DE2140771A1 - Electronic calculating machine - Google Patents

Description

Applicant: Sybron Corporation, 1100 Midtown Tower, Rochester, New York, 14604, USA

Electronic calculating machine

The invention relates to an electronic calculating machine which can automatically perform divisions.

Often the numerator and denominator characterize physical variables such as pressure, temperature, material properties, time or the same quantities, which for whatever reason can become zero. In particular, a temporal averaging can be performed involve division of a variable so that when performing a step averaging for a series of intervals a division by zero can occur at the beginning of the interval.

In known electrical calculating machines, the Means for performing a division a high gain amplifier for amplifying a counter voltage, wherein a negative feedback loop is provided around the amplifier so that the output of the amplifier reflects the counter voltage. The loop eventually becomes through a denominator voltage controlled to make the feedback proportional to the denominator voltage. Therefore represents the output voltage of the amplifier the counter voltage, which is determined by the Denominator is divided. If the denominator voltage is sufficiently small is, the feedback becomes zero as a result, so that the

209810/0287

-2- 2U0771

The amplifier's output signal no longer matches the input signal can correspond.

It is therefore the object of the invention to design an electronic calculating machine of the type mentioned in such a way that division by zero is prevented.

An electronic calculating machine differs according to the invention of the calculating machines of the type mentioned in that a part of the controlled by the denominator voltage Feedback loop in the result is replaced by an element that provides feedback as a function of the denominator if it becomes zero or sufficiently small. The invention is particularly useful for systems for averaging of moisture content is applicable along the width of a moving paper web. A measuring instrument probes to determine the moisture content from the paper web transversely to its direction of movement. The average moisture content a fraction of the width of the web is obtained by dividing the moisture content along the width of that fraction is integrated. This mean value is divided by the time it takes to measure the moisture content of this Fraction has elapsed, so the resulting quotient is the average moisture content of a strip is the path, the width of which is proportional to the measurement time. This procedure is used for the nearest fraction of the web and the the following fractions repeated. Therefore, a set of averages of moisture content results for a series of Strip of web. However, every time the averaging starts in a strip, the elapsed time is the same Zero, so the division of the integrals of the moisture content by "the time is taken into account according to the invention.

The elapsed time is determined by an integrator in the form of a capacitor placed at the beginning of each strip is discharged so that feedback occurs during a short time interval corresponding to the beginning of a strip the mentioned element is retained.

In summary, therefore, the essential characteristics of a

209810/0287

electronic, calculating machine according to the invention to be seen in that division by zero is prevented by applying a counter voltage to a high gain amplifier which has a negative feedback loop created by a denominator voltage which is part of the negative feedback loop ineffective when this voltage becomes zero. However, the effect caused by this is avoided by this part of the loop is replaced depending on a denominator voltage equal to zero.

The invention is to be explained in more detail with the aid of the drawing. Show it:

Fig. 1 is a schematic circuit diagram of an electronic Calculating machine according to the invention, suitable for performing divisions; and

Fig. 2 a device for averaging for a Calculating machine according to the invention.

In Fig. 1, the numerator voltage e and the denominator voltage e "the non-inverting input connections 1 and 2, respectively the differential amplifier 10 or «20, which have a high gain. A feedback voltage e ,, and a Sawtooth voltage e "" are applied to the inverting terminals 11 and 22 of amplifiers 10 and 20, respectively. In this example, all four voltages are negative.

The feedback voltage e ,, is derived from the output voltage e _n at the output terminal 4 of the amplifier 10 via a negative feedback loop, which essentially consists of a capacitor 5 connected in parallel to the resistor between the terminals 4 and 11, which resistance is provided by the resistors 6 and 7 and the field effect transistors 8, 9 and 12 is given.

The gate electrodes of the field effect transistors 8, 9 and 12 are connected to the output terminals 13, 14 and 15 of the amplifiers 30, 20 and 40, respectively. Field effect transistors 9 and 12 are connected to source and drain, the drain electrode of the Field effect transistor 9 is connected to the common node CC. The source electrode of the field effect transistor

209810/0287.

