Pontigny, Jacques A. (Montmorency, FR)
Collineau, Claude J. (Epinay/Seine, FR)

04/833646

12/07/1971

06/16/1969

Download PDF 3626164       PDF help

Click for automatic bibliography generation

Coulter Electronics, Inc. (Hialeah, FL)

702/100

377/50, 702/46, 377/12, 708/101

G01N15/12; G01N15/10; G06F15/36

235/92 (30)/ 235/92 (31)/ 235/92 (34)/ 235/92 (56)/ 235/150.3,153,151.3 324/71CP 328/34,46,120 307/220,226,271

Morrison, Malcolm A.

Dildine Jr., Stephen R.

We claim

1. A method for progressively and automatically correcting the digitalized accumulation of counting data which normally is subject to random error, comprising the steps of:

2. A method according to claim 1 further comprising the step of

3. A method according to claim 2 further comprising the step of

4. A method according to claim 2 further comprising

5. A method according to claim 1 further comprising the step of

6. A method according to claim 1 further comprising

7. A method according to claim 6 further comprising

8. A method according to claim 6 further comprising

9. A method according to claim 1 further comprising the step of

10. A method according to claim 1 and accomplishing said receiving and detecting by

11. A method according to claim 10 further comprising the step of

12. A method according to claim 11 further comprising

13. A method according to claim 1 further comprising the step of

14. A method according to claim 13 further comprising the steps of

15. A method according to claim 14 further comprising

16. Apparatus for progressively and automatically correcting the digitalized accumulation of counting data which normally is subject to random error, comprising:

17. Apparatus according to claim 16 further comprising

18. Apparatus according to claim 17 further comprising

19. Apparatus according to claim 17 wherein

20. Apparatus according to claim 17 wherein

21. Apparatus according to claim 16 further comprising

22. Apparatus according to claim 16 wherein

23. Apparatus according to claim 22 wherein

24. Apparatus according to claim 22 wherein

25. Apparatus according to claim 16 wherein

26. Apparatus according to claim 25 further comprising

27. Apparatus according to claim 26 wherein said obtaining means includes

28. Apparatus according to claim 16 further comprising

29. Apparatus according to claim 28 wherein said initiating means comprises

30. Apparatus according to claim 29 wherein said initiating means further comprises

31. Apparatus for automatically and progressively applying a digitalized statistic correction to a progessive accumulation of data which is transmitted in the form of a train of data pulses that is subject to random error due to the mode of transmitting such data, said error following a statistically preascertainable course, said apparatus comprising:

32. Apparatus according to claim 31 wherein said correction pulse generator comprises

33. Apparatus according to claim 32 wherein

34. Apparatus according to claim 32 wherein said plurality of inputs includes

35. Apparatus according to claim 32 wherein

36. Apparatus according to claim 35 wherein

37. Apparatus according to claim 36 wherein

38. Apparatus according to claim 32 wherein each said pulse establishing circuit comprises

39. Apparatus according to claim 38 wherein

40. Apparatus according to claim 38 wherein

41. Apparatus according to claim 40 wherein said counter comprises

42. Apparatus according to claim 41 in which said data pulses are each derived from a particle in a fluid suspension, and said apparatus further comprises

BACKGROUND OF THE INVENTION

This invention is directed to a method and counting circuitry which provide a statistic correction to a detected train of count pulses, such that effective random loss of the phenomena being counted does not induce ultimate counting error.

Now well known in the art of electronic particle counting and analyzing is apparatus marketed under the trademark "Coulter Counter." Such apparatus and portions thereof are disclosed in several U.S. Pats., for example, Nos. 2, 656,508; 2,985,830; and 3,259,842. A significantly important portion of such apparatus is the minute scanning aperture or scanning ambit relative to or through which pass and are detected single particles at a rate often well in excess of 1,000 per second. Because of the physical parameters of the scanning aperture, particles, rate of flow, etc. there frequently results the coincidence of two particles in the scanning ambit. As a result, there is effectively scanned and detected only one particle, not two.

