The disclosure describes a printer in which characters are moved
past a sheet support on an endless belt. Hammers move from a rest
position to a strike position in order to strike the characters
against the sheet. The time at which the hammers are moved from
their rest positions so that the characters can be accurately struck
irrespective of belt speed variations by means of a strike correction
circuit arrangement which includes a first sensor means positioned
with respect to the endless belt for generating a synchronizing
pulse in response to the movement of a reference indicator on the
belt past a first predetermined position and a second sensor means
associated with the endless belt for generating a check pulse in
response to movement of another indicator means on the belt past
a second predetermined location. A logic circuit generates a signal
What is claimed is:
1. In a strike connection circuit arrangement for an impact printer
comprising a carrier for moving a character past a sheet support
at a nominal speed, reference indicator means for identifying the
location of the character on the carrier, first sensor means associated
with said carrier for generating a synchronizing pulse in response
to the movement of the reference indicator means past a first predetermined
location, at least one strike hammer adapted to be energized from
a rest position to a strike position in contact with the character
to effect a print out of the character on the sheet while the carrier
is moving, control circuit means operably connected to said strike
hammer for controlling the energization thereof including a logic
unit responsive to said synchronizing pulse, a delay generator connected
to said logic unit for generating a delay pulse in response to said
synchronizing pulse and a strike amplifier responsive to said delay
pulse for controlling the energization of said strike hammer, the
second sensor means associated with said carrier for generating
a check pulse in response to movement of said indicator past a second
predetermined location, circuit means for generating a deviation
signal representing the deviation of the actual carrier speed from
the nominal carrier speed in response to the time interval separating
said synchronizing pulse and said check pulse, and said delay generator
being responsive to said deviation signal to thereby control the
time in which the delay generator generates said delay pulse, said
time being proportional to the value of the deviation signal, and
said time of generation commencing with the receipt of the synchronizing
pulse by the delay means.
2. A strike correction circuit arrangement as set forth in claim
1 wherein the character support is an endless type-carrying belt
tensioned over two pulleys.
3. A strike correction circuit arrangement as set forth in claim
1 wherein said circuit means for generating a deviation signal includes
a first clock, an overall strike time measuring stage connected
to receive the output of said clock, the synchronization pulse and
the check pulse for calculating during the time interval separating
the synchronization pulse and the check pulse the number of pulses
N equivalent to the actual strike time as dictated by the actual
speed of the sheet support and the difference .DELTA.N between N
and N' where N' is the number of pulses equivalent to the strike
time relative to the nominal speed of the support.
4. A strike correction circuit arrangement as set forth in claim
3 further including a second clock, a counter connected to said
measuring stage and adapted to receive therefrom a signal corresponding
to .DELTA.N, said counter having a second and third input, means
for connecting at the second input the output of said clock and
at the third input the synchronizing pulse to thereby allow said
synchronizing pulse to be transmitted to the delay generator.
5. A strike correction circuit arrangement as set forth in 4 wherein
said synchronizing pulse is transmitted to the delay generator after
a time t.sub.r equal to the product .DELTA.N and h, where h is the
period of the pulses supplied by the second clock.
6. In a strike correction circuit arrangement for an impact printer
comprising a carrier for moving a character past a sheet support
at a nominal speed, reference indicator means for identifying location
of the character on the carrier, first sensor means associated with
said carrier for generating a synchronizing pulse in response to
the movement of the reference indicator means past a predetermined
location, at least one strike hammer adapted to be energized from
a rest position to a strike position in contact with a character
to effect a print out of the character on the sheet while the carrier
is moving, amplifier strike means for energizing the strike hammer,
control circuit means operably connected to said amplifier means
for controlling the energization thereof including a logic unit
and a delay generator connected for generating a delay pulse in
response to said synchronizing pulse, said strike amplifier being
connected to the output of the delay generator and responsive to
said delay pulse for controlling the energization of said hammer,
the improvement comprising second sensor means associated with said
carrier for generating a check pulse in response to movement of
the indicator means past a second predetermined location, logic
circuit means having a first and a second input, said first input
being connected to said first sensor means for receiving said synchronizing
pulse, said second input being connected to said second sensor means
for receiving said check pulse and said logic circuit means having
an output connected to said delay generator to thereby control the
time at which the delay generator produces the delay pulse as a
function of the speed of the belt.