12 is connected to the output terminal 4. One end of each of the resistors 6 and 7 is connected to the drain electrode of the field effect transistor 8 while the other end of the resistor 6 is connected to the output terminal 11. The other end of the resistor 7 is with the source-drain connection of the field effect transistors 9 and 12 and the source electrode of the field effect transistor 8 is connected to the output terminal 4.

The amplifier 30 is a high gain differential amplifier having a non-inverting input terminal 3 and an inverting input terminal 33, the former being connected to the input terminal 2 and the latter being connected to W of a source B for a reference voltage.

The amplifier 40 is an inverting unit gain amplifier whose input terminal is the same as the output terminal of amplifier 20 is.

The sources for the various input voltages are indicated by a box, and relate to a sawtooth voltage S, a denominator voltage N, a reference voltage B and a numerator voltage Z. The output voltage e_. at terminal 4 is a the quotient Q using device which is connected to the connection 4 and e.g. a control device or a Display device is.

The purpose of the sawtooth generator, the amplifiers 20 and 40 and the field effect transistors 9 and 12 is that the feedback voltage e ₁ corresponds to the denominator voltage e 2. For this purpose, the voltage e _{22 should} have such a profile that it starts at zero and increases linearly to a given value which is equal to or greater than the highest value of the voltage e ₂ to be expected, and which repeats itself at a given rate , which is much greater than any rate at which the voltage e "can change its magnitude. The voltage e ₂ , on the other hand, may be the output of a measuring device which detects a condition such as temperature, material properties or the like and causes the voltage e _{2 to have} a predetermined quantitative relationship with this condition. The voltage e, can have practically the same character as

209810/0287

2U0771

have the tension e «.

Since the amplifier 20 is a differential amplifier with high gain, when the voltages e 1 and e _{22 are} supplied, it emits an output voltage e ₁₄ at the terminal 14, the sign of which depends on which of the voltages e_ and e "" is greater. If at a given moment the voltage e _{0 is} greater

is

as the voltage e is ₂₂ then ax voltage e ^. positive, while e- _{4 is} negative when the voltage e " _{2 is} greater. The output voltage e ,,. of the amplifier 20 also always has the opposite sign, but the same size as the voltage e ₁₄ .

The amplifier 20 is designed so that the voltages & - .. and e _{15 is} always greater than the blocking voltages of the field effect transistors 9 and 12. Therefore, and because voltages e, «and e, ₅ always have the opposite sign, one of the field effect transistors is always conductive and the other is switched off, the conductive or non-conductive state depending on which of the voltages e_ and e ₂ _ is greater.

When the field effect transistor 12 is conductive, a feedback loop is present, which consists of the capacitor 5 in Parallel connection to resistors 6 and 7 and the drain-source resistor of the field effect transistor 12 is provided. If the field effect transistor 12 is non-conductive, then that is Input terminal 11 with the common circuit point over the series resistance of the resistors 6 and 7 and the drain-source resistance of the field effect transistor 9 are grounded.

If the voltage e _{2 is} greater than zero, but less than the maximum value of the voltage curve, then the point in time at which the voltage rise is just slightly greater than the voltage e ₂ corresponds to the magnitude of the voltage e ₂ · Therefore is during any given The voltage rise is the time during which the field effect transistor 12 is conductive, proportional to the magnitude of the voltage e ₂ · Furthermore, the overall gain of the amplifier 10 is proportional to the reciprocal value of this time. Therefore the voltage e at the terminal 4 is proportional to e, / e ".

209810/0287

2U0771

The above-mentioned mode of operation is known per se. However, it is also known that in known calculating machines it is not permissible that the voltage e _? Becomes zero, or that a minimum value must not be exceeded. The reason for this can be seen in the fact that the charging of the capacitor 5 is no longer caused by the voltage e_ when the voltage e is less than e "^ ^st . ^Since the field effect transistor 9 is conductive, the amplifier 10 and the Capacitor 5 as an integrating device and cause the voltage e _{Q to} reflect only the specifics of the amplifier 10, such as, for example, its offset voltage and the bias current, which is known per se. While the time constant of the discharge of the capacitor 5 is fixed by the selection of the capacitance of the capacitor and the resistors 6 and 7, so that the capacitor 5 cannot discharge significantly during a given voltage curve, e ₂ changes slowly enough compared to the sawtooth voltage at e " ₂ that, if e ₂ becomes zero or near zero, it is generally maintained so long that capacitor 5 is discharged to an extent that, as a result, there is no feedback at terminal 11 at what point in time the amplifier 10 reaches the saturation current.