Although such coincidence is random in time, it follows a statistically ascertainable course from which curves, tables, and formulas are obtainable. A relatively simple one of such formulas is:

n'=k(n ² /1000)

In which

n' = the total number of coincidences, i.e., the required addend;

k = a constant which relates primarily to the physical parameters of the scanning elements of the apparatus and the average particle size; and

n = the detected number of particles, the augend. Accordingly, the desired or corrected count N will equal the sum of n+n'.

Heretofore, the operator of a "coulter Counter" would obtain the augend count by analysis of a suspension of particles and then would refer to a coincidence correction chart which presented the proper correction or addend for a very large selection of augends. The sum thereof would then be the corrected count which the operator would record.

Although the use of charts provides an accurate result, it is both time-consuming and prohibits the fully automatic recording and processing of the corrected counts. Also, of course, the accumulating count during analysis is uncorrected.

The use of analog, nonlinear meters and/or elements in the output state of a "coulter Counter" has also been accomplished with limited success; however, in many used a direct reading digitalized output is greatly to be preferred.

SUMMARY OF THE INVENTION

Accordingly, it is a primary object of this invention to provide method and circuitry which provide a digitalized, direct reading corrected output for a counting apparatus that requires a continuously applied correction factor.

Another object of the invention is to provide automatic, digitalized coincidence correction circuitry for particle analysis apparatus.

A further object of the invention is to provide digitalized coincidence correction circuitry which operates in a feedback mode to incrementally supply addends to the particle count as it is being accumulated.

To achieve the above and other objects and overcome the deficiencies in the prior art, the invention provides method and circuitry exemplified by a plurality of decade counters serially coupled to the output of the particle scanner. Certain of the decade counters provide, at predetermined count levels, enabling and triggering signals to correction establishing circuits of a correction pulse generator, which in turn generate one or more correction pulses that are fed back to at least one of the decade counters. Because of the interconnection of the circuitry elements, the incremental corrections progressively cause the sum or corrected count to follow a true count curve.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a block diagram of a particle analysis apparatus including the present invention;

FIG. 2 is a diagram of the decade counter and correction pulse generator portions of the invention;

FIG. 3 is a graph showing a theoretical statistic correction curve and curves and other indicia related to the method and operation of the invention;

FIG. 4 is a schematic diagram of the combination of the delay circuit and one of the correction establishing circuits in a preferred embodiment of the correction pulse generator of the invention;

FIG. 5 is a chart showing waveforms related to the schematic in FIG. 4; and

FIG. 6 is a schematic diagram of a preferred form of the correction pulse generator of the embodied invention.

DESCRIPTION OF A PREFERRED EMBODIMENT

Although practice of this invention is not limited to its use with a "Coulter Counter," such will be the described embodiment, particularly with reference to FIG. 1. Such an apparatus comprises a detector 1, including a beaker 2 filled with a sample liquid containing the suspension of particles to be counted. A glass tube 3 of substantially smaller diameter extends into the beaker, is closed at its lower end, and is connected to a source of vacuum (not shown) and to a siphon manometer 4. The lower end of the tube 3 has an aperture 5 in its lateral wall. The aperture is of microscopic size and, for example, is in a sapphire member fixed permanently in said wall.

The aperture tube 3 and all the other accessories are filled with a compatible fluid or additional amounts of the sample liquid. Two electrodes 6 and 7 are provided on each side of the aperture 5 and are connected to conductors 8 and 9, which constitute the output terminals of the detector.

The manometer 4 is employed for causing a given amount of the sample liquid to pass through the aperture. It comprises a column of mercury 10 and its free end is open to the atmosphere. The column of mercury is so arranged that, in the course of its movement, it touches contacts 11 and 12 which serve respectively to start and stop the counting procedure. For this purpose, the column of mercury can be grounded.

To operate this detector, the operator opens a cock 13 leading to the source of vacuum so that the column of mercury moves into an unbalanced condition as illustrated. After closure of the cock 13, the column of mercury 10 tends to resume its state of balance and in doing so draws the sample liquid through the aperture 5. The column of mercury then comes in contact with the electrode 11 and an electrical potential is created across the electrodes 6 and 7 to produce a current in the liquid through the aperture. Now, in passing through the aperture, the particles in suspension modify the impedance between the two electrodes 6 and 7, so that said current is modulated and produces a series of pulses.

