Note: Descriptions are shown in the official language in which they were submitted.
55~3
71~ ~
BACKGROUND OF_THE INVENI'ION
This invention relates to an ultrasonic ranging
system, and to a camera into which such a system is
incorporated.
Ultrasonic ranging systems fox cameras are
disclosed in U.S. Patent 3,522,764, German Patent 864,048
and I.B~M~ Technical Disclosure Bulletin, Volume 9, No. 7,
December, 1966, pp. 744-745. In each of the systems,
ultrasonic energy is transmitted toward a subject to be
photographed, and the subject reflects energy back to the
camera. Characteristics of the transmitted and received
signals are compared neeessitating separate sending and
receiving transducers, and a control signal representative
of subject distance is produced. The control signal is
used to drive the lens mount of the camera to a position
functionally related to subject distance whereby the sub-ject
will be in focus.
In the U.S. patent No. 3~522J764~ ultrasonic
bursks are transmitted at 40kHz with a period no less than
~he time required for sound to travel twice the rnaximum
subjeet distanee for which the lens mount is to be adjusted.
Eehocs from the subjeet, with the same periodicity as the
transmitted bursts, are received in the intervals between
the transmitted bursts, the time between transmission and
reeeption of a burst being related to subject distance and
used to establish the duty cycle of a first pulse generator.
A seeond pu~se generator, with the same frequency as the first,
is assoeiated with the lens mount, but the duty cycle o~ the
second generator depends on the position of the lens mount.
Movement of the lens mount takes plaee until the duty cycles
of the -two pulse generators are equalized.
~1- ~
S~7~
In the Ge~nan patent and in khe I.B~M. public~tion,
ultrasonic wave trains are frequency modulated with a period
no less than the time required for sound to travel twice the
maximum subject distanc~ for which the lens mount is -to be
adjusted. The echoes from a sub~ect are thus frequency
modulated with the sam~ periodicity as the transmitted
energy, so that the subject range can be established by the
instantaneous di~ference between the frequencies of the
transmitted and received signals.
In addition to requiring two transducers, the
known prior art systems have the disadvantage of requiring
a considerable amount of time to establish subject range.
At room temperature, sound travels at about 340 meters per
second so that the time required fox a burst to reach a
target at 7 meters, which is the maximum distance for which
focusing is usually required, is about 20 msec. Thus, each
period of the above systems must be about 40 msec: and if
ten periods are required to establish subject range, the
about 0.5 second9 is consumed in achieviny camera focus.
Such el~psed time 1~ relativ01y long with respect to human
re~lexes, with the result that this photography has to
proceed in two successive and distinct steps: on~ involving
focusing, and one involving shutter actuation.
For general ranging purposes, an ultrasonic
pulse distance measuring device is disclosed in U.S. Patent
No. 3,454~922 wherein separated fixed frequency bursts are
transmitted by a transducer, and a re~lected echo is
received by the same transducer after a period of time
related to target range.
Wh~n a fixed ~re~uency ultrasonic sotlnd is
utilized, it has been found that subjects within the
acceptance angle of the transducer, and within the field
of view of the camera, are often undetected. From experi-
S mental work~ it appears that reflections from various poin-ts
on the subject may interfere with each other thus cancelling
or so weakening the echo at the receiver, that the latter
cannot respond, and the subject remains undetected. This
phenomenon is most noticeable ~or subjects that are rela-
tively close to the transducer.
In the context of photographing a subject, the
latter is considered to be close to the camera when it is
within 2 meters from the camera. Since many photographs
are taken with subjects at this relatively close range,
the failure of the transponder to receive an echo from a
relatively close subject would normally cause the lens
mount to be improperly positioned.
The side lobes associated wi-th the radiation
pattern of ultrasonic transmission give rise to another
problem. An o~-axis target relatively closo to the trans-
ducer, and located within an attenuated side lobe of the
antenna pattern~ may have a surface condition or other
characteristic which produces an echo of a strength comparable
to a subject located on-axis at a considerable distance from
the tran9ducer. In such case, the lens mount of -the camera
may be set in accordance with the distance to the off-axis
target rather than the sub~ect being photographed, and an
unfocused exposure would result. While the transducer can
be made more directional by increasing its area, this expedient
increases the size of the ranging equipment associated with
the camera, and neutralizes one of the basic reasons for going to a single
~ransduce~ which is to reduce the size and weight of the ranging equipment
It is there~ore an object of the present invention to provide a
new and improved ultrasonic ranging system wherein the problems and the
deficiencies outlined above are reduced or substantially overcome.
It is a further object of the invention to provide an improved
automatic focusing camera.
Another object is to provide a method of ranging for photographic
operations.
SUMMARY OF THE INVENTION
.
According to a broad aspect of the present invention, there is
provided in a sonic ranging system including means actuatable for transmit-
ting a burst of sonic energy toward a subject, for receiving and processing
an echo ~rom said subject, and for producing a range signal related to the
range of said subject) the improvement comprising means for controlling said
transmission means to produce a sonic burst having one portion varying in
frequency and another portion o~ relatively constant frequency.
