the signal on terminal 42 so that power amplifier
16 can be driven into saturation without outputting
21 a malfunction signal on terminal 40.
22 Referring now to FIGURE 2, section 44
23 of the malfunction circuitry of malfunction deteetor
~4 36 is shown in gxeater detail. Malfunction circuitry
44 is comprised of filter means 46, compare means
26 48 and OR circuit means 50. Filter means 46, compare
27 means 48 and OR circuit means 50 are connectad in
28 tandem. The output from OR circuit means 50 appears
29 on terminal 40 and indicates when-there is a malfunction.
35~P~
The main theory underlined detection of
2 circuit malfunction is that in normal operation
3 of power amplifier; the error voltage ES (see FIGURE.
4 2) which appears on terminal 38 from malfunction
detection means 26 will be very close to zero (offset
6 term plus output/open loop gain). A malfunction
7 will cause Es to rise above or fall below a few
8 millivolts of ground. By monitoring the fall or
9 rise and comparing it with a standard range of reference
voltages, one can determine when malfunction occurs
11 in power amplifier 16.
12 ,However, it is well known that at initial
13 startup; i.e., whenever the ampli~.'ier is running
14 at high fxequencies or if the slew rate of the amplifier
is exceeded, compare means 48 will output a signal
16 which'will activiate OR circui~ me~ans 50 and output
17 a malfunction signal on terminal 40 when in fact
18 there is no malfunction.ing in amplifier 16. To
19 alleviate this false alarm, filter means 46 is incorporaked
as shown in FIGU~E 2.
-21- ~ Referring now to FIGURE 2A and FIGURE
22 2B alternate embodiments of filter means 46 are
2~ shown.
24 . ' In FIGURE 2A, filter means 46 is comprised .-
of a conventional RC filter 56. Filter 56 is comprised
26 ' of terminal BB and resistor 60 in series with capacitor
27 62. The time constan~'of filter 56 is substantially -
.
28 ' equivalent ~o~the.dominant pole of the ampliier.
~ '
. -12
:
1 Whenever filter 56 is substituted for filter means
2 4 6 in FIGURE 2, the signal on terminal 64 will not
3 be outputted until power amplifier 16 is running
4 ~t steady state. With this configuration, any transient
signal which is caused by s~artup will not be able
6 to reach compare means 48 and therefore the reliability
7 of the system is enhanced.
8 FiLter 58 (FIGURE 2B) is an a~ternative
9 to filter 56. Basically filter 58 is comprised of
10 - terminal CC~ voltage follower 66, resistor 68 and
11 capacitor 70. Voltage follower 66 is connected
12 in series with resistor 68, and the combination
13 is in series wi*h capacitor 70.
14 ! , Although filter 56 and ~ilter 58 are needed
for accurate malfunction detection, there are several
16 applications wherein a filter is not required. For
17 example,-if the input signal (Ein on terminal 32)
18 is restricted (i.e.; no high frequencies and/or
19 low slew rate) then a through wire, instead of ilter
56 and filter 58, càn be used to connect terminal
21 38 to termlnal 64. In one embodiment a straight
2? ~hrough wire is used to connect terminal 38 to terminal
23 64 and there was no degradation in the malfunction
24 detection-circuitry.
The signal from filter means 46 appears
26 on terminal 64. Termlnal 64 .i5 attached to comparator
27 means 48. Comparator means 48 is comprised of comparator
28 72 and comparator 74. Comparakor 72 is the positive
29 comparator and will output an error slgnal on terminal
76 whenever the error voltage ES from malfunction
-13-
.
.. ' '
1 detection means 26 exceeds the allowable few millivolts
2 range in the positive directionO Comparator 72
3 has a positive terminal which is tied to terminal
4 64 and a negative terminal which is tied to ground
via positive reference batterv 78. The voltage
6 on battery 78 is equivalent to the maximum possible
7 voltage which is allowable for proper operation
8 of the ampliiier 16. In other words, whenever errox
9 voltage Es exceeds the vo].tage on battery 78 a malfunction
si~nal will appear on terminal 40. For example,
11 in one embodiment battery 78 was set with 50 millivolts.