The conductors 8 and 9 are connected to a pulse amplifier 14 which, owing to a shaping circuit therein converts the input series of pulses into rectangular signals having a fixed width and height. The output of the amplifier 14 is connected to a decade counter array 15 comprising two transistor decades 16 and 17, each of which consists of four flip-flop circuits which are interconnected in the known manner as to afford a division by 10. The decade 17 is followed by an amplifier 18 connected to a series of three thyratron decades 19, 20 and 21.

The amplifier 18 converts the signals which issue from the decade 17, and have a width varying as a function of the mean counting frequency, into constant width pulses and amplifies them in such manner as to render them capable of acting on the thyratrons of the following decades. In a preferred embodiment, each of the latter comprises a shaping thyraton, a series of 10 counting thyratrons, and a series of 10 interrogation thyratrons. The counting and interrogation thyratrons are coupled in pairs through their cathodic load in such manner as to interrogate the decades, as will be explained hereinafter.

A control circuit 22 is connected to the electrodes 11 and 12 of the detector 1 and moreover to a multivibrator 23, which produces pulses at low frequency for example at 3 Hz. The control circuit also acts on switches 24, 25 and 26 which bring associated systems into circuit at appropriate instants. For this purpose, the control circuit comprises conventional delay circuits which will not be described in detail.

The output of the multivibrator 23 is connected, through a shaping circuit 27 and the switch 24, to an interrogation lead 28 connected to the interrogation thyratrons of the decades 19, 20, and 21. The multivibrator is also connected to the inputs of a series of three amplifiers 29, 30 and 31 whose outputs are connected to a printer 32 comprising, for each digit position of the registered number, a drum or number which is individually movable step-by-step by the effect of pulses applied to its drive means. Each of the amplifiers 29, 30 and 31 includes a blocking input 33 which is connected to one of the series of interrogation thyratrons of the considered decade through conductors 34.

A correction pulse generator 35 is associated with the counter array 15 and will be described hereinafter in considerable detail as will other circuit portions illustrated in FIG. 1, but not yet mentioned.

A source of power 36 is provided which suitably feeds each circuit of the apparatus through conductors which, in the interest of greater clarity, have not been shown.

FIG. 2 shows the block details of correction pulse generator 35 and includes the counter array 15 of which the shaping amplifiers 14 and 18 have been omitted. This circuitry is so arranged as to introduce a statistic correction in the count effected by the decades 16, 17, 19, 20 and 21 in such manner that corrected count is as near as possible to the theoretical count afforded by the formula given hereinbefore. Accordingly, an automatic correction is made each time the decade 19 emits a pulse, that is, for each thousandth pulse counted by the decade counter array. The correction cycle is divided into four parts having ranges respectively from 2,000 to 9,999; 10,000 to 39,999; 40,000 to 79,999; and 80,000 to 100,000. In each of these ranges, a different amount of correction is made. Thus, in the order or the ranges just mentioned, 20, 100,200 and 400 addend counts are added to the registered, augend, count each time the decade 19 of the thousands produces an output pulse. It should be mentioned that no correction is made below the number 2,000 in respect of the described example.

The correction pulse generator 35 comprises four correction circuits each consisting of a correction pulse establishing circuit and a control circuit, these correction circuits serving respectively for the corrections of 20,100,200 and 400 per 1,000 pulses counted. For this purpose, a lead 37 takes a signal from the output 38 of the decade 19 and transmits it to a delay circuit 39. The output 40 of this delay circuit is connected to four correction establishing circuits 41 to 44 through input leads 45 to 48. The pulses issuing from the decade 19 are also applied to the control circuit 49 through a lead 50 connected to the lead 37. The pulses issuing from the decade 20 and arriving at a terminal 51 are simultaneously applied though a lead 52 to the input leads 53 to 55 of control circuits 56 to 58, respectively. Further, a lead 59, connected to the decade 21, transmits a pulse to a triggering input of the control circuit 57; whereas, a lead 60, also connected to the decade 21, transmits a pulse to a triggering input of the control circuit 58 when the decade counter reaches the numbers 30,000 and 70,000 respectively.

An output from the correction establishing circuit 41 is connected to an input of the tens decade 17 through a lead 61. This circuit 41 transmits two pulses to this decade 17, i.e. makes a correction of 20 for each thousandth pulse registered by the decade counter array.