The present invention provides an ultrasonic rangillg system for
a camera having a lens mount moveable to a position at which a subject being
photographed is in focus. The system includes an ultrasonic transducer that
responds to a keying pulse by transmitting a relatively short burst of fre-
quency modulated ultrasonic energy, and a synchronized receiver for process-
ing an echo signal produced by the transducer on receipt of an echo within a
predetermined time interval (termed the receiver ranging time) following the
burst. The receiver produces a range signal wi~h a characteristic linearly
related to the distance of a subject being photographed from the camera. ~or
receiver synchronization, a variable Q filter in the receiver filters echo
signals produced by the transducer, with the Q increasing during the predeter-
mined time interval thereby avoiding the need for a matched filter.
It has been found, experimentally, that the strength of the return
from close subjects is highly
_~_
B
)S'7i3
dependent on the frequency oE the ultraso~ic burst. If a~
one frequency the echo would be very small due to inter-
ference~ there are other frequencies at which the echo
would be large. On the other hand, the return from remote
subjects is much less dependent on frequency.
The preferred form of the frequency-modulated
burst is a leading half in the form of a chirp (using radar
terminology) wherein the frequency decreases with time,
and a trailing half in the form of a constant frequency
equal to the lowest frequency of the chirp. Because of
the chirp, energy at many different frequencies will be
incident on a subject increasin~ the probability that some
of the frequencies will be reflected back to the transducer
by relatively close subjects even if other frequencies are
lost because they set up cancelling interference patterns.
~he preferred format of the burst is to be distinguished
from the approach taken in U.S. Patent No. 2,433,78~ whic~
shows a single bursk, frequency modulated ultrasonic rangincJ
~y~tem without a constant requ~nc~ portion. In -the pr~sent
invention, the frequency values and their time-wise distri-
bution are salected to minimize both the ef~ects of absorp-
tion at all ranges and the cancell~tion of echoes at close
range without adversely affecting the signal-to-noise ratio
for long range echoes.
During the initial portion of the receiver ranging
time in which any return will be from a relatively close
subject and will likely include many di~erent frequencies
in the chirp, the Q of the filter is made low insuring that
the filter bandwidth will be wide enough to accommodate all
of the chirp fre~lencies. The low Q allows more noise
Cl 5~7~3
reception, however, the latter is offset by ~he strong sig~l
from close subjects. During -the final portion of the
receiver ranging time in which any return will be from a
relatively distant subject and will, due to absorption
probably exclude the higher chirp frequencies, the filter
is altered to a higher Q with the bandwidth narrow and
centered on -the constant frequency. Thus the receiver has
reduced sensitivity to the effects of cancellation with only
a slight compromise in the signal-to-noise (S/N) ratio for
remote subjects.
The lower value of Q of the filter during the
initial portion of the receiver ranging time has the added
benefit of decreasing the response time of the filter (more
rapid rise in fil-ter response) for echoes from sub~ects close
to the camera as compared to subjects more remote therefrom.
Consequently, the accuracy will be better ~or subjects that
are clo~e than for more remote subjects, an advantageous
situation in photography w~ere the focus is more sensitive
-to exrors in the len~ mount position for close subjects.
rrhe admittance of the variable Q filter increases
duriny the receiver ranging time causing the receiver, in
effect, to amplify echo signals from remote subjacts more
than nearby subjects. This technique decreases the angular
sensitivity of the transducer and discriminates against
echoes from nearby off-axis targets which result from side
lobes of the angular field pattern of the transducer~
The receiver of the present invention may also
include detector means responsive to the output of the filter
for producing a range signal when the output exceeds a pre-
dete~nined level. The time of detection measured Erom the
keying pulse is representative of -the subject aistance.
i'7~
I~he invention also consists in a camera havin~
an ultra~onic ranging system like that described above.
Such camera has means responsive to the range signal for
moving the lens mount of the camera to a position determined
by the delay between the keying pulse and the range signal.
~S
Embodiments of the present invention are illus-
trated in the accompanying drawings wherein:
FIGo 1 is a block diagram of a general form of
an ultrasonic ranging system according t~ the present
invention showing the system incorporated into a camera;
FIG. 2 is an idealized response diagr~m of a
variable Q filter,
~IG. 3 is a detailed circuit diagram of a pre-
ferred form of the ranging system according to the present
invention;
FIG. 4 is a waveform diagram showing idealized
waveforms occuring at various locations in the ~ystem oi
Figure 3;
FIG, 5 is a circuit diagram of a voltage gen0rator
employed in the system of FIG. 3;
F~G. 6 is a block diagram of an alternate embodi-
ment of the transducer drive shown in the ranging system
of Figure 3; and
FIG. 7 is a waveform diagram of -the ultrasonic
hurst produced with ~he embodiment of Figure 6.
DESCRIPTIOI~ OF THE PREFERRED EMBODIMENT
Referring now to FIG. 1, reference numeral 10
designates a camera into which an ultrasonic ranging sys-tem
11 according to the pr sent in~ention has been incorporated.
~z~
Camera 10, which is shown in schematic ~orm, includes
housing 12 within which film 13 is supported opposite
lens mount 14 which is axially displaceable along the
optical axis 15 between spaced terminal positions. At
one terminal posi~ion, lens mount 14 is posi~ioned to
focus subject 16 onto the plane of film 13 when the subject
is close to ~he camera, e.g., about 25 cm away. At the
other terminal position of lens mount 14, subject 16 will
be in focus when it is located beyond say 7.5 m from the
camera. The position of lens mount 14 between the two
terminal positions for bringing a subject into focus, is
a predetermined function of the distance to the subject,
such function being highly non-linear and being termed the
lens/subject function.