The above-described effect that e ₂ becomes zero or almost zero Ψ can be prevented by the amplifier 30 and the field effect transistor 8. During normal operation, if e _{2 is} not equal to zero or differs therefrom by an amount which corresponds to a predetermined minimum value, the amplifier 30 generates a voltage e, ₃ with such a magnitude and a sign that the field effect transistor 8 is switched off , in which state its drain-source resistance is infinite for practical purposes and therefore does not affect the operation of the amplifier 10. As can be seen from the illustration, the voltage source B connected to the input terminal 33 supplies a fixed reference voltage e ₃ - for a reference voltage, which is just set to such a value that e, ₃ turns on the field effect transistor 8 when e _{2 is} equal to zero or has a value so small that it corresponds to the influence of the value zero. if

209810/0287

therefore e _{2 is} just slightly larger than e ^ ₃ , then the voltage e, _{3 changes} its sign and switches off the field effect transistor 8. When the field effect transistor 8 is switched on, there is feedback to the terminal 11 via the capacitor 5 in parallel with the series resistance of the resistors 6 and the drain / source resistance of the field effect transistor 8. While the connection 11 can still be in connection with the common node CC via the field effect transistor 9, the resistor 7 is selected so that it has such a value in relation to the drain / source resistance of the field effect transistor 8 that a sufficient Feedback voltage is present when the field effect transistor 9 is conductive, at the connection between the resistor 7 and the drain electrode of the field effect transistor 8 in order to control the output signal of the amplifier 10, so that e. 'A suitable, predetermined fixed multiple value of e, received at this point. In particular, it is expedient to select the resistor 7 to be sufficiently large so that e _{n is} practically equal to e i at this point in time. In addition, the resistor 7 limits the drain current to acceptable values when the field effect transistors 8 and 9 are both conductive.

Numerous modifications of the exemplary embodiment shown in FIG. 1 are possible. For example, various reversals of the specified polarities can take place, as well as those that are given by the nature of the related components. For example, e "and e ₂ _ could be positive voltage gen. In this case, the connections 2 and 22 would be the inverting and non-inverting input connections of the amplifier or the field effect transistors 9 and 12 would be p-conducting and not η-conducting, as shown.

The field effect transistor 12 could also have a p-channel Field effect transistor can be replaced, in which case the amplifier 40 could be omitted because the then existing Gate electrode could be connected directly to terminal 14.

The task of the field effect transistors is essentially to switch over with an activated resistor to be carried out, which is negligibly small in comparison to resistors 6 and 7, while a disconnected resistor

209810/0287

" ⁸ " 2U0771

stand is present that is large enough that in the event of a shutdown the circles are effectively open. In Figure 1, the drain and source electrodes are distinguishable, although not is essential, in particular because field effect transistors are available in which the drain and source are interchangeable.

In the embodiment of Figure 2, a moisture measuring device F is moved by a mechanism M at right angles to the web and the direction R of web travel. The measuring device therefore moves along a zigzag path along which the moisture content is measured. If the web and the measuring device move at constant speed, then it can be seen that the web can be viewed as being divided into zones a, b, c and d running in the longitudinal ψ direction, each of which is divided into sub-zones a ,, a, a, a., etc., with each sub-zone corresponding to a single determination of an average moisture content. While the division into zones and sub-zones partly depends on considerations such as the intended use or the interpretation etc. of the average values, it can also depend on the type of averaging, as will be explained in more detail below.

When the web is moving and the measuring device in the transverse direction is moved and measures the moisture content of the web, mean values are determined and zones and sub-zones are defined essentially by the division system in Fig. 1 and certain modifications. To avoid repetition, only that part of the system in FIG. 1 is shown in FIG. 2 / which is required for the modifications mentioned and to be able to explain the use of the system in FIG. 1 in the moisture measurement. So much for Fig. 2 in Fig. 1 Contains parts shown, these are denoted by the same reference numerals.

The moisture measuring device is with an integrating circuit I connected, whose box in the figure corresponds to the box for the counter voltage Z in FIG. Correspondingly the box for the denominator voltage N in Fig. 1 is through a box representing an integrating circuit I is replaced.

209810/0287

_9_ 2H0771

The integrating function of the integrating circuits is given by the capacitors 16 and 18, parallel to which field effect transistors 17 and 19 are connected. The voltages at terminals 1 and 2 in this arrangement represent the Voltages with respect to the common node to which capacitors 16 and 18 are charged.

The integrating circuit with the capacitor 16 is connected to the humidity measuring device F in order to obtain a voltage of the measuring device to integrate what voltage is the moisture content the trajectory that was determined by the measuring device. Therefore the voltage at connection 1 is the time integral of the output voltage according to the humidity measurement.

The integrating circuit with the capacitor 18 is connected to a fixed DC voltage source, which is used as cell 23 is shown to integrate the voltage of cell 23. Therefore, the voltage at terminal 2 is the time integral of Voltage of cell 23. Since the voltage of the cell is constant, the time integral simply indicates the elapsed time, or finally the length of the path of the measuring device relative to the web.

It can be seen that the quotient at terminal 4 is an average value of the moisture content in the paper web reflects. However, the integral circuits cannot be indefinite integrate, because of the properties and parameters of practical integrators. That's why the capacitors need periodically discharged. This takes place with the aid of the control device St, which is connected to the mechanism M, for example is to determine an appropriate time during the movement of the moisture meter at which the capacitors be discharged. In this case, the control device outputs a field effect transistors 17 and 19 that switch on Pulse from, approximately at the point in time after a sub-zone has been crossed by the measuring device. This impulse lasts just long enough to discharge capacitors 16 and 18 to the common node potential. After this The field effect transistors 17 and 19 are at the end of this pulse

209810/0287

2U0771

switched again and the integrating xnr xchtungen begin to reintegrate.

When the capacitors discharge, the voltages at terminals 1 and 2 become zero or very low, and at the same time as the pulse for the capacitor discharge, the control device emits a pulse which is fed to the gate electrode of the field effect transistor 8 via a terminal 131, to turn on the field effect transistor 8. Therefore, in FIG. 2, the control device and the terminal 130 corresponding to the amplifier 30 and the terminal 13 in Fig. 1 and k practice DIE same function with respect to the field effect transistor 8.

Finally, the difference between the two figures is how the denominator voltage becomes zero or near zero. In Fig. 1, the occurrence of the denominator voltage zero, which is however generated, is detected. In Fig. 2 is the denominator voltage Zero an averaging event. As in practice, in Fig. 2 the occurrence of the denominator voltage is not zero to be expected, so it is more convenient that the control device the field effect transistor 8 switches on, to be provided as a special detection device for the state with the voltage zero, like amplifier 30.

The triggering pulse preferably occurs at the connection 130 ″ somewhat before the triggering pulse for the field effect transistors 17 and 19 and lasts a little longer. The reason for this is to be seen in the fact that the time integrator is essentially a sawtooth generator, which produces a voltage that increases from zero with a constant increase to a given value and instantly becomes zero again when it has reached this given value. In fact, the condenser is draining 18 does not instantly go to zero and there is some uncertainty about the point at which the voltage across the capacitor 18 then begins to follow the sawtooth shape. Therefore there is an interval which marks the end of a sawtooth pulse and overlaps the beginning of the next in which the voltage at terminal 2 can become zero, but otherwise of the ideal case deviates. In the invention, the field effect transistor

209810/0287

2U0771

8 is turned on during this interval (whichever is connected to port 4 is marked, so to speak, that e has the predetermined value, which is determined by the resistor 7, so practically the value of e, which is zero at some point in this interval) so that no division occurs, except when the voltage e "has the desired sawtooth shape. Important it is here that to the extent that the e "from the ideal form to deviates from the appropriate extent, there is an error in the elapsed time measurement.

The successive pulses emitted by the integrator are generated with the capacitor 18, form a sawtooth voltage with the frequency of the discharge as a result of capacitors 16 and 18. This frequency must be small in comparison to the sawtooth frequency at the terminal 22, analogous to the limitation of the change in the denominator voltage in FIG. 1.

From the above it can be seen that the subdivision of the railway basically depends on the limited capacity of the integrators. For example, there are limits for the magnitudes of the stresses that occur in the operation of the device which is used for the moisture measuring device, of the integrating devices, so that the maximum Voltage for capacitors 16 and 18 is slightly less than one or the other of these limits. The subdivision is however, it is actually an advantage that the waveform at terminal 4 is correlated to the zones and the sub-zones. Therefore, by examining the waveform at the terminal 4, a wet spot on the web can be detected and checked the zone or sub-zone in which it occurred. If, furthermore, the mean values are to be used, Determining the total moisture content of a large section of the material web can be caused by wet stains Effects are filtered out because the normal moisture content a few percent by weight of the paper in a normal sub-zone, while a wet spot is in a sub-zone a brief peak voltage is generated, which is more pronounced the smaller the sub-zone is.

209810/0287

_ ₁₂ _ 2U0771

The mechanism for traversing the zones of the web, the moisture meter, the sawtooth generator and integrators are common devices of known type. The control device can be any device, which emits the control signals in the form of pulses at suitable points along the movement path of the measuring device. There are therefore no difficulties in producing suitable circuits which the above-described working program of the Control device can perform.

Claims

209810/0287

Claims (7)

-13- 2H0771 Claims / l) Electronic calculating machine for carrying out divisions, with a high gain amplifier before a feedback and negative feedback loop connects the input to the output of the amplifier so that the amplifier emits a first signal at its output which is substantially proportional a second signal fed to the input thereof, the feedback loop comprising an element which is variable as a function of a third signal which is fed to it, so that the amplifier is the first Signal generated proportional to the size of the second signal divided by the size of the third signal, and where the No feedback signal as a result of the feedback loop Change of the feedback element as a function of the third signal when this has practically the value zero, and with a signal generator which is connected to the feedback element, to the feedback element the third To supply a signal, which signal generator has the property of generating the third signal with a value corresponding to zero, characterized in that a second feedback element is independent of the first feedback element connected between the output and the input of the amplifier is that the second feedback element is operable between first and second states that the first state is such in which the second feedback element has no feedback between the output and the input causes, while the second state is one in which the second. Feedback element a feedback between output and input causes which second feedback element is normally in the first state, and that a control device is effective when the magnitude of the third signal is practically zero to the second feedback element to switch to the second state. 2. Calculating machine according to claim 1, characterized in that that the control device on the third 209810/0287 Signal responds when its size is practically zero to switch the second feedback element in the second state. 3. Calculating machine according to claim 1, characterized in that that the control device periodically reduce the size of the third signal practically to the value zero and can switch the second feedback element to the second state practically at the same time. 4. calculating machine according to claim 3, characterized marked ψ is characterized in that the signal generator comprises a voltage ramp generator for generating a ramp signal and a sawtooth generator for generating a Sägeζahnsignals that the rise signal is the third signal and the first feedback element in common by the rise signal and the sawtooth signal is controlled so that there is practically no feedback when the ramp signal is practically zero. 5. Calculating machine according to claim 4, characterized in that that the control means the first feedback element in the first state during a time interval switches, which begins before the rise signal falls practically to the value zero, and which ends after the rise signal the next time began to rise above zero. 6. Calculating machine according to claim 1, characterized in that the signal generator is an integrating circuit having a capacitor in which a fourth signal can be stored, so that the size of the third signal represents the time integral of the fourth signal, and that means for deriving the stored fourth signal from the capacitor is provided to such an extent that the third signal is practically zero. 209810/0287 2U0771 7. calculating machine according to claim 1, characterized in that that the device responsive to the third signal is a field effect transistor at its gate electrode a device is connected to switch off the field effect transistor, except when the size of the third Signal practically zero is that this circuit device responds to the third signal if this is practically the same Zero becomes a drain-source resistance of the field effect transistor to cause which can provide the predetermined amount of negative feedback, and that the drain-source resistance is connected in parallel with the first feedback element so as to provide the second feedback element. 209810/0287 Empty soap