The outputs from the correction establishing circuits 42 to 44, respectively send, for each pulse received at their inputs 46 to 48, one, two or four correction pulses to the input of the decade 19 and advances the latter one, two, or four ranks of 100 each respectively. This is carried out through diodes 62 and a lead 63 and initiating elements to be described in connection with FIG. 4.

The correction pulse generator 35 is started and stopped by a starting and resetting circuit 64 which receives signals from the control circuit 22 in FIG. 1 through leads 65 and 66. Finally, the circuits described hereinbefore are fed from an input terminal 67 connected to the power source 36.

It should be mentioned that the embodiment described preferably employs as triggering elements cold-cathode thyratrons which have been found to be particularly suitable for the purpose; however, it is obvious that these elements can be replaced by other electronic triggering components, such as for example, vacuum tubes or semiconductors, providing that the appropriate technological adaptations are made.

Turning next to FIG. 3, there is shown a graph having a plurality of different sets of indicia along its exponentially spaced abscissa and ordinate divisions. Extending upwardly along the ordinate of the graph is the cumulative value of addend or correction applied to the decade counter array labeled "Total Correction." Extending upwardly along the right margin of the graph are the four previously mentioned ranges which commence at 2,000; 10,000; 40,000; and 80,000 corrected counts then stored in the decade counter array.

The "Detected Pulses" or augend is marked along the abscissa with indicia commencing at 700 and terminating past 80,000. Directly therebelow, on the same scale, but with necessarily earlier points of intersection of the same count values, is the "Corrected Count" index.

A theoretical or statistic correction curve 68 is plotted to show to the progressively required total correction with reference to the detected or uncorrected number of pulses which would be emitted from the amplifier 14 in FIG. 1. Thus, at any point on the curve 68 there is a corresponding total amount of required correction and this data could be tabulated to provide a correction table of the type earlier mentioned as a form of the prior art in which manual count correction is possible.

However, in order to render the correcting method and apparatus automatic, an automatic correction curve 69 is first established in the form of steps which is as near possible to the statistic correction curve, taking into account a certain previously imposed accuracy. The limits of this accuracy are indicated by a pair of curves marked +1 percent and -1 percent designating their value limits. Of course, it is possible to be more or less exacting in the accuracy by providing an automatic curve having a greater or smaller number of steps and consequently a greater or smaller complication of the electronic circuits of the apparatus.

Before discussing the schematic details of FIGS. 4-6, the overall operation of the invention will be discussed from the point in time that the 1,000th detected pulse has been received by the decade count array 15. At such time, the decade 19 emits at its output 38 a triggering signal to the control circuit 49, by way of the leads 37 and 50, which enables the correction establishing circuit 41, but does not elicit therefrom any correction signals. Upon receipt of the 2,000 detected pulse, the decade 19 again produces a trigger signal which, by way of the delay circuit 39 and the leads 40 and 45, is coupled to the correction circuit 41 and causes it to produce two correction pulses. The latter are applied to the input of the decade 17 through the lead 61. The decade 17 thus advances two ranks and therefore adds 20 to the registered count, the indicated total or corrected count then being 2,020 pulses. Thereafter, each time the decade 19 emits an output signal, the correction circuit 41 sends two pulses to the decade 17 so as to add 20 to the registered count. Thus, when the counter indicates the number 10,000 the detector has delivered only 9,840 pulses.

At the count of 10,000, the decade 20 emits a pulse at the output 51, the control circuit 56 is actuated through the leads 52 and 53 to enable the circuit 42, while the pulse establishing circuit 41 is simultaneously inhibited (by circuit means subsequently to be discussed). When once again the decade 19 emits a pulse, the pulse establishing circuit 42 is triggered by way of the delay circuit 39 and the input lead 46 and the circuit sends one pulse to the input of the decade 19 through the lead 63 so as to advance it one rank and cause it to make a correction of 100. Each time, the decade 19 thereafter emits a pulse a correction of 100 is made by the pulse establishing circuit 42.