In a manner to be described below, ultrasonic
ranging system 11 produces a range pulse 17 that is
delayed with respect to a keying pulse 18 by a period of
time linearly related to the distance of subject 16 from
the camera. ~ocusing mochanism 19 associated with the
cant~ra responds to the pulses 17 and 18 by moving the lens
mount 14 to an axial position at which subject 16 will be
in focus.
The focusing mechanism includes a logic circui*
20 which, in response to a range signal 21 generated
by range pulse generator 22, produces a train of
- 8 ~
~lfF~ i'7~
puls~s whose number is representative of the axial pos:ition
of the lens mount at which the subject will be in focus~
Such pulses are gated inko counter 23 and used for the
purpose of driving motor 24 which is mechanically con~
nected by means 25 to the lens mount 14. In addition,
means 25 is also connected to a feedback sys-tem such as
an auxiliary pulse generator 26 so that rotation of motor
24 under the control of the contents of counter 23 causes
auxiliary pulse generator 26 to produce a predetermined
number of pulses for each unit displacement of lens mount
14. Logic means 20 responds to the output of auxiliary
pulse generator 26 for the purpose of determining when
lens mount 14 has been moved to the position determined
by the contents of counter 23 thus bringing the subject
into focus.
rrhe ultrasonic ranging system of the present
invention designated by reference numeral 11 includes
ulkrasonic transducer means 27 which may include an electro-
stat.ic transducer element o~ the Sell-type similar to that
disclosed in the article: Geide) K: "osc.illation Char-
acteristics of Electroacoustic Transducers using the Sell
Principle", Acusti~a, Vol. 10, pp. 295-303 (19603. The
nature of the main and sids lobes of the preferred form
of element 27A depends on the output pattern of the backing
plate (not shown) of the element. For a transducer of
given siæe driven at a given frequency, the narrowest beam
is produced by a constant outpu-t across the transducer
element. For example, a transducer of the type disclosed
in the above article, with a 3.5cm diameter active area
driven at 50kHz, has a half power angle o~ 6" off-center.
,9_ ,
1~L;2q~5'7f3
Th~ ~:irst zero occurs at 'L3 and t~he Flrst si~e lo'be at 1~
These anyles are about inversely propor-tional to the diameter
of the transducer and to the frequency and -the first side
lobe may have a relative power of -17.6dB for transmit ancl
receive conditions; combined the relative power for the
system is approximately -35dB. Improved patterns would
have slightly larger angles but smaller side lobes
Transducer means 27 is physically Located
adjacent lens mount 14 and has a transmission pattern 28
with a main lobe 29 closely matching the field of view 30
of the lens mount. Associated with -the main lobe of the
pattern are side lobes 31J ~he precise form of the main
and side lobes depending upon the specific design of the
transducer element.
Ranging system 11 also includes control voltage
generator 35, frequency modulator 32 f'or driving transducer
means 27 and causing the latter to transmit a 'burst of
ultrasonic energy towards subject 16 in response to a ]ceying
pulse 18 applied to yenerator 35, and receiver 33 for
20 ~ proceCsing an echo signali~t produced by the transducer in
response to receipt of an echo Erom the subject within a
predetermined interval of time (hereinafter t0rmed the
receiver ranging time) following the burs~.
In ~peration, a manual input to the system, such
a~ for example, the depression of a camera pushbutton (not
shown), is converted, by leading edge detector 34, into
keying pulse 18 which is applied to control voltage gen-
erator 35. The output of generator 35 controls frequency
modulator 32 which causes transducer means 27 to produce a
frequency-modulated burst. Generator 35 is eEective to
-10-
517&~
modulate the output voltage o e modulator 32 such that,
during half the transducer hurst, the frequency changes
between the limits of 65 to 50 kHz; and during the other
half o the transducer burst, --he frequency remains constan-t
at ahout 50kHz.
In view of experimental results showing that the
return from a close subject is highly dependent on the
frequency of the incident ultrasonic burst in the sense
- that cancellation of an echo can occur at certain freq-
uencies, the provision o~-the chirp will insure the presence
of many frequencies incident on the subject. At least some
of the frequencies will be reflected back to the transducer
without being cancelled. The provision of the 50k~z con-
stant frequency during hal~ o the burst minimizes the
effects of absorption on ultrasonic energy thus insuring
a return from a remote subject under adverse ambient con-
ditions. It is known, ~or example, that the reElected
signal power varies exponentially with the di~t~nce o~ th~3
subject and approximately invers~ly with the ourth power
of the distance to the subject. For example, an approxi-
mately 60dB variation in reflected signal power occurs
when a subject is moved from 25cm to 5m using a 50kHz
signal at 20C. It has been found from experiments that
the ~bsorption, and the variation of absorption, with
temperature and humidity, increase rapidly with frequency.
Generally speaking, the lower the frequency, the lower the
absorption. For the frequencies of the preferred burst,
the lowest absorption, for a given temperature and
humidity, occurs for the 50kHz signal.