12Likewise, comparator 74 is a conventional
13 negative comparator wi.th its negative terminal tied
14 to terminal ~4 and its positive terminal tied to
ground via negative reference batt:ery 80. Negative
16 reference battery 80 stores a negative voltage,
17 the value of which is equivalent t:o the maximum
lB negative swing which.can he experienced by malfunction
19 detection means 26. Whenever the swing in th~ negative
direction at malfunction detection means 26 exceeds
21. the value on negative reference battery 80,.an error
.
22signal is outputted on terminal 82. The signals ~:
23 on terminal 76 and terminal 82 are ~ed into OR circuit
24 means 50 which in turn outputs a signal on terminal
25 40 indicative of a malfunction.
26. Referring now to FIGURE 3, malfunction
27 detection circuitry B4 of malfunction detector 36
28 is shown~ As was mentioned previously, in some
29 operation ~for exampler when amplifier 16 is used
to control the drive motor of a library system~,
1 it is advan~ageous to operate the amplifier in a
2 saturation state during peak acceleration periods.
3 This can be done at the expense of reduced sensitivity
4 to malfunct~on detection. To overcome this lack
of-sensitivity, malfunction detection circuitry
6 84 hereinafter called gain switch stage 84, may
7 be incorporated as shown in FIGURE 3. Essentially,
8 gain switch stage 84 monitors the error signal on
9 terminal 38 and the input signal on terminal 42.
As long as the input signal on terminal 38 is below
11 a predetermined level, amplifier 16 should not be
12 in saturation, and buffer amplifier 86 will provide
13 a relatively high gain for the ma:Lfunction signal
. . .
14 on terminal 38. Therefore, for relatively low and
moderate signal levels, a high degree of sensitivity
16 ma~ ~e maintained. At higher input signal levels;
17 i.e., signal levels which drive amplifier 16 into
18 sat~uration, the gain switch s age will reduce the
19 gain of buffer amplifier 86. This has the effect
of widening the sample window or reference band
~1 to handle the overdrive signal. However, the reduced
22 malfunction detection sensitivity will only be felt
2~ at high signal levels. By positioning the gain
~4 switch stage prior to filter means 46, switch transient
will not cause false signals.
26 Essentially, gain switch stage 84 is comprised
27 of buffer amplifier 86, a single pole double throw
28 gain switch means B8 and a~ magnitude threshold circuit
29- ,means 90. Gain switch means 88 can also be a single
pole single throw switch. Magnitude threshold circuit
-15- -
.
l means 90 is a compare circuitxy which is substantially
2 analogous to compare means 48. It stores the minimum
3 or threshold value of the input signal which wîll
4 drive amplif.ier 16 into saturation. The stored
value is then compared with the actual input signal
6 which appears on terminal 42, and whenever the value
7 of the signal on terminal 42 is equal to or greater
8 than the threshold value a signal is outputted
9 .on terminal 92 which switches gain switch means
.10 88 to the high terminal and thereby forming a conductive
ll path between ground and buffer amplifier 86. Likewise,
12 if the value of the input signal on terminal 42
13 is below the value stored in threshold circuit means
14 90, gain switch means 88 will be in contact with
the low terminal and therefore there will be no
16 . conductive path between ground and buffer amplifier
17 86.
18 Bu~fer amplifier 86 is comprised of a
l9 positive terminal which is connected to terminal
38 and a negative input which is connected to terminal
21 94. The output of buffer amplifier 86 is fed back
.22 to terminàl 94. The output of buffer amplifier
23 . 86 is fed back to terminal 96 via a feedback resistor
24 98. Terminal 96 is connected to gain switch means
88 via lead lO0 through resistor 102 hereinafter ~ .
26 called second resistive means 102. Likewise, terminal
27 96 is tied to terminal 94 via lead 104, while terminal
28 94 is tled to gro-md through resistor 106, here- -
2g inafter.called first resistive means 106. The output
-16-
1 from buffer amplifier 86 appears on terminal 108.
2 This completes the detailed description of the preferred
3 smbodiment
4 peration
As was previously mentioned, the malfunction
6 detector is based on the theory that during normal
7 operation of power ampliier 16, the voltage Es
8 . at malfunction detector means 26 should be ideally
9 zero. However, due to the offset term and output/open
loop gain, Es would swing about the zero point.
11 The swing is generally in the area of plus or minus
12 a few millivolts. By monitoring the.swing in voltage
13 Es and comparing the swing with standard reference
14 .voltage, one is abla to pinpoint malfunction in
amplifier 16 over its entire operational range.