When the decade 21 attains the number 30,000, it sends a pulse on the lead 59 to the control circuit 57 so as to enable it. Upon the arrival of the following output signal, at the terminal 51 of the decade 20 representing the number 40,000, the control circuit 57 is triggered by way of the leads 52 and 54 and enables the pulse establishing circuit 43 and also inhibits the pulse establishing circuit 42. Upon the arrival of the following pulse at the output of the decade 19, number 41,000, the pulse establishing circuit 43 sends two correction pulses to the input of the decade 19 so as to provide an addend of 200. Thereafter, each time the decade 19 emits a pulse, a correction of 200 is made in the count of the counter array 15. Thus, when the latter indicates the corrected count of 40,000, only 37,100 pulses have been delivered by the detector 1.

When the decade 21 attains the corrected count of 70,000, a pulse is sent to the control circuit 58 so as to enable it through the lead 60. Thereafter, as soon as the decade 20 emits the following output pulse, number 80,000 on the leads 52 and 55, the control circuit 58 enables the pulse establishing circuit 44 and inhibits the pulse establishing circuit 43. The following pulse emitted by the decade 19, number 81,000, produces four pulses at the output of the pulse-establishing circuit 44, these pulses being sent to the decade 19 and providing an addend of 400. This is thereafter repeated until the counter array 15 is completely full.

It will be observed that the counting could very well stop before the counter 15 is full, the number finally registered depending on the number of particles contained in the measured volume of the sample liquid. Moreover, the counting is stopped by the column of mercury 10 when it touches the contact 12. At that time, the given volume of the liquid sample has passed through the measuring aperture 5. The stopping of the counting by means of the column of mercury 10 also triggers the multivibrator 23 through the control circuit 22, and the latter closes the switches 24 to 26 at the appropriate moments.

The multivibrator 23 transmits the interrogation pulses to the shaping circuit 27 and these pulses are thenceforth applied to the interrogation thyratrons of the decades 19 to 21 so as to interrogate them. At the same time, the multivibrator 23 transmits pulses to the amplifiers 29 to 31 which are connected to the printer mechanisms corresponding to the hundreds, thousands and ten thousands through the switches 26.

The interrogation pulses have two particular purposes. They cause the drums of the registering mechanisms of the printer to advance simultaneously step by step, and they interrogate the interrogation thyratrons of the decades 19 to 21. As soon as, in the interrogation series of the decade concerned, the interrogation thyratron corresponding to the energized counting thyratron is reached, a blocking pulse is sent to the corresponding amplifier through the lead 34 so that the amplifier is blocked. The drum of the printer 32 then stops at the number represented by the interrogation thyratron that was energized.

In order to further facilitate an understanding of the invention and its operation, a preferred embodiment of one of the correction pulse establishing circuits will be described with reference to FIG. 4, this circuit being similar to each of the circuits 41-44 and associated with a delay circuit similar to the circuit 39.

As shown in FIG. 4, two thyratrons 70 and 71 have their anodes connected to a +200 v. source, for example, through anode resistors 72 and 73. The cathode of the thyratron 70 is grounded through an RC circuit comprising a resistor 74 and a capacitor 75. The grid of the thyratron 70 receives control pulses through a capacitor 76 and it is polarized by a +100 v. source, for example, through a resistor 77. The cathode of the thyratron 70 is also connected to a coupling capacitor 78 which is connected to the grid of the thyratron 71. This grid is polarized at the voltage of +100 v. through a resistor 79. The cathode of the thyratron 71 is grounded through an RC circuit consisting of a resistor 80 and a capacitor 81 and is connected to the output of the circuitry through a diode 82. Note that the RC time constant of the resistor 74 and the capacitor 75 is much greater than that of the resistor 80 and capacitor 81.

The operation of this circuitry is as follows with reference to FIG. 5. An input or count pulse 85 is applied to the capacitor 76 and causes, after a certain inherent delay t, the ionization of the thyratron 70 which, being conductive, charges the capacitor 75. This charging continues until the voltage at the terminals of the capacitor has such value that it lowers the anode-cathode voltage of the thyratron below the ionization value. The thyratron thus becomes nonconductive. The time constant of the RC circuit 74 and 75 is so calculated as to produce the conduction curve 86 in which a threshold voltage 87 defines the conduction range of the thyratron 70.