)5 ~
IJncler adver~e conditions of temperatUxa and
humidity, the higher fr~quencies in the burst are likely
to be attenuated. Hence, thev will be most effective for
clo.se subjects, which is precisely where interference is
encountered were a single frequency used in the burst. The
50kHz portion of the burst is the least attenuated of all
of the fre~uencies present and so is suitable for subjects
remote ~rom the camera.
When subject 16 is relatively close to transducer ~-
means 27~ the frequencies in the return signal incident on
the transducer element wlll contain most of the frequencies
in the chirp except those ~requencies which have been
cancelled by reason of interference. When the sub~ect 16
is located more remotely from transducer means 27, the
return signal is most likely to contain those ~requencies
least attenuated by the environment, namely the ~requencies
close to the lower frequency o the chirp.
The 65 to 50kHz chirp portion and the 50kHz
constant portion of the burs~ can be arranged in eight
dif~erent combinatlons. Half of these lnvolve ~tep~type
discontinuities at the transition between the two portions.
For reasons relating to simplification of the electronics
for developing the control voltage that drives the modulator,
the latter discontinuous hurst arrangements are not preferred.
Turning to the other four possible (continuous)
burst arrangements~ it shoula be noted that th0 chirp
portion could rise or fall to the constant frequency and
could precede or ~ollow the constant frequency portion.
Now, at near distances, interference or so-called spsckle
is a problem while at far distances the signal-to-noise
57~3
ratio is of most concern. Based upon the above, it is
preferred to have the chirp firs-t which provides the
greater ranging accuracy of operating with the leading
edge of the burst thereby permitting greater ~ocusing
accuracy desired for close subject dis-tances. On the
other hand, since lower frequencies are less absorbed,
their use is preferable for far distances whare signal-
to-noise ratio is important. Consequently, a leading
chirp which falls to a lower constant frequency was
preferred with the frequency~of the burst to start at
65kHz and drop to 50kHz in 0.5 msec, and then remain
c~nstant at 50kHz.
Referring again to FIG. 1, Xeying pulse 18, upon
initiating the operation of control volta~e generator 35
and the transmission of a frequency-modulated ultrasonic
burst from transducer means 27, is also applied to blanking
gate 36 of the receiver 33. Blanking gate 36 produces a
level khat is app~ied to the output o~ the receiver 33 so
as to enable the output approximatel~ 0.4ms ~ollowiny the
tormination o~ the transducer burst, and then maintains
the output in active operation for a predetermined period
of time, termed the receiver ranging time, which is preferably
about 40ms in duration. In this interval of time, sound at
sea level at 20C tra~els from the transducer to a target
located about 7.3m, and returns to -the transducer. The
0.4ms delay in enabling the output provides suf~icient time
for the transducer element of transducer means 27 to
stabilize following termination of the burst. As a result,
the delay time defines the closest subject distance that
can be accommodated by the ranginy system, n~nely about
-13-
~Z~S'7~
25cm. AS shown in E'IG. 1, the output of the blankiny yate
36 may also be applied to a pre~ampli~ier 37 to enable
the latter following the indicated delay.
An echo signal~ produced by transducer means 27
on receipt of a return from subject 16, is applied via ~
line 38 to the pre-amplifier 37 whose output is processed
by filter network 39 which as later explained in detail
with regards to FIG. 3 includes variable Q filter 40, and
amplifier 41 whose gain is optionally variable. Level
detector 42 produces a range pulse 17 when the output of
variable gain amplifier 41 reaches a threshold level.
Associated with variable Q filter 40 is programmed
Q control circuit 43 which is responsive to keying pulse 18
for increasing the Q of the filter during the receiver ranging
lS time. The center frequency of the filter is the lowest
frequency of the burst, namely 50kHz in the present situation
Control 43 may typically change the Q o the filter 40 from
a value of 5 to a value of 70 during the receive~ ranging
time.
The admittanae of Q ~ilter 40 as a function of
frequency is shown in FIG. 2 for diffexent value~ of the
parameter Q. When tha Q of the filter 40 is low, as indicated
by curve 44 in FIG, 2, the bandwidth of the filter 40 will
be relatively wlde, and in factj will be sufficiently wide
to pass all of the chirp frequencies~ The Q of the filter
40 i9 rela-ti~ely low during the initial portion of the
receiver ranging time within which~subjects close to the
transceiver will supply an echo return to transducer means 27.
When the ~ of the filter 40 is relatively high
during the terminal portion of the receiver ranging time,
-14-
1)S'7~3
the bandwiclth o~ the ~ilt~r will b~ rclatively narrow and
can be optimized with respect to slgnal-to-noise ratio.
The 50kHz portion of the burst will be most effective in
reaching a subject remote from the transducer and will be
strongly present in any return therefrom. Since the
relatively narrow bandwidth of the filter occurs durlng
the latter portion of the receiver ranging time, this is
consistent with subjects remote from the camera.
As indicated in FIG. 2, the admit-tance of filter
40 when its Q is low will be significantly less than its
admittance when the Q filter is relatively high. Con-
sequently, the impedance of filter 40 will be higher during
the initial portion of the receiver ranging time than during
the final portion thereof. This has the effect of attenuating
the output of pre-amplifier 37 for echo signals produced
when a subject is close where the amplitude of the echo is
likely to be large. The output of filter 40 thus tends
to have a level independent of subject range.