.16 In operation, an input signal Ein is applied
17 to terminal 32. This signal can be derived from
18 any conventional driving means, for example, a voltage
-19 sourceO The inP~t signal is then registered at malfunction
detection means 26. Simultaneouslyr any signal
21 appearing at the output of amplifier 16 i5 fed back
22 through feedback terminal 28 to malfunction detection
23 means 26. Malfunction detection means 26 outputs
. 24 an error signal which is the weigh~ed difference
in ~he sum of the input signal on terminal 32 and
26 the signal which appears on terminal 28. lf the
27 error signal Es is within a few millivolts (i.e.,
28 ~ positive or negative) of ground, the amplifier is
29 operating normally and will not be disturbed.
-17- , . -
-
'
5~
1 The permissible voltage swing of Es is
2 stored on positive reference battery 78 and negative
3 reference battery 80. The voltage Es from malfunction
4 detection means 26 is constantly compared with the
reference voltages stored on positive reference
6 battery 78 and negative reference battery 80. Whenever
7 Es exceeds either of these voltages, an error signal
8 will appear on terminal 40. The signal is used
9 to close down the power supply and/or to signal
the microprocessor.
11 In order to pxevent false alarm due to
12 slew rate or high frequency triggering of the amplifier,
13 filtering means 46 is incorporated so that the compare
14 means will not be activated until the system is
running at steady state. The time constant of the
16 filter is substantially equivalent to the dominant
17 pole of the amplifier.
18 ~s was previously stated, in some operation
19 it is advantageous for one to operate the amplifier
when it is saturated. In order to achieve this
21 goal, gain switch stage 84 is incorporated. In
22 operation, the error signal which appears on terminal
23 38 is fed ~nto buffer amplifier 86 Simultaneously,
24 the input signal Ein is fed into magnitude threshold
circuit means 90. Magnitude threshold circuit means -
26 90 will compare the input signal on terminal 42
27 with a standard predetermined signal. As long as
28 the input signal is not greater than or equal to -
29 the predetermined threshold value, gain switch means
. .
-18- ;
.
. .
1 88 will attach to the high terminal which results
2 in a closed circuit. With the closed circuit configuration,
3 the gain of buffer amplifier 86 is one plus the
4 feedback resistor 98 divided by the parallel combination
o resistive means 106 and resistive means 102:
6 1 ~ R98/ R106 R102
R106 ~ R102
7 Where R98 = feedback resistor 98
8 . and R106 = resistive means 106
9 and R102 resistive means 102
Buffer ampli.fier 86 enables the malfunction detPctor
11 to maintain its sensitivity as long as amplifier
12 16 is not driven into saturation.
13 If the input signal (Ein~ exceeds or is
14 equal to the stored predetermined signal of magnitude
threshold circuit 90, the gain of 'buffer amplifier
16 86 is reducedO Although the malfunction de~ector
17 is le~s sensitive, it will detect any malfunction
18 which will occur while the amplifier is being driven
~9 at a saturated rate. In order to reduce the gain :
of ~uffer amplifier 86, the signal which appears
21 .on terminal 92 will switch gain switch means 88
22 to its low terminal thereby rendering a nonconductive
23 path between ground and buffer amplifier 86. In
24 other words, there ls an open circuit. In the open
configuration, the gain of buffer amplifier 86 is
26 reduced. Theoretically, the gain is equivalent
27 ~: 1 + R98~R106 By effectively increasing the ~ -
28 resistance in the circuit, the gain of buffer ampli~ier
~9 86 is reduced.
.
1 The advantage of the above described invention
2 enables one to monitor a wide band amplifier for
3 malfunction over its entire signal range.
4 By varying the gain of gain switch stage
84 as a function of the input voltage, one can allow
6 the amplifier to be operated in a saturation mode.
7 While the invention has been particularly ~ :
8 shown and described with reference to a preferred
9 embodiment thereof, it will be understood by those
skilled in the art that various changes in form -
ll and detail may be made therein without departing
12 from the spirit and scope of the invention.
.
.
,~ .
.
" ' ~ '. ''' ' , ' ', ' ' ".
- . : . .: ,
: :
,. . .
` , ' ' ~ . .
-20- , , : `
.' . '
.
.