The operation of the thyratron 71 is identical to that of the thyratron 70. Thus, a pulse applied to the grid of the thyratron 71 triggers the latter and charges the capacitor 81 until the voltage at its terminals reaches such value that the thyratron ceases to conduct. As mentioned before, the charging time of the capacitor 75 is much longer than that necessary for charging the capacitor 81. Consequently, during the time which elapses between the thyratron conduction range of the curve 86, during which the capacitor 78 is also charged, the latter periodically transmitting a part of its energy to the grid of the thyratron 71 by way of a pulse train 88, the thyratron 71 is periodically conductive and periodically charges the capacitor 81. The number of times that the capacitor 81 can be charged and discharged therefore in particular depends on the capacity of the capacitor 78.

For a resistor 74 of 560 ohms, a capacitor 75 of 47 microfarads, a resistor 80 of 1 megohm and a capacitor 81 of 0.22 microfarads, a value of 1.5 microfarads must be chosen for the capacitor 78 to obtain one pulse, 2.7 microfarads to obtain two pulses and 5.1 microfarads to obtain four pulses, the latter as in the pulse train 89, per input pulse 85.

In the embodied apparatus, there are employed only the single delay circuit 39 such as that described hereinbefore, and the four correction pulse establishing circuits 41-44, each of which comprises an input capacitor, such as the capacitor 78 having the appropriate value. A schematic thereof is shown in FIG. 6. Thus, the output lead 40, which is connected to the junction point of the capacitors 75 and 78 in FIG. 4 is similarly connected to each of the correction pulse establishing circuits 41-44 through their respective capacitors 78 in FIG. 6.

As previously mentioned, each of these pulse establishing circuits is associated with a respective control circuit 49, and 56-58, comprising a thyratron 90 which is capable of polarizing, when rendered conductive, the grid of the thyratron of the associated pulse establishing circuit. This is effected through the respective leads 91-94 also shown in FIG. 2. The thyratrons of the control circuits can be polarized by a polarizing voltage through the lead 52 to the input leads 53-55. Thus, the thyratrons are polarized only upon the counter array 15 being filled up to 10,000. On the other hand, the thyratron of the control circuit 49 is polarized as soon as the operator has started the apparatus and ionized the thyratron 95 of the circuit 64 by means of the terminal 65.

The thyratrons of the apparatus are arranged in two main groups A and B as to their anode source. For this purpose, the power source 36 comprises two distinct parts each of which consists of a common anode resistor 96 for the considered group, this resistor being in series with a diode 97 between two supply lines 98 and 99 brought respectively to potentials of +400 and +200 v. The terminals A and B are at the junctions of the respective resistors and diodes and are connected to the anode terminals of the thyratrons of the respective groups, these terminals having the corresponding letter references in FIG. 6. Owing to this anode coupling of the two groups of thyratrons, only a single thyratron per group can be triggered; and the thyratron, which is triggered at a given moment by means of a voltage applied to its grid, renders nonconductive the previously conductive thyratron in respect of which the anode voltage is then too weak to maintain the ionization.

The operation of the circuitry shown in FIG. 6 will now be described in detail from a starting point at which it will be assumed that none of the thyratrons of the circuitry is conductive. Starting up is achieved by the application of a pulse at the terminal 65 of the start circuit 64, its thyratron 95 being rendered conductive and the grid thereof being permanently polarized by a potential of +100 v. The current which is established in the thyratron 95 supplies a potential by way of a lead 100 to the grid of the thyratron 90 in the control circuit 49 so as to polarize it. The pulse establishing circuit 41 is then enabled to receive the thousandth pulse recorded by the counter.

At the moment the decade 19 counts the thousandth pulse, it transmits a pulse to the delay circuit 39 through the lead 37 and as the thyratron 70 is permanently polarized by a potential of +100 v. on its grid, it is rendered conductive. At the same time, the thyratron in the control circuit 49, previously polarized through the lead 100, is rendered conductive by way of the lead 50 and thus polarizes the thyratron 71 in the correction circuit 41 through the concerned lead 91.

Upon the arrival of pulse 2,000 in the counter, the decade 19 transmits a new pulse to the delay circuit 39 so that the thyratron 70 previously rendered nonconductive by its cathode capacitor 75, is rendered once more conductive, and owing to the polarization of the thyratron 71 and to the value of the capacitor 78, the latter will produce two pulses which are transmitted by the output lead 61 of the correction circuit 41 to the tens decade 17 so as to cause it to advance two ranks.

An addend value of 20 is thus added to the uncorrected count value of 2,000 and the counter jumps to the number 2020 as shown in the lower left corner of FIG. 3. This correction is thenceforth made each time the decade 19 produces an output pulse corresponding to the number 1,000, so that at the arrival of the pulse 9,840 at the input 101 of the counter 15, the latter displays in fact the coincidence corrected number 10,000.

As soon as the pulse corresponding to the number 10,000 issues from the decade 20, the thyratron in the control circuit 56 is rendered conductive through the leads 52 and 53. Simultaneously, the corresponding thyratron in the control circuit 49 is turned off owing to the common anode coupling to point A in the power source 36 as earlier explained; this operation thenceforth preventing the conduction of the thyratron in the correction circuit 41 whose polarization is eliminated. On the other hand, the grid of the thyratron in the correction circuit 42 is polarized through the associated lead 92. As soon as the decade 19 once more records the arrival of 1,000 pulses; that is, when the counter displays 11,000, it transmits the signal to the grid of the thyratron 70 in the delay circuit which renders it conductive. Then, owing to the value of the capacitor 78 in the circuit 42, only one pulse is produced by the thyratron 71 therein, this pulse being transmitted to the input of the decade 19 through the concerned diode 62 and the lead 63. As the decade 19 advances one rank, a correction value of 100 is added to the number displayed by the counter 15, which consequently displays the number 11,100. At this moment, the detector has transmitted to the counter only 10,840 pulses. This correction of 100 appears on FIG. 3 with associated reference 102 and returns the automatic correction curve 69 from -1 percent error to very close to the theoretical correction curve 68. This correction is repeated thereafter each time the decade 19 emits an output signal until the counter displays the number 40,000. However, at the number 30,000, the decade 21 transmits a pulse through the lead 59 which thus polarizes the thyratron in the control circuit 57. As soon as the counter displays the number 40,000, the decade 20 renders the same thyratron conductive through the leads 52 and 54. At this moment, the counter has received at its input terminal 101 only 36,940 pulses. Owing to the common anode coupling, the thyratron in the control circuit 56 is turned off and simultaneously eliminates the polarization of the grid of the thyratron 71 in the associated correction circuit 42. The latter can no longer produce correction pulses. On the other hand, the thyratron in the control circuit 57 polarizes the grid of the thyratron in the correction circuit 43 which, upon the arrival of the following pulse at the output of the decade 19 which is, at number 41,000, becomes conductive twice, effecting the correction value of 200, two pulses being applied to the input of the decade 19. This correction of 200 is made every thousand pulses until the counter 15 displays the number 80,000. However, at that moment, the counter has received only 68,300 input pulses. Because of the correction beginning at 40,000, the automatic correction curve 69 again swings away from the -1 percent curve toward the +1 percent curve; however, nearing the count of 80,000, the -1 percent curve is again being approached.

The arrival of the pulse 70,000 polarizes the thyratron in the control circuit 58 through the lead 60 so that when the decade 20 emits a pulse when it is once more filled by the following 10,000 pulses, that thyratron is turned on and the thyratron in the prior control circuit 57 is turned off, which renders the thyratron in the correction circuit 43 inoperative. On the other hand, the thyratron in the correction circuit 44 then is able to produce pulses. Consequently, upon the arrival of the pulse 81,000, the decade 19 excites the thyratron 70 in the delay circuit 39 so as to produce four pulses in the correction circuit 44 owing to the value of the capacitor 78. These pulses are applied to the decade 19 through the considered diode 62 and the lead 63. The decade 19 then advances four ranks so as to make a correction of 400. The display of the counter 15 is thereafter corrected in this way every thousand pulses until the total filling of the counter. If the apparatus is not wholly stopped by elimination of the source voltages on the lines 98 and 99, the thyratron in the circuit 58 remains energized until the arrival of the following starting pulse at the terminal 65. The latter, in rendering the thyratron 95 conductive, turns off any one of the thyratrