вход по аккаунту


Патент USA US2118069

код для вставки
May 24, 1938.
A. F. coNNl-:RY
Filed July 25, 195e
2 sheets-sheer 1
May 24, 1938.
Filed July 25, 1956
2 Sheets-Sheet 2
È»mxä w
6105/? /v- Cown/¿wy
ß .
Patented, May 24, 1/938
Alder F. Connery, Brooklyn, N. Y., assignor to
The Commercial Cable Company, New York,
N. Y.,'a corporation of New York
Application July 25,\l936, Serial No. 92,551
y 5 Claims.
(Cl. ¿178-70)
The present invention relates to signal regener
ating apparatus for re-shaping and re-transmit
ting distorted telegraph signals. More particu
larly it relates to fork controlled regenerators of
5 the type disclosed in the copending application to
the winding is from left to right in the drawings
will cause the armature of the relay to move in
the same direction, that is, fromieft to right.
This same arrangement has been- followed with
respect to every winding of every polarized relay
Marion H. Woodward, iiled July 1, _1936, Ser. No. ' in both figures of drawings, so that in general
the same simple rule may be applied in studying
the operations of any of the polarized relays. A
It is an object of the present invention to pro
vide an improved generator circuit in which the Y current from left to right will move the armature
w various characteristics of the circuit are readily from left to right and vice versa..
It is also an object of the invention to provide
Connected to the contacts of receiving relay 2
are the operating windings |03, | 04 of the dot and
an improved fork correcting circuit of the type in
dash locking relays 3 and l', respectively. These
which corrections of a predetermined duration
i.; are applied to the fork when the phase of the fork
and of the signal diiîer from each other by more
than a predetermined amount. More particularly,
it is an object to provide such a fork correcting
circuit in which the above mentioned predeter
mined duration of the corrections is measured
by a reliable, accurate circuit which is readily ad
justable and whose time constant is independent
of the rapidity with which the corrections recur.
Another particular object of the invention is to
23 provide such a fork correcting circuit in which
the predetermined amount of phase difference
between the signal pulses and the fork contact
closures is also measured by a readily adjustable
circuit having the desirable characteristics of re
30 liability, accuracy and constancy of time.
The invention may be best understood by ref
erence- to the drawings, in which Fig. 1 and Fig.
2, when placed together, form a schematic repre
sentation of one preferred forni of regenerator in
35 accordance with the present invention. In the
drawings, | is an incoming cable over which sig
nals are received from a distant transmitter (not
shown). It will be assumed that the signals are
in the so-called “Submarine Morse" code in which
40 the dots and dashes consist of positive and nega
tive pulses oi' equal length and in which zero
or no current intervals of this same length are
used for spaces between letters and words. It
will also be assumed that the signals are trans
45 mitted at `a substantially constant rate as from a
tape transmitter.
Relay 2 is a receiving relay of the moving coil
type which normally stands with its contact lever
midway between its two contacts and is polarized
,3o so that it will be actuated to the rig'ht by a signal
pulse of positive current (a dot) and to the left
by a-signal pulse of negative current (a dash).
For convenience in reading the drawings, the
winding of this relay 2 has been arranged so that
55 aícurrent whose conventional direction through
relays 3 and 4 are also provided with biasing
windings 203, 204, respectively, through which
current is applied to bias the armatures toward
the left, as shown in the drawings. Each of these
relays is also provided with two locking windings
303, 403, 304 and 4104, respectively. These lock
ing windings are so connected in series with the 20
contacts of the relays 3 and 4, that the relays
are locked in whichever- position they happen
to occupy, whenever current is ñowing through
the relay contacts. Connected to the contacts of
the dot and dash locking relays, respectively, (and 25
in series with the above mentioned locking wind
ings) are the operating windings |05, 205 and
|06, 206 of dot and dash transmitting relays 5
and 5. These transmitting relays 5 and 6 have
no predetermined bias in any particular direction 30
but are provided with senil-locking windings 305,
306 which are so connected to the contacts of the
relays themselves that the relays are partially
locked in whatever position they happen to be.
None of these semi-locking windings, however, 35
produces a strong enough locking bias to prevent
therelay from moving to a new position when
the appropriate operating winding is energized.
The contacts of transmitting relays 5 and 6 are
connected to a special ungrounded transmitting 40
battery '| and the armatures of the relays are
connected to the outgoing cable 8 and to the out
going ground connection, as shown in Fig. 2.
The above described chain of apparatus per
forms the function of retransmitting to outgoing
cable 8 signals whose character is determined by>
the signals in incoming cable I. Such a simple
repeating equipment, however, is not capable by
itself of performing any useful regeneration of
the signals. The timing impulses necessary for 50
regeneration are provided by a separate portion
of the equipment known as the “timer”, shown
within the dot and dash lines in Fig. 1. This
“timer” is connected to the dot and dash locking
and transmitting relays 3, 4, 5 and 6 over wire 55
9, in such a Way that the operating windings IUS,
2‘05, |06 and 206 of the transmitting relays can
not be energized excepting at the moments when
negative potential is connected to wire 9 by the
with the above described pulses from relay I0, are
not derived directlyfrom the fork contacts but
from a fork pulse relay I5 which is operated un- ,
der the control of the fork through a special delay
chain ` I2, I3, I4.
'I'he timer itself has for its principal element
a self-driven tuning fork II which is arranged
to periodically connect wire 9 to a source of nega
tive potential. The normal speed at which this
10 fork operates is primarily determined by the nat
ural resonance period of the fork itself, but is
also capable of being slightly varied by means of
fork speed rheostat III, which is connected in
the self-energization circuit of the fork. By
15 means of this rheostat the normal running speed
This chain determines how
much delay will be provided between the closure
of the fork contacts and the operation of the fork
pulse relay and also independently determines
the duration of the operation of this relay. Both
of these time intervals can be readily adjusted 10
and will remain exceedingly constantv at the value
to which they are set.
In this special timing chain, I2 is a vacuum tube
whose cathode is connected to a point of zero or
neutral potential (designated -_I-) in the drawings.
20 between the fork and the signal, corrector pulse
The screen grid of this tube is connected through
a suitable resistor to positive potential. The con
trol grid 4l2 of this tube is connected through
condenser II2 to a source of positive potential
this condenser being shunted by an adjustable 20
windings III) and 2I0 are connected to the same
contacts of receiving relay 2 which are used for
either an inner or an outer contact of fork II. 25
of the fork is adjusted to be slightly faster than
the average center-hole speed of the incoming
In order to maintain the required synchronism
relay IIJ is provided. This relay is equipped with high resistance rheostat 2 I2, hereinafter referred
two operating windings III) and ZIO, as Wellas _ to as the “fork-pulse delay rheostat.” Grid M2
one biasing winding 3III. The two' operating is also connected through a suitable low resistance
3I2 to a switch 2I I, which may be connected to
controlling the re-transmitting equipment, pre
We will assume that the switch is connected as '
viously described.
shown in the drawings to an outer contact of the
These windings are so con
nected that the energization of either one of them
will cause the armature of relay I0 to move to the
A30 right, the biasing winding being arranged to
move this armature to the left.- The relative
strength of the biasing and operating windings
are so chosen that the relay will be rapid not only
in operation, but also in its return to normal.
35 Therefore, no` matter how rapidly the armature
of relay 2 may travel in moving from its left hand
contact to its right hand contact or vice versa,
relay I0 will not hold over and remain energized,
but will release at the moment when the arma
40 ture of relay 2 leaves one contact and will again
re-operate when the armature of relay 2 arrives
at the other contact. Thus, it will be seen that
relay IIJ will operate or move from its left' to its
right contact at the beginning of each dot or
dash transmitted over the incoming cable
to receiving relay 2, excepting in the case of
a dot which immediately follows another dot,
fork, so that negative potential is normally ap
plied to grid M2 except when the fork is at
tracted toward its inner contacts. Thus, in the
condition assumed, and shown in the drawings, '
the potential of grid 4I2 will be highly negative
with respect to the cathode of tube I2, and con
denser II2 will be charged. Anode SI2 of this
tube is connected to a source of positive potential 35
through the winding of relay I3.