7. A strike correction circuit arrangement as set forth in claim
6 wherein said logic means includes a first clock for generating
a series of pulses, an overall strike time measure stage connected
to receive said series of pulses for calculating during the interval
separating the synchronization pulse and the check pulse the number
of pulses N which is equivalent to the actual strike time as dictated
by the actual speed of the support, a second clock for supplying
a series of pulses each having a period h, the actual strike time
being equal to the product of N multiplied by h, a subtraction circuit
connected to receive the pulses N upon receipt of the check pulse
at said second input, said subtraction circuit being arranged to
calculate the difference .DELTA.N between N and N', N' being the
number of pulses equivalent to the strike time relative to the nominal
speed of said support, which is equal to the product of N' multiplied
by h, a register connected to receive the pulses corresponding to
the difference .DELTA.N, a downwards counter connected to receive
the output of the register, said downwards counter having as additional
outputs the output of said second clock and a synchronizing pulse,
said synchronizing pulse being transmitted to said delay generator
through said counter, said synchronizing pulse being transmitted
after a time t.sub.r equal to the product of .DELTA.N multiplied
8. A strike correction circuit arrangement as set forth in claim
7 wherein said first clock includes a settable clock and an adjustable
clock connected in parallel, the settable clock supplying a fixed
number of pulses and the adjustable clock supplying an additional
number of pulses which may vary between a lower limit and an upper
9. A strike correction device as claimed in claim 7 wherein the
strike time measure stage comprises a first and a second counter,
the first counter being arranged to enable the second counter to
count pulses supplied by the first clock when it receives a synchronizing
pulse and preventing said first counter from so counting when it
receives the check pulse.
10. A strike correction circuit arrangement as set forth in claim
7 wherein the strike time measure stage and a subtraction circuit
are formed by one and the same counter, said counter being preset
to a negative value equal to (-N').
11. A strike correction circuit arrangement as set forth in claim
6 wherein the character support is an endless type-carrying belt
tensioned over two pulleys.
BACKGROUND AND SUMMARY OF THE INVENTION
The present invention relates to an arrangement for correcting
the way in which print hammers strike. In particular it is intended
to correct for mis-alignment in impact printers which results from
variations in the speed of the character support.
Modern impact printers have a character support (a rotary drum
or a linear character support) which moves cyclically in front of
a text or sheet support such that all the characters are made available
at each print position.
At each print position there is a corresponding strike member which
releases a print hammer at the required time, that is, when the
character to be printed is at the appropriate print position.
This release is caused by the action on the hammer of an actuator,
normally an electromagnetic or electrodynamic actuator, which is
operated by a strike amplifier.
Although their performance is similar from the point of view of
printing speed, printers having a linear support are preferred to
drum printers at the present time because of the better standard
of print they give.
Among linear supports, endless type-carrying belts are well known.
Such a belt is described, for example, in French Pat. No. 1602392
which was filed on Nov. 19 1968 corresponding to U.S. Pat. No.
3844211. This belt is a metal one, made of steel, which is tensioned
over two pulleys having parallel axes of rotation and which is provided
on its top edge with N type-carrying fingers in the form of flexible
tongues, the entirety of these N fingers consisting of a whole number
of identical sets of different characters.
The belt passes horizontally before the text support and the group
of strike members with a uniform linear motion. Along the bottom
edge it has a line of N synchronization holes, each of these holes
being associated with one and only one of the N characters. At a
level different from that of the N holes is situated another synchronizing
hole known as the start-of-belt hole which corresponds to one particular
character among the N characters, such as the first in the series
of characters making up one of the aforementioned sets.
With the line of N holes is associated a first sensor (magnetic
or opto-electronic) which is connected to the framework of the printer
and which gives a signal each time a hole, and thus a character,
passes in front of it. This signal is amplified and shaped and is
transmitted to the logic unit which controls the printing of the
A second sensor, which is likewise connected to the framework of
the printer, is associated with the "start-of-belt" hole
and gives a signal each time the latter passes in front of it. The
signal, once amplified and shaped, is also transmitted to the logic
unit of the printer.
Combining the N signals transmitted by the first sensor with the
signal from the second sensor makes it possible to locate any of
the characters with certainty, that is, it is always known which
character-carrying finger is passing in front of a particular hammer.
To make allowance for the response time of the strike members,
it is necessary for a certain length of time to elapse between the
moment at which each synchronizing hole is picked up and the moment
at which the hammer actually strikes the character which corresponds
to this hole. To this end:
1. the axis of symmetry of each synchronizing hole is equidistant
from pairs of successive fingers; and
2. each strike amplifier has in series with it a monostable circuit
termed a "delay" circuit.
When a hole has been picked up by the logic unit of the printer
and the corresponding character needs to be struck by the hammer
which is in coincidence with the character, the logic unit gives
an order to the monostable circuit which then emits a so-called
delay pulse of adjustable duration td. When this pulse drops to
logic zero, the strike amplifier associated with the hammer provides
the hammer with the energy needed for it to strike. Duration td
depends on the strike time, i.e., the interval which elapses between
the time at which the logic unit of the printer gives the order
to strike and the time at which the hammer effectively strikes the
Duration td of the delay pulse is adjusted so that the strike time
is the same for all the hammers (this operation is termed "alignment
of the hammers"), and this duration is a function of the performance
required from the printer and thus of the speed of the belt.
The pulleys on which the belt is mounted are driven by motors having
a rotational speed subject to variations in the supply or mains
voltage. If the delay times (and thus the strike times) have been
adjusted for only a single belt speed, any variation in the belt
speed upsets the striking action and results in mis-alignment of
the print. The characters are struck too early or too late and this
detracts from the quality of the print.
To remedy this state of affairs, there are a number of solutions
in current use. When the drive is by means of synchronous motors,
an associated inverter is used. When the motors are DC, they are
However, these solutions are cumbersome and expensive and still
allow variations in speed which remain troublesome.
The present invention makes it possible to overcome these drawbacks
by advancing or retarding the time at which the monostable delay
circuit produces the delay pulse depending on whether the speed
decreases or increases. The delay time is increased or reduced in
this way, and the strike time of the hammers is adjusted in the
most favorable way as a function of the changes in the speed of
A correcting arrangement made in accordance with the invention
can be used with an impact printer which has a linear character
support. The support moves at a nominal speed V past a text or sheet
support and is provided with the same number of synchronizing reference-points
as there are characters. A fixed detector which detects the reference-points
is arranged close to the path followed by them and produces a so-called
synchronizing pulse each time a reference-point passes in front
of it. The printer also has a strike hammer responsive to a synchronizing
pulse for moving from a rest position to a strike position in contact
with a predetermined character on the sheet.
When used with such a printer, a preferred form of the invention
would include means for generating a deviation speed signal representing
the deviation of the actual carrier speed from the nominal speed.
Strike means are provided for moving the strike hammer from the
rest position to the strike position in response to a strike pulse.
The strike means typically could include a monostable multivibrator
which pulses a strike amplifier that releases the strike hammer.
Delay means apart from the multivibrator generate the strike pulse
after a delay time period proportional to the value of the deviation
speed signal, said delay time period commencing with the receipt
of the synchronizing pulse by the delay means. By using this apparatus,
the strike hammer will strike the character at the proper time irrespective
of carrier speed variations.
In accordance with other more detailed features of the invention,
the correcting arrangement comprises:
a second fixed detector (C) for detecting the reference points
located at a distance from the first detector equal to an odd number
of half inter-character spaces. The second fixed detector generates
a check pulse each time a reference point passes in front of it;
a logic correcting circuit associated with the two detectors comprising:
a first time-base clock;
a circuit for measuring overall strike time;
a subtractor circuit;
a downwards counter; and
a second time-base clock associated with the downwards counter.
Chronologically, the logic correcting circuit operates as follows:
the measuring circuit receives at a first input the synchronizing
pulse, at a second input the check pulse corresponding to the same
reference point and at a third input the pulses supplied by the
first time-base clock;
during the time interval separating the synchronizing and check
pulses, the logic correcting circuit calculates the number N of
pulses equivalent to the actual strike time relative to the actual
speed of the character support, this actual strike time being equal
to the product of N multiplied by h, h being the period of the pulses
supplied by the second time-base;
finally the logic correcting circuit transmits an actual speed
signal corresponding to the number N to the subtractor circuit as
soon as the check pulse is received.
The subtractor circuit calculates the difference N between N and
N', which is the number of pulses equivalent to the strike time
relative to the nominal speed of the said support and is equal to
the product of N' multiplied by h. In other words, N' corresponds
to the nominal speed of the support.
This difference N is a deviation signal which is transmitted to
the register and then to the downwards counter. The downwards counter
receives at a first input the synchronizing pulse and at a second
input the pulses of period h from the second time-base, and allows
the synchronizing pulse to be transmitted to the delay generator
after a time tr which is equal to the product of .DELTA.N multiplied
DESCRIPTION OF THE DRAWINGS
The invention will be better understood from the following description
of a preferred embodiment of the invention which refers to the accompanying
drawings in which:
FIG. 1a is a schematic perspective view of one embodiment of the
character support of a printer in which striking is corrected as
a function of variations in speed by a preferred embodiment of the
FIG. 1b is a schematic top plan view of the apparatus shown in
FIG. 2 is a block diagram of a preferred form of logic and control
circuitry made in accordance with the present invention;
FIG. 3 is a graphic representation, as a function of time, of the
respective trajectories of the fingers and of a hammer of the printer
shown in FIG. 1a; and
FIG. 4 is a diagram showing the definitions adopted as a convention
for the minimum and maximum strike times.
DESCRIPTION OF THE PREFERRED EMBODIMENT
The invention will be better understood if one recalls the principles
on which a printer having a type-carrying belt is constructed and
The belt 100 is shown mounted in the printer on two pulleys 101A
and 101B, the axes Y1 and Y2 of which are vertical. To make FIG.
1a easier to understand, only part of the belt 100 is shown with
its character-carrying tongues 102 to 105. The tongues are deliberately
shown overlarge in relation to the belt 100 so as to make FIG. 1a
In the same way, hammers 110 and 114 are shown, very schematically,
in FIG. 1b. The endless belt 100 moves linearly at constant speed
between hammers 110 and 114 on the one hand and the text or sheet
support 106 and an anvil 107 on the other.
The speed of the belt 100 is adjustable and depends on the printing
speed required from the printer. Fingers 102 to 105 are associated
with synchronizing holes 122 to 125 respectively, in such a way
that one and only one synchronizing hole corresponds to each finger.
As an example, the synchronizing holes may be situated between
the type-carrying fingers, as shown in FIG. 1a. Thus, hole 122 is
situated between fingers 102 and 103 hole 123 between fingers 103
and 104 etc. If belt 100 moves in the direction of arrow F, as
indicated in FIG. 1a, hole 122 is associated with finger 102 hole
123 with finger 103 and so on.
For reasons which will become clearer later, a certain number of
synchronizing holes 130 to 142 are shown on the front face of the
belt in FIG. 1a.
This belt 100 also contains a special synchronizing hole 126 termed
a "start-of-belt-hole" which is situated at a different
level from the line of synchronizing holes.
With the line of synchronizing holes is associated a sensor C1
termed the synchronizing sensor (shown symbolically by an arrow)
which gives a synchronizing pulse LNS each time a synchronizing
hole passes in front of it, this pulse being transmitted to the
strike-controlling logic circuits of the printer.
The start-of-belt hole 126 has associated with it a sensor C3 which
gives a pulse when the hole in question passes in front of it, this
pulse too being transmitted to the logic circuits.
By combining the two signals transmitted by sensors C1 and C3 it
becomes possible, in a known way, using the logic circuits, to locate
any of the synchronizing holes and thus any of the fingers. Since
the geometric relationship between the sensors, the hammers and
the fingers is known, a strike order can be given to print any desired
character on the belt.
In accordance with the invention, a second sensor C2 identical
with the first two, is arranged at the same level as the synchronizing
holes (in the same way as C1) at a distance from sensor C1 equal
to an odd number of half-intervals between characters. In a particular
embodiment of the invention, this distance is equal to 25 half intervals,
i.e., 12.5 intervals. Sensor C2 produces a BVS pulse each time a
synchronizing hole passes in front of it.
A preferred arrangement according to the invention is shown in
the form of a block diagram in FIG. 2. Since the arrangement operates
in exactly the same way for every hammer, the subsequent explanation
will deal with only a single hammer.
FIG. 2 illustrates various constituent parts of the preferred arrangement
according to the invention, namely:
sensors C1 and C2;
a first adjustable-frequency clock 1;
a circuit 2 for measuring overall strike time,
including counters 21 and 22;
a subtractor circuit 3;
a register 4;
a downwards counter 5; and
a second frequency clock 6.
This correction circuit controls the monostable delay unit 7 and
the strike amplifier 8 for the hammer in question, (110 for example).
So that the principle on which the present arrangement operates
may be understood, it is necessary to review the way in which a
hammer functions during a strike, this being illustrated in FIG.
When an LNS synchronizing signal is emitted, the strike-controlling
logic circuits actuate the monostable delay circuit 7 for the hammer
in question. This monostable circuit 7 supplies, beginning at time
to, a pulse equivalent to logic 1 of duration td (delay time). When
this pulse returns to logic 0 at time tl = (to + td), it actuates
the strike amplifier 8 which supplies energy for a strike to the
strike module. The hammer is then set in motion and describes a
trajectory M1 (FIG. 3.) During this time, the finger to be struck
by the hammer is describing a trajectory D1 moving at the nominal
speed V of the belt. The hammer strikes the finger in question at
time t2 at the point F1 where the two trajectories M1 and D1 intersect
along the time axis OT1.
There will now be considered the case where the belt suffers a
reduction in speed .DELTA.V. Its speed is now (V - .DELTA.V). The
finger in question describes a trajectory D2. 1 hammer therefore
strikes the finger at point F'1 at time t2. Length F1 to F'1 represents
the misalignment due to the variation in speed .DELTA.V. It is clear
that the hammer will strike only a part of the finger and that,
as a consequence, the finger will improperly print on the text support.
For the print to be good, it is necessary for the hammer to strike
the finger at point F2 at time t3 (representing the point where
line D2 intersects the time axis OT1). Consequently the trajectory
of the hammer needs to be altered. For this there are two possible
a. the delay pulse may be lengthened by a time tr = (t3 - t2) so
that a new delay time t'd is obtained which equals td + tr = td
+ (t4 - t1) where (t4 - t1) = (t3 - t2).
b. the time to at at which the monostable circuit 7 gives a pulse
may be delayed. This time becomes t'o= (to + tr) (shown on the auxiliary
time axis O'T2).
It is the second solution which is adopted in practice.
The object of the arrangement is thus:
1. to calculate the delay time tr; and
2. to delay the time at which the delay pulse emitted by the monostable
delay circuit 7 is triggered.
Before beginning to described the way in which the arrangement
operates, some new standard strike times will be defined:
a. the standard strike time TFC = t2 - to (see FIG. 3).
This strike time, which depends on the physical, mechanical and
electrical characteristics of each hammer, is fixed for good and
can in no case be altered.
b. the mean strike time TFMO
This time TFMO is equal to TFC + tpv + tpc, where tpv is the margin
of error of the strike time resulting from variations in the speed
of the belt (due to the characteristics of the system driving the
pulleys) and tpc is the margin of error of the strike time resulting
from inaccuracies in the geometrical setting up of the sensor (C2),
the position of the sensor being such that at speed V, the strike
time of a hammer is effectively the same as the mean strike time
c. the minimum strike time TFMin which in practice is equal to
d. the maximum strike time TFMax = TFMO + tpv + tpc.
The arrangement illustrated in the drawings operates as follows:
Let the speed of the belt be different from V and equal to V -.DELTA.V.
When a first synchronizing hole, for example, hole 130 passes in
front of sensor C1 the latter emits a first synchronizing pulse
LNS which is transmitted to counter 21 of measuring circuit 2. In
response to the first LNS pulse, counter 21 enables counter 22 to
receive pulses from clock 1 and prevents any BVS pulses from being
transmitted to counter 22. After counter 21 has received 13 LNS
pulses, the next BVS pulse is gated to counter 22 as a check pulse.
That is, the BVS check pulse generated by the passage of hole 130
in front of sensor C2 is gated to counter 22. In response to the
BVS pulse generated by hole 130 counter 22 stops counting pulses
from clock 1 transfers the accumulated count to subtractor circuit
3 and is reset to 0. In response to the fourteenth LNS pulse, the
BVS pulses are again prevented from being transmitted to counter
22 and counter 22 again begins to count pulses from clock 1. In
other words, BVS check pulses are transmitted through counter 21
to counter 22 only between every 13th and 14th LNS pulses. The frequency
dividing and gating circuitry of counter 21 required to achieve
this mode of operation would be obvious to those skilled in the
digital electronic arts.
The interval of time which elapses between the LNS pulse and the
BVS pulse produced by the same synchronizing hole is termed the
measuring time Tm.
During time Tm counter 22 in the measuring circuit counts N pulses
from clock 1 which are equivalent to the actual overall strike time
TFG for the hammer. If h is the period of the pulses from the second
clock 6 N .times. h must equal TFG. It is therefore clear that
the frequency of the pulses from the adjustable clock 1 will have
been adjusted in such a way that during time Tm (the time taken
by the belt 100 to move a distance equivalent to 12.5 character
intervals) the number of pulses N is indeed such that N .times.
h = TFG. In other words, the number of pulses N counted by the circuit
2 is proportional to the speed of the belt 100.
At the conclusion of measuring time Tm, the number N is transmitted
by the measuring circuit to the subtractor circuit which subtracts
a number N' of pulses such that N' .times. h = TFC (Standard/strike
time). At the output of the subtractor circuit are received a number
of pulses .DELTA.N = N - N' such that .DELTA.N .times. h = tr (see
FIG. 4). This number .DELTA.N is transmitted to the register 4
which in turn, transmits the number to the downwards counter 5.
At a first input, counter 5 receives a series of pulses from the
second clock 6 and counts down for a length of time equal to .DELTA.N
.times. h, i.e., tr. At a second input, counter 5 receives the signal
LNS from sensor C1. Once the number of pulses generated by clock
6 has reached .DELTA.N, signal LNS is allowed through and is transmitted
to the monostable delay circuit 7 with a delay tr. This circuit
thus in fact gives the delay pulse with a delay of tr.
The adjustable frequency clock 1 comprises:
a settable clock 11; and
an adjustable clock 12.
The settable clock 11 emits a fixed number of pulses Nf such that
Nf .times. h is slightly less than the standard strike time. Clock
11 comprises a quartz crystal oscillator and a frequency divider.
This clock delivers pulses at a rate which is not continuously variable
in a precise manner. To obtain pulses at a precise and variable
rate, an adjustable analogic clock 12 is used. This clock permits
a continuous variation in the rate at which it produces pulses.
The adjustable clock 12 supplies a number of pulses lying between
two numbers A and B such that A = (Nmax - Nf) and B = (Nf - NMin)
where Nmax and Nmin represent the numbers of pulses which are equivalent
to the maximum strike time TFMax and the minimum strike time TFMin
respectively and are such that
Nmax .times. h = TFMax
Nmin .times. h = TFMin
For example, if 1022 pulses per second are desired from clock 1
settable clock 11 is set to produce 1000 pulses per second and adjustable
clock 12 is adjusted to produce 22 pulses per second.
As previously described, circuit 2 for measuring strike time comprises:
a counter 21 for counting LNS pulses; and
a counter 22 for counting the pulses supplied by the first clock.
The counter for counting the LNS pulses generates signals to instruct
the pulse counter 22 to stop and start. When the first LNS signal
is received, counter 21 allows counter 22 to count the pulses supplied
by clock 1 until signal BVS arrives. The latter then stops counter
22 from counting. In the interval between signal BVS (equivalent
to 12.5 LNS) and the thirteenth LNS signal, number N is transmitted
to subtractor 3 and number .DELTA.N = N - N' is transmitted to register
4. When the thirteenth LNS signal arrives, the cycle of operation
is repeated as before.
To sum up, over a period of time the successive cycles of correcting
operations progress as follows. First to be considered is an initial
cycle C1 of thirteen LNS signlas. During this cycle C1 the correcting
arrangement calculates a correction of .DELTA.N pulses equivalent
to a delay tr1. During the cycle C2 of thirteen LNS signals which
is immediately after cycle C1 in time, the correcting arrangement
applies the delay tr1 (equivalent to N1) to the monostable delay
circuit 7 while it calculates a new delay tr2 (which is usually
different from tr1) equivalent to .DELTA.N pulses, and so on. In
broad terms, during a cycle Ci, the correcting arrangement applies
a delay tr (i - 1) to the monostable delay circuit 7 while it is
calculating a new delay tr(i) which will be applied to the monostable
circuit 7 in the next cycle (Ci = 1).
In a modification of the invention, the subtractor circuit 3 is
done away with and the counter 22 then performs the function of
The counter in question is preset to the value -N'. Since it counts
N pulses during the measuring time Tm, at the end of this time it
will contain the value N - N' + .DELTA.N, which is then transmitted
to register 4.
The preferred embodiments may be altered without departing from
the true spirit and scope of the invention as defined in the accompanying