In generalJ howe~er~ the gain of amplifier 41,
which is a part o overall filter network 39, can he
programmed using gain control circuit 46 associated with
this amplifier. Control circuit 46 responds to the appli-
cation of keying pulse 18 by producing a control signal which
causes the gain of amplifier 41 to increase during the
receiver ranging time. As a consequence of this operation~
relatively weak echo signals associated with objects
relatively remote from the transducer will be amplified
to a greater extent than relatively strong echo signals
from a subject closer to the transducer.
-15-
- lZ~Si7E~
Tho advantages :Eor varying the gain of th0 overall
filter network i~ explained in connection with the presence
o~ side lobes 31 shown in FIG. 1, and the pres~nce of target
16A within one of the side lobes. Since -target 16~ is out
of the field of view of the objective lens of the camPra,
it is highly desirable for the ranging system to discriminate
against target 16~ i.n favor of subject 16 which is within
the field of view. The variation in admittance of filter 40
alone, or together with the variation in gain of amplifier
41 if filter 40 is not sufficient, cooperates in achieving
this discrimination as explained below~
The return from target 16A will reach transducer
means 27 before the return from subject 16 which is more
remote from the transducer than target 16A. Not only will
the signal strength of the return from target. 16A be
relatively low by reason of the strength of the side lobe,
but the operation of filter 40 and amplifier 41 wlll be
such as to insure that the signal reaching level detector
42 will be below the threshold of the detector. By the
time t~e retu.rn from 3ubject 16 rsaches transduce~ means 27,
the admittance of the filter will have increased (which
means -that the impadance of the fil-ter 40 to the echo
signal will have decreased) from its previous value presented
to the return ~rom target 16A. In addition, the gain of
amplifier 41 will have increased from its previous value.
Conssquently, the output of amplifier 41 will exceed the
threshold of detector 42 and cause range pulse 17 to be
produced at a point in time corresponding to the range of
subject 16 from the transducer.
-16-
5i78
In addition to the change~ in ~ o ~1'L-t~r 4() J
which change the admittance o~ the filter in a way tha-t
beneficially decreases the angular sensitivity of the
transducer and discriminates against off-axis targets,
another beneficial result is achieved. Such result arises
because the rise time of filter 40 is ~aster when the ~ is
relatively low than when the Q is higher. The relatively
faster rise time thus occurs in connection with echo
signals associated with objects relatively close to the'
camera. Since the rise time allows early detection of the
leading edge of the echo~ the more rapid ri~e time results
in greater accuracy for the generation of range pulses
associated with close subjects. This is consistent with
the requirements for a camera since ~ocus is more sensitive
to errors at close range than to errors at distances more
remote frQm the camera.
The pre~erred construction of an ultrasonic
ranging system is shown in FIG. 3, and is designated by
reerence numeral llA. System llA includes ultrasonic
transducer element 27A, modulator 50 for driving the
element and c~u~ing it to transmit a frequency-modulatad
burst of ultrasoni~ energy toward a subject in response
to the application of keying pulse 18 to the modulator,
and receiver 33A, including filter 51, for processing an
echo signal produced by element 27A in response to its
receipt of an echo from a subject ~not shown~ within a
predetermined time interval following the burst (i.e., within
the ranging time of the system). Receiver 33~ produces a
~ range ~ L 17 delayed with respect to keying pulse 18 by
a period of time ~ linearly related to the subject distance
(i.e., twice the time ~or sound to travel between the trans-
ducer and the subject,).
l7-
r~lrning aga.in to FIG. 3, a modulator 50 incluaes
gate generator 52~ voltage generator 53, voltage controlled
oscillator 54, amplifler 55, transformer 56, and decoupling
diodes 57. A keying pulse 18 applied at ~b) to the inpu-t
of gate generator 52 causes the latter to produce a gate
signal 58 at (c~. A single shot multivibrator with auto-
matic delay reset or an RC delay in combination with a
Schmitt trigger circuit may be employed ~or the gate gen-
erator 52. As shown in FIG. 4(c~, gate signal 58 has a
duxation of about 40 msec, corresponding to the time
required for sound to travel about 7 meters from the
transponder to a target and back again~ such distance
corresponding to an infinity focus position of the lens
mount. For subjects beyond 7 meters, the lens mount is
positioned at its infinity ~0CUS3 and this would cause
the subject to be in focus.
In response to the gate signal 58, voltage gen-
erator 53 produces the time-variable voltage pul8e 59 shown
in FIG. 4(d). ~n exemplary voltage generator is ~hown in
E'IG. 5 wherein a convorltional pulse generator 98, ~or
example, a single shot multivibrator with automatic
delayed reset upon triggerirlg by the gate generator 52
produces a l.Oms input pulse (the length of the burst which
is delivered to parallel circuit legs 100 and 102). The
latter feeds appropriate voltage, from junctions 103 and 105.
through a pair of diodes 104 and 106 respectively to an
output junction 108 such that the latter transmits the
higher voltage of either junction 103 or 105. During
application of the input pulse, a capacitor 110 feeds a
decaying voltage to the junction 105 w~lile a resistor ll2
-18
S~71~
provides a constant low voltage at junction 103 such that
the output junction 108 initially experiences the decaying
voltages from the junction 105 until the latter equals that
of the junction 103 thereby producing shaped voltage pulse
59. Specifically, the voltage produced by generator 53
jumps from zero to six-~olts at the start of the yate
pulse and then decreases, substantially linearly, to about
four-volts in 0.5 msec, thereafter remaining substantially
at four-vo~ts for another 0.5 msec and then decreasing to
zero. The voltage values and variations are those actually
used, but those skil~ed in the art will recognize that
both the values and variations can be selected within a
wide range consistent with the. other components of the
modulator.
The shaped voltage pulse 59 (see FIGo 4)
applied to voltage control oscillato:~: 54 causes the latter
to develop a matching requency-modulated burst. r~he
fre~uency of the burs~ varying substantially linearly from
abou~ 65kEIz to about 50kHz duriny the time the output
2V voltage (pulse 59) of generator S3 decreases from s.ix to
~our volts, and then remains substantially constant at about
50kHz during the time the output voltage remains at about
four volts. After amplification by amplifier 55, the
~requency-modulated burst is applied to the primary 60
of transformer 56 whose secondary 61 is applied to trans-
du~er element 27A ~hrough decoupling diodes 57. The outpu-t
voltage of secondary 61 is made as high\as possible for the
transducer 27A, for example, about 300 volts peak-to peak
with a diode-capacitor combination 62 thereby establishes
a bias voltage on element 27A of about 150Vdc after
-19-
7~
~evoral cycle~. ~h~ outpu-t voltaye drives e~ement 27~
causing it to radiate a highly directional, correspondingly
frequency-modulated bursk of ultrasonic energy as indicated
by arrow 63 in FIG. 3~
The values of driving and bias voltages on
element 27A are based on using a 6 ~m Mylar film in the
element. m us, these values would be optimiæed for films
of different thicknesses such that the output of the element
and its sensitivity to echoes are simultaneously maximized.
In addition, the Q of the output circuit, which depends
in part on the c~pacitance of the transducer, should be
relatively low in order for the chirp to be transmitted
with a constant amplitude and with no significant dependence
on the capacitance of the ~ransducer. By maintaining a
relatively low Q, the energy o~ the system dies rapidly
at the termination of the drive voltage thus allowing the
transducer element to quickly reach a quiesence condition
at which it can receive echoes from relatively clo~e
su~j~cts.
Decoupling diocle~ 57 furlctiorl to decouple trans-
former secondary 61 from the transducex element during
reception of an echo. During transmission the voltage
drop of about 0.7 volts across the diodes is so small with
respect to the 300 volt peak-to-peak driving voltage, that
the decoupling diodes have no effect on transmission.
During reception, however echo signals produced by element
27A are in the range of 2 ~V to 20 mV and the diodes are
in effect an open circuit to the echo signals.
An echo from a subject is symbolically indicated
at 64 in FIG. 3, and the resulkant echo signal produced
~20-
1~ 8
by element 27~ is proce~sed b~ receivex 33A whlch includes
pre-amplifier 65, filter 51 referred -to previously, means
66 for varying the Q of the filter during the receiver
ranging time, and detector means 67 for converting an echo
signal to range pulse 17. During the receiver ranging
time~ the dc voltage on element 27A remains substantially
at 150Vdc. The input impedance of pre-~nplifier 65 is
matched to the transducer element impedance (about 12k ~ ).
The output ,mpedance of the pre-amplifier is selected to
be compatible with the highest Q of filter 51J which is
about 70. The gain of the pre-amplifier is about 48dB.
Filter 51 is an LC filter comprising secondary
61 of transformer 51 which ~urnishes the inductance of
the fi7ter, and capacitors 6R, 69 between which the output
of pre~ampli~ier 65 is connected. A tap 70 relatively
close to the ground connection of the ~econdary applies
~he output of the filter to input 71A of high-impedance
oukput amplifier 71 through resistor 72 which has a value
of about lK~ , Re~istor 72~ which is in parall~l w:ith
~he LC circuit of t,he filter, is a part of pre-progrc~mmed
Q control means 66 for varying the Q of the filter. Q
control means 66 also includes current yenerator 73 and
dynamically variable resistance means 74 serially con-
nected to resistor 72 and input 71A. Resistance means
74 comprises fixed resistor 75 (of about lM~ ) in
parallel with diode 76 arranged to conduct during the time
that the current generator 73 supplies current.
To provide the best signal-to-noise ratio for
echo signals from subject~ most remote from the camera
~i.e., about 7 meters), the difference between the center
~2~5i7~3
requency of the rilter and the re~uancy a~ w~ch t~e
response power drops to hal~ f, is related to the con-
stant frequency pulse length as follows: 0.2/(constant
frequency pulse length). In the preferrea embodiment,
the half-power bandwidth of the filter should be about
0.8kHz neax the end o~ the receiver ranging time when echo
signals from remote subjects are being processed. During
the initial portion of the receiver ranging time, when echo
signals from subjects close to the camera are being pro- -
I0 cessed, the bandwidth of the filter must be such as to
pass all of the frequencies of the chirp. Thus, the filter
must initially have a hal~-power bandwidth of about 30kHz~
the center frequenay being about SOkHz. The required
changs in bandwidth is achleved by chaing the Q of the
filter from about 5 near the beginning of the ranging time
to about 70 near the end. For a typical LC circuit (with
a c~pacitance of about 300pf), the resistance in parallel
with the circuit must vary from about lKl~Lto abou~ lM~ a.
The r~æiskance in parallel with the LC c:Lrcuit
o ~ilter 51 is the e~eeative resistanca o~ resistor 72
in ~eries with the parallel combination of rasiskor 75
and diode 76. The dynamic resistance of diode 76, for
small ac signals supplied by pre-amplifier 65, is ahout
inversely proportional to the dc current flow through the
diode, and will be substantially independent of the exact
diode characterlsti~s. When the dc current flow is rela-
tively high, the dynamic resistance of the diode will be
significantly smaller than the resistance of resistors
72 and 75~ with the result that the effective resistance
of ilter 51 will aepend substantially only on resistor 72.
-~2-
;3S'78
Conse~uently~ the Q o~ the filter 51, under the condition
of high current flow through di.ode 76~ wi~l depend on the
value of resistor 72 whi.ch is selected, taking the
inductance and capacitance of the filter 51 into account,
to provide a Q which establishes the filter bandwidth
su~h that all of the frequencies of the chirp are pa~sed
by the filter.
When the current through diode 76 is relatively
low, the dynamic resistance of the aiode will be of the
same order of magnitude as the resistance of resistor 75
with the result that the effective resistance of filter
51 will depend essentially on the resistor 75. Consequently,
the Q of the ~ilter under the condition of low current
flow throug7n diode 76 will depend on the value of resistor
lS 75 which is selected to establish a bandwidth matched as
well as possible to the fixed frequency of the burst.
The time-wise variation in the current supplied
to diode 76 ~y current generator 73 is such as to insure a
suitable variation in the effective value of the resista:nca
in parallel with the LC circuit of ilter 51. To this end,
generator 73 is responsive to gating pulse 58 for producing
a dc current that i8 initially high for a relatively short
time at the beginning of the receiver ranging time, and
monotonically decreases as indicated by curve 77 in FIG. 4(e).
Curve 77 is of a type that can be produced by rela-tively
simple components with more assurance of repeatability and
stability as compared to a monotonically increasing curve.
A suitable current generator may be provided by a networX
of RC circuits, each having different time constants the.reby
providing diminishing current. Curve 77 has three portions:
-23-
a transitional portion 77A lasting about 4 m~ec duriny
which the current rapidly drops to a substantially constan-t
level, an initial portion 77B lasting about 6 msec during
which the current remains substantially constant7 and a
terminal portion 77C during which the current decreases
substantially linearly. During portion 77A, the effective
resistance in parallel with the LC circuit, increases,
but the Q of the filter is primarily dependent on the
value of resistor 72 and changes only slightly as indicated
in FIG. 4(). During portion 77B~ the Q of the filter
remains substantially constant. Thus, for objects within
about 1.5m from the transducer element, the bandwidth and
admittance of the filter are substantially constant. For
objects beyond 1.5m, the bandwidth gradually decreases as
the filter Q increases, and the filker aamittance docreases
see FIG. 4tg). The change in filter admittance can be
thought of as changes in effective gain o~ the receiver/
t~e effectiv~ gain being relatively low and substantially
constant for objectæ up to about 1.5m ~rom the transducer
element, and increa9în~ ~or subject Aistances beyond t~:is
point, which is in accordance with the principles set
forth above.
Hence, the bandwidth remains essentially
constant, at a relatively broad value for approximately
9 ms or approximately one-fifth to one-quarter of the
predeterminea range time of 42 ms and then gradually is
narrowed during the remainder of the range time. Stated
otherwise~ the bandwldth and Q etc. remain constant during
the initial portion of the range time when echos may be
received from close up ~ubjects (up to 1.5 meters from
-2~
5'~
the c~nera) so as to ensure reception of all chirp fre~uencies
and produce fast filter rise. Then, the bandwidth is na~rowed
during the remainder of the range time to increa~e the
signal to-noise ratio.
The output of filter 51 i5 applied to input 71A
of amp~ifier 71 which has a high input impedance to prevent
loading the filter and lowering its Q. The back-to-back
diodes of Q control means 76 and the location of tap 70,
limit the transmit pulse~ but the amplifier 71 is still
overdriven to some extent. It is designed to recover
rapidly,, however, and can handle a filtered echo signal
after about 0.3 msec ~ollowing the transmit pulse. The
output impedance of this ampli~ier is low; and the gain is
about 65dB.
The output of amplifier 71 is applied to detector
net~ork 67 which includes a conventional clc~mplng circuit
80, an RC integrator 81 and a detector ~2. circult 80 is
driv~n by blanking generator 83 which produce.s a blan~iny
pulse 8~ in respon~e to gate pulsa 58 as shown in FIG. 4(h).
Pulse 84 lasts about 1.5 msec; and during this pulse,
clamping circuit 80 is effective to clamp the detector
input to ground.
Following the blanking pulse, an echo signal
pa~sed by filter 51 and amplified by amplifier 71 is xecti
fied and correlated. The integrator 81 is constructed such
that several cycles of an echo signal must be applied to the
integrator within a given time span (say n . 2 msec) for the
voltage to build up on the capacitor so as to reach the
thre~hold of amplifier 84 which thereby forms a range pulse
17. While the correlation technique utilized does not
)57~
improve the signal-to-noise ratio of the input signal,
it does provide a ~i~ter for discriminating against single
spikes such as may be present due to logic circuits
associated with the mechanism for moving the lens mount
of the camera.
As noted, the varied Q filter provides increasing
gain during the transmit-receive time . However, the gain of
the amplifier 71 may also be varied during the transmit-
xeceive time by means of a r~mp generator 96 which in response
to the gating pulse 58 of gate generator 52 applies con-
tinuously increased gain control to amplifier 71 during
the predetermined transmit-receive time of approximately
42 msec thereby providing substantially continuously
increasing ampliier gain as tha interval progresses.
With only slight modification, the circuit des-
cribed above is capable of operatiorl in several different
modes. As noted i.n the pre~erred mode~ a single burst~
namely a single one msec duration transmit pulse, is
utilized. l'his arrangement without amplifier ramp gain
permits detection o~ ob~ects at a distance of from 25cm
to about 5m, the total time needed to measure the distance
baing 19ss than about 35 msec; the latter being extended
to about 9m when amplifier ramp gain is employed. This
mode is preferred with a snap~shot type camera since
focusing can be effected and exposure completed with a
single manual input.
In an alternate embodiment, the chirp portion
of the burst may be digitally formed in a st~pped arrange-
ment as shown in FIG. 6 wherein in response to the 42 msec
gate pulse 58 from gate generator 52, a s~epped frequellcy
-26-
s~
pulse 120 i9 provided by msans o~ a alock 12Z and programmed
divider 124. ~s shown in FIG. 7, the pulse 120 falls in
a series of small steps from 65kHz to 50k~Z.
In another mode of operation, several different
pulses could be used, one for each of several di~ferent
ranges. For exampleJ a short pulse could ba used for
objects from lOcm to lm. A second and longer pulse could
be used for longer distances. The Q of the filter would
have to be adjusted to the pulse Length, however; and the
maximum Q would be different for different pulses. Since
the signal-to-noise ratio is proportional to the square
root o~ the pulse length, a change in pulse length would
permit the range to be increased. If a system has a 5m
range with a 0.5 msec pulse, a system with a maximum pulse
length of 5 msec could provide range information for objects
up to 6.5m. The Q of the filter, however, would preferably
be ten times higher.
The present invention is also capable of baing
used in a conkinuous pulsing mode suitable or use with a
movie camera enabling the ocu~ to be adjusted continuou ly
during the time the camera i9 running. Using a ~low drive
for moving the lens mount, an integration of the echoes
would be achieved improving the signal-to-noise ratio.
The variation in filter ~ in the preferred
25 embodiment has many advantages when used in a ranging system
for a camera. In most cases, the need for a separate ramp
gain in the general form of the invention shown in FIG. 1
may be eliminated. SecondlyJ the need for a matched filter,
which is required for reception of a chirp is eliminated,
thereby simplifying the circuitry yet permitting the use
-27-
5'78
o a chirp which reduces sensitivity to intere~rence fo~
~lo~e objects wi~hout significantly comprising the signal-
to-noise ratio f~r remote objects. Thirdly, the filter Q
is low for near~by objects where absolute accuxacy is most
needed, and a low Q filter has a fast rise time insuring
good accuracy. Finally, an electronic drif-t in output
fre~uency~ as compared to the filter frequency, or the
frequency shift produced by the movement of the target
(Doppler effect), affects only the far distance~ where the
consequences are less important for a photographic camera
The larger the drift or frequency shift, the nearer will
be the affected distance. This gradual influence has to be
compared with the total disappearance of the echo at a
certain drift at all distances for a constant ~ system.
The selections of frequencies, pulse durations
and the like ~re by way of example only, it being under-
stood that these parc~meters are chosen in accordance with
the use to which the ranging system o~ the present invention
is applied. For example, the requency o e 50kHz was
selected on the basis o~ the largest distance to be
measured reliably in the presence of acoustic and the~nal
noise. Other considerations are involved in frequency
selection, such as transducer size, acceptance angle, etc.
Finally, the disclosed technique for varying the
effective resistance of the ~ilter is by way of example
only~ The resistanca can be in series with the LC circuit
instead of parallel; and this expedient would require dif-
ferent values of resistors. Alterna-tive to the use of a
current-controlled diode, an FET -transistor operating in
the depletion mode could be used~ preferably with the
transistor used in a parallel rnode.
-2~-
5 78
Ranging systems 11 and 11~ ha~e appliaa~ility
more general than incorporation into cameras. For example,
they are applicable to arrangements for assisting blind
persons, or assisting in maintaining proper spacing between
vehlcles, or any operation requiring ranging inormati.on.
It is believed that the advantages and the
improved results furnished by the apparatus of the present
invention are apparent from the foregoing description of
the preferred embodiment thereof. Various changes and
: 10 modi~ications may be made without departing ~rom the spirit
and scope of the invention as sought to be defined in the
claims that follow.
: ~ :
:: :
-29-