Under the control of the contacts of this relay
I3 is vacuum tube I4 whose cathode is connected
to neutral potential (i) and whose screen grid
is connected through a suitable resistor to pos
itive potential. Control grid 2M of this tube is
connected to a source of negative potential
through a variable high resistance rheostat Ill,
hereinafter referred to as the “fork-pulse dura
tion rheostat.” When relay I3 is operated, how -45
ever, this grid 2I4 becomes also connected to pos
itive potential through a two-microfarad con
or a dash which immediately follows another ' denser II3. Thus, at the moment of operation
of relay I3, control grid 2H becomes strongly
of two dots in succession or two clashes in suc
positive through the
H3 and then slowly
negative potential as
through rheostat H4.
cession is relatively less frequent than other com
binations, it can be seen that relay ill will be op
erated by a comparatively large proportion of the
signal pulses, thus providing a suñicient number
I I3 whenever relay I3 returns to its back contact.
Anode 3M of this tube is connected through the
Since the incoming signal consists predomi
nantly of dots and dashes, with only a small per
centage of zero impulses, and since the occurrence
discharge of condenser 50
returns to its normal
this condenser charges
A ‘suitable discharge re
sistor 2I3 is connected to the back contact of re
lay I3 in such a way as to discharge condenser 55
of opportunities for comparing the signal timing4 winding of fork-pulse relay I5 to positive poten
tial, the connection being arranged to pass also
with the fork timing.
'I'he armature of relay I0 is connected through through relay I3. The above described chain of 60
a two microfarad condenser 4I0 to a source of apparatus II, I2, I3, Il and I5, together with
positive potential and the left hand contact of the associated resistors and condensers, serves to
this relay is arranged to -discharge the condenser cause the operation of fork-pulse relay I5 in re
through a suitable resistor 5I0. Thus, upon each
operation of relay I0 a brief pulse of positive po
tential Ais transmitted out over the right hand
contact of the relay. This impulse does not con
tinue as long as the relay continues in the op
erated state but only occurs for a very short time,
70 each time the relay actually operates. These
pulses are used for comparison with a diiîerent
set of pulses produced by the fork to determine
whether or not the timing of the fork and signal
differ by more than the tolerated amount.
The fork pulses which are used for comparison
sponse to the vibrations of fork II. The delay
interval between the opening of the outer contact 65
of fork II and the operation of relay I5 is adjust
able-by means of “fork-pulse delay rheostat” 2 I2.
The duration of the‘operation of relay I5 is sepa
rately and independently adjustable by means of
“fork-pulse duration rheostat” IM.
The* back contact of relay I5 is connected
through a protective resistance II5 to negative
potential so that normally, upon the operation
of corrector pulse relay I0, condenser M0 will
be discharged through this resistance H5. The
2,1 18,009
front contact of relay I6 is connected- to the left
hand winding 'oi correction pick-up relay I8.
Pick~up relay I8 is provided with only one con
‘ tact in order to make it quick acting. Thel arma
ture of -this relay I 8 is connected to a source o!
positive potential and> its single front contact is
connected to its own locking winding I I 6, through
the winding of auxiliary correction relay I'I and
wire 20 which extends to the bias correcting re
10 lay, later to be described. Auxiliary correction
relay I1 is provided with two contact levers II‘I
and 2I1. Lever 2I‘I is not absolutely essential to
the operation of the device but is merely pro
vided to insure reliability. This lever cooperates
15 with its back contact to open the connection of
discharge resistor 2Il described above, for the-
purpose of preventing an'absolute tie-up in case
the front and back contacts oi' relay I8 become
short-circuited by thin lilaments oi' welded metal.
20 Ordinarily a welded short circuit of this'type is
only temporary and will disappear after a few
openings and closures of the effected contacts.
Because of the nature of the present circuit, how
ever, it would be possible for such a short c_ir
cuit at the pòint mentioned to tie'up the system
so that no further contact motions would occur;
thus, the diillculty would not clear itself. By the
provision of lever 2I'I and its back contact, such
a tie up is prevented. Lever II‘I is connected to
30 negative potential and the front contact asso
ciated with this lever is connected through an
relationship of these windings I2! and 228, how
ever, the bias correcting relay operates in a some
what diiïerent manner from the corrector pulse
relay‘III-belng operated to the left in response
to a dot signal and to the right in response to a
dash signal. Since relay 25 has no bias winding,
and since it is of the two-position type whose
armature is incapable of remaining midway be
tween its contacts, the relay will, during no
current or zero signal pulses, remain in the posi
tion to which it has last been moved. The arma
ture of this relay 25 is connected to wire 28
which, as shown in Fig. 1, receives a pulse o!
positive polarity each time the correction pick
up relay operates. 'I'he contacts _of this relay 2l
are connected through its respective locking wind
ings 325 and 425 to _two terminals of the auto
matic bias corrector schematically illustrated be
low the relay. 'I'his automatic bias corrector is
of the type fully described and shown in Patent
No. 1,929,879 to A. F. Connery.
The function of this automatic bias corrector
is to compensate for unsymmetrical conditions of
the line circuit or of the transmitting or receiving
This- is performed by biasing the
receiving relay 2 (by means of an auxiliary wind
ing or its normal receiving winding). The auto
matic bias corrector acts to change the amount
' equipment.
of bias so applied, by small steps between a maxi
mum value of “dash” bias and a maximum value
of “dot” bias. In response to a positive pulse upon
its left hand terminal (which is produced in re
sponse to the occurrence of a correction during
a dot signal as hereinafter explained) the auto
the closureof lever II1 against its’front contact . matic bias corrector moves one step in the direc
will .provide an uninterrupted energization of the tion of increased dot bias (or decrease dash bias).
fork magnet for the duration of the closure. The In response to a positive pulse on its right hand
characteristics of the fork are such that this terminal (which is produced in response - to a
energization will not only decrease the amplitude correction duringv a dash signal as hereinafter
of the fork vibrations but will also slightly slow explained) the automatic bias corrector movei
one step in the direction of increased dash bias
up the vibrations of the fork. Rheostat 3II en
ables the intensity of this slowing action to be (decreased dot bias). Thus, if the corrections
regulated and is hereafter referred to as the of the fork at a given time are occurring primarily
“fork-slowing intensity rheostat.” ’I'he back con
on dot signal pulses, the bias applied to the re
ceiving relay will be cumulatively altered in the
tact of lever I I1 is used to control a third vacuum
45 tube I8 whose function is to determine the dura
dot direction until the symmetry of the signals i:
so restored that corrections in general take place
tion of the corrections applied to the fork.
'I'he cathode of tube I8 is connected to neutral equally on dots anddashes.
The above detailed description of the appara
potential (i) while its screen grid is connected
through a suitable resistor to positive potential. tus and its interconnections not only provides ar
The control grid II8 of this tube is connected to adequate disclosure of how to construct the pre
a source of positive potential through a condenser ferred embodiment of the present invention, bui
2 I8 which is shunted by a variable high resistance almost sufñces to make its operation obvious. Foi
rheostat 3I8 (referred to as the “fork-slowing completeness, however, a brief description ol' th(
duration rheostat”). -This grid is also connected operation will be given.
vIn the timing equipment the self-driving tunI
through a suitable low resistance ‘I8 to the backv
contact of lever II'I of the auxiliary correction ing fork II has been adjusted so that it is vibratI
adjustable rheostat 3II to the driving magnet of
fork Il. Since this connection does not pass
through the outside contact of the fork itself,
relay I'I. Normally, therefore, the potential on
grid I I8 is highly negative relatively to its cathode.
Anode 5I8 of this tube is connected to positive
potential through the winding of correction ter
minating relay I9. This relay I9 is lprovided with
a singlefront contact which is connected so that
upon the operation of relay l5, it short circuits
65 the locking or right hand winding ‘of pick-up
relay I6.
ing at a normal speed slightly higher than th(
average center-hole speed of the incoming signal
Simultaneously, the corrector pulse relay Il i
operating sporadically under thecontrol of thl
dots and dashes of the signal as received by rela:
2. For deiiniteness in my description, it will b
arbitrarily assumed that the desired relation be
tween the fork vibrations and the operations> o
relay IIJ is as follows: the fork just commences tÁ
In addition to the re-transmitting equipment open its left o_uter contact at the instant whicl
and the timing equipment described above, there y corresponds to the center of anincoming signa
.is provided some further apparatus shown on the pulse as received by relay 2. It will be also arbi
drawings at the left of Fig. 2 and marked “Bias trarily assumed that at the present time this de
equipment." Operating windings I25 and 225 sired timing relation is exactly fulfilled. .
The instant when fork II opens its left han(
of the bias correcting relay 25 are connected in
series with windings IIII and 2I0 of relay I0 so outer contact will be employed as the refereno
that this relay 25 is operated from the contacts point from which to compute time for fork II a
of the receiving relay 2. Because of the reversed well as for the chain of apparatus I2, I3, I4, I
2,118,069 n
directly controlled by this fork I I. The time from
this point will be measured in degrees on the basis
that one complete cycle of the fork equals 360’.'
Thus the opening of the left hand outer contact
of the fork occurs at fork time 0°.
At the instant of the opening of the above not
the operations of fork pulse relay IB. In other
words relay I0 will deliver its brief surge of cur
ed contact (that is, at zero degrees) the- charge
Assume now that the fork gradually Igains in
phase with respect to the incoming signal (be
cause of its slightly higher frequency). Finally,
on condenser II2 commences to leak oiI through
rheostat 2I2. Thus the potential of grid ‘I2 rises
exponentially at a rate determined by the “fork
pulse delay rheostat” 2I2; After a certain in
terval, the potential of grid 4I2 is suillciently high
so that the space current of tube I2 operates relay
I3. ’I'his relay will be assumed to close its make
contact at K’ degrees of fork time. At the in
stant when the make contact of relay I3 closes,
grid 2H of tube I4 becomes positive through con
.denser II3. As a result of this positive potential
on grid 2 I4, the space current of tube I4 becomes
more than suillciently great to operate the, fork
pulse relay I5.
rent before fork time »K°. In this case, the con
denser 4I0 will discharge through resistance III
and therefore pick-up relay Il 'will not be ener
the fork will get so far in advance of its proper
timing relative to the signal that fork pulse relay
I5 will'operate before the corrector pulse relay Il.
In this case, upon the operation of relay I0 pick- ’
up relay I6 will be operated, locking itself and
energizing auxiliary correction relay I'I in an 15
obvious manner. The operation of auxiliary cor
rection relay I1 will apply to the drive magnet of
the fork an uninterrupted energization superposed
on the regular self-interrupted driving energiza
tion. Because of the characteristics of the fork, 20
This relay I5 therefore closes , this uninterrupted energizatlon will slightly slow
its make contact at a time only slightly later
than K'°; we will call the moment of this contact
closure “fork-time K°”. The positive potential
on grid 2H, however, exponentially decreases
commencing from the instant when it is applied
(from fork time K'°) because of the charging of
condenser II3 through rheostat Ill. After an
interval, therefore, the potential of grid 2li be
comes sumciently negative to cause the release of
down the rate of vibration as long as it continues
to be applied. 'I'he operation of auxiliary correc
tion relay I‘I will also permit condenser 2|! to
discharge through rheostat 3 I8 so that the poten 25
tial of grid II8 will exponentially rise. After an
interval of time determined by the adjustment of
rheostat 3I8, the potential of grid IIB will be
suil‘lciently positive so that the space current of
tube Il will operate relay I9. Finally, the opera 30
tion of this correction terminating relay I9 willv
short circuit the locking winding of pick-up relay
relay I5. 'I'he time at which this relay in releas
ing leaves its make contact will be called fork
time C°. This time is accurately adjustable byl IB thus permitting the latter to release. 'I‘his re
meansof the “fork-pulse duration rheostat" II4. lease of relay I8 causes the release of relay I1,
From the above description it will be seen that the termination of the fork-slowing current, and
the armature of relay I5 remains closed against
the restoration of grid II8 to its normal poten
its make contact during the interval from K° to
tial. The restoration of this grid to normal re
,C°. ` Both the duration of this interval and the
leases relay IS so that the correction equipment
-time of its commencement are accurately and
is all in condition for another operation.
independently adjustable.
'I‘hese time intervals ._
are determined by the charging or discharging
rates of condensers through adjustable resist
'I‘he inherent accuracy of such an ar
rangement is enhanced by the fact that the ex
.ponential charging and discharging curves are
not used in the vicinity of their asymptotic end
regions (since the negative and positive poten
tials available for the control grids of tubes I2
>and I4 are designed to be far in excessief the val
ues acquired for complete cut-off and for relay
operation- respectively).
Independently ofA the above traced train of
events, corrector pulse relay I0 is simultaneously
operated in response to the incoming signalas
' received by relay 2. As explained previously, this
relay I0 does not operate in response to every
pulse but the sporadic operations of this relay
In the above description of the operation of the
timing equipment there has been described how
an unusually retarded signal cooperates with a
fork which has slightly advanced in phase to
cause a correction which slows down the fork for
a predetermined duration of time. The intensity
oi the slowing or braking current is also prede
termined, so that the net result of the correction
is to setback the phaseof the fork by a deilnite
amount, generally of the order of 15° or 20°. If,
in some manner, the fork is so far ahead in phase
that it requires more than this amount of correc
-tion, a second correction will immediately be ap
plied to it at the next operation of corrector pulse
relay I0.
ordinarily with this system the signals which> 55
cause corrections to take place are those signal*
elements whose eil’ective starting points are un
usually retarded with respect to the general aver
when they do occur- are precisely timed with re
spect to the vibrations of fork Il_except for two age center-hole timing of the incoming message.
'I‘he unusually retarded signals areîln general
(l) The irregularity of individual signal pulses produced by the transmission of a dot imme
with respect tothe average center-hole speed of - diately following one or more dash signals or by
the transmission of a dash signal immediately
the signal.
following a dot signal. Because of the charac
’ . (2) The deviation of the fork from its normal
time relationship with thecenter-hole speed of terlstics'of the transmission cable I, such signals
(which require a complete reversal from positive
the signals.
to negative potential or vice versa) are trans
So long as the sum of these factors _does not ex
ceed a predetermined tolerated amount, correc à mitted more slowly than the average signal.
In the absence of any particular bias favoring
tion pick-up relay II will not be actuated. 'I'his
tolerated amount of phase difference between dot or dash reception, these unusually retarded
fork II and relay l0 is determined by the length - signals will consist almost equally of dots and
of the time interval which has been called K°. dashes. IWhen, however, the receiving relay or
Normally, when the desired relationship between transmission line or transmitting equipment pre
sents some sort of dissymmetry, it may occur that
the signal and fork exists, the operations ot cor
rection pulse relay I0 will occur shortly before dots. for example, are more readily responded to
than' dashes. In this case, the unusually retarded
signals which cause the operation of the correc
tion pick-up relay will predominantly, if not
wholly, consist of dashes.
Upon each operation of correction pick-'up re
lay I6, a positive pulse is sent out over wire 20
through one or the other of the contacts of relay
` 25 to the automatic bias corrector. Relay 25'is, of
course, operated to its left hand position by dots
and toits right hand position by dashes. There
fore, in the above assumed case in which dots were
more favorably received so that the corrections
took place predominantly at the time of reception
of dashes, the pulses received by the automatic
15 bias corrector will be predominantly on its right
hand terminal. This will cause the compensat
ing bias applied to receiving relay 2 to be cumu
latively altered in such a direction as to be more
favorable for the reception of dashes. Similarly,
20 if the predominating signals are dashes (so that
the late signals which cause correction are mostly
dots) the automatic bias corrector will be actu
ated mostly by signals on its left-hand terminal
and will therefore cumulatively vary the bias in
25 the direction which is most favorably for dots.
The operation of the re-transmitting equipment
itself is very simple. In accordance with the
signal received by relay 2 dot or dash locking
relay 3 or l is operated. Then, at about the center
30 of the signal pulse, a timing impulse is received
on wire 9 from the tuning fork II. This4 impulse
on wire 9 normally locks the dot or dash locking
relays in whatever positions they then occupy,
and energizes the appropriate windings _of re
35 lays 5 and 6 to cause these to take up positions
corresponding to relays 3 and 4 respectively.
When the brief pulse on wire 9 terminates, relays
3 and 4 are again free to move in accordance with
the incoming signals but relays 5 and 6 will now
40 remain in their new positions until the arrival
of the next pulse on wire 9. Thus, it will be seen
that this regenerator equipment determines the
character of the pulses transmitted (whether
to give the desired relation of the timing pulse on
wire 9 with respect to the signals received by re
lay 2. By the inherent operation' of the fork
correction apparatus the timing relation between
fork Il and the incoming signals will be main
tained such that normally fork pulse relay I5
will operate slightly later than corrector relay
pulse I0. ,Therefore„ since the time of opera
tion of fork pulse relay I5 is K° after the open
'ing of the left center contact of the fork, any 10
variation ofthe time interval K° will ultimately
result in a changein the timing relation between ,
the vibrations of fork I I and the incoming signal.
The adjustment of “fork pulse duration rheo
stat" II4 controls the duration of the fork pulse,
i. e. the time during which relay I5 remains
operative. This time should be longer than the
ordinary variations between advanced signals and
retarded signals.- For example, if some of the
signals are transmitted 30"l earlier than the 20
average and other signals are transmitted 38°
later than the average, the duration of the fork
pulse should be at least 68°. On the other hand,
the duration of the fork pulse should not be so
long as to allow any possibility of extending over 25
into the next following signal pulse, even if this
signal pulse happens to be unusually early. For
this reason (on the above mentioned assumption
of 30° maximum leading distortion and 38° maxi
mum lagging distortion) the fork pulse must be 30
less than 292° in length. Also the total time C°
(which is the delay of the fork pulse plus the
duration of the fork pulse) should be less than
360°. With the above assumption of signal dis
tortion, and the previous arbitrary assumption 35
that 0° of the fork cycle occurs at the center of
the average signal pulse, this will mean that the
fork pulse should be less than 142° in length. It
will thus be seen that the adjustment of the
fork pulse duration rheostat is not at 'all critical 40
unless the distortions encountered in the signal
are unusually severe.
Although the above described embodiment of
45 of the incoming signal pulse. This determination
the invention comprises a fork Whose natural
speed is higher than the average center-hole 45
is effective at a time which is about the center of
the incoming pulse so as to be less affected by dis
tortion. Furthermore, it will be noted that the
speed of the incoming signal, and a correcting
system >for the fork which acts to delay its phase
by a predetermined amount, it is obvious that the
time at which the outgoing pulses start and end
is not determined from the incoming signal but
from the timing device. Thus, the signals trans
converse system can be used in which the fork
vibrations are normally slower than the average 60
center-hole rate of the signal. In such a sys
mitted out over the outgoing cable 8 are not only
re-shaped so as to have square tops but are also
tem the correcting signal must be arranged to
set forward the phase of the fork (for example,
by speeding up its vibrations for a certain length
plus, minus, or zero) by means of the character
re-timed so as to have a uniform, even space from
55 the commencement of each signal element to the
commencement of the next one. It will also be
noted that these signals are slightly shifted with
of time).
It will be noted in the above described embodi
ment the corrections occur when the interval
between the opening of the left outer fork contact
respect to the incoming signals, since each trans
mitted pulse commences at approximately the4 and the operation of relay I0 becomes greater
60 center point of the incoming signal pulse; this than the time interval determined by the delay 60
however is ordinarily of no consequence.
chain Il, I2, I3, I4, I5. It is obvious that with
For best operation, it has been fovund that the out changing the spirit of the invention, the sys
fork speed should be set approximately one cycle tem could be arranged so that the correctionsl
per second faster than the average center
taire- place when the interval between the fork
speed of the incoming signals. The “fork
vibration of the operation of relay I0 becomes 65
slowing duration rheostat” should be adjusted less than the interval determined by the delay
so that the correction will last several fork cycles
(say 5 or 10). Theñ, the “fork-slowing intensity
rheostat” should be adjusted so that the total
phase retardation caused by one correction of the
fork is of the order of 15° or 20°. This will mean,
of course, that approximately 18 to 24 correc
tions per second will take place during ordinary
'I'he “fork-pulse delay rheostat” is now adjusted
chain. In general, however, it is a feature of this
invention to correct from one side only, i. e. to
correct only when the interval between the vibra
tion of the fork and the operation of relay Il) 70
becomes unequal in one particular sense to theinterval determined by the delay chain.
Also, it should be noted that while the above
described embodiment of the invention employs
a delay circuit between the fork contact _and th‘e
fork pulse relay l5 in order to permit the phase.
of the fork relative to the signal to be varied, it
is possible to attain the same result by arranging
the delay chain I2,l I3, I4, i5 under the control of
s corrector pulse relay ill instead of under the con
trol'oi’ tuning fork li (or both chains together
might be used). Either of these arrangements is
somewhat superior to the use of a delay chain in
wire 8 .between the fork and the retransmitting
10 relays because the fork vibrations themselves are
employed for timing the regenerator without the
interposition of any apparatus.
What I claim is:
l. In a telegraphic signal regenerator a re
15 ceiving relay, a vibratory timing member, a
brator, a space `discharge tube coupled to said
contact device, a partially reactive circuit hav
ing an adjustable time constant, connections be
'tween said space discharge tube and said partial
ly reactive circuit whereby the current through
said tube is controlled by said partially reactive
circuit to determine an adjustable time interval
andcorrecting means for energizing said mag
net -in response to the condition'that the time.
interval between said ilrst device and said con 10
tact device is greater than-the time interval de
termined by said space discharge .tube and as
sociated circuit.
4. A telegraphic signal regenerator compris
ing a self-driven vibrator, a magnet for influ
transmitting relay jointly controlled bysaid re » encing the speed thereof, a first device operated
responsive to incoming signals, a second device
ceiving relay and said member, a speed iniluenc
j ¿ing magnet for varying the vibration ratepf said,>` operated in response to the motion of the vi
member, means for initiating corrections when
2_9 ¿the assynchronism between the receiving relay
and the vibratory member exceeds a certain value,
a condenser, a resistor, a space discharge device
vhaving a control grid, connections whereby the
. vpotential of said -condenser is varied by current
2_5 through said resistor in response to the initiation
brator, a space discharge tube, a partially reac
tive circuit having a predetermined time con
stant, connections whereby the current through
said tube is controlled by said partially reactive
circuit to> determine a definite time interval, cor
recting means for controlling the energization of
said magnet invresponse to an inequality of the 25
of acorrectlon' by said initiating means, other _ interval between the operation of said first and
~` Vconnections between said dischargeA device and
said condenser such that the space current of said
' discharge device is varied in response to the po
second devices, and the time interval deter
trolling the energization of said magnet under
`the control of said space current to vary the speed
of said member for a predetermined time..
>2. A telegraphic signal regenerator comprising
35 a self-driven vibrator, a magnet for iniluencing
the speed thereof, a ilrst means operated respon
stant, connections whereby the current through
mined by said space discharge tube,- a sec
ond space discharge tube, a second partially
30 tentlal of said condenser andv means for con- » reactive circuit having an adjustable time. con 30
Asive to incoming signals, a second means oper
ated in response to the motion oi' the vibrator,
a space discharge tube coupled to said second
40 means, a partially reactive circuit having an ad
justable time constant, connections between said
space discharge tube and said partially reactive
circuit whereby the current through said tube is
controlled by said partially reactive circuit to
45 determine a deñnite time interval and correcting
~ means for controlling the energization of said
magnet in response to inequality in one sense by
a predetermined amount of the time interval be
tween the operation of said first and second
50 means and the time interval determined by said
space discharge tube.
3. Atelegraphic signal regenerator comprising
a self-driven vibrator, a magnet for influencing
the speed thereof, a first device operated re
the said second tube is controlled by said circuit
to determine a second interval and holding
means'to maintain for the duration of said sec
ond time interval the energization of the magnet 35
which is caused by the correcting means.
5. A telegraph signal regenerator comprising
a self-driven vibrator, a magnet for inñuencing
the speed thereof, a first device operated in re
sponse to incoming signals, a second device oper 40
ated in response to the motion of the vibrator,
corrector means for controlling energization of
said magnet in response to a predetermined dif
ference in time between the operation of said
first and second device, means for determining 45
said predetermined time difference comprising a
space discharge tube coupled to said second de
vice and a partially reactive circuit having an
adjustable time constant coupled to said- tube,
and connections for energizing said corrector 50
means in response to an inequality between the
length o'f said time difference in açredetermined
sense and the length of time determined by said
adjustable time constant circuit.
55 sponsive to incoming signals, a contact device
operated in _response to the motion of the vi
Без категории
Размер файла
1 115 Кб
Пожаловаться на содержимое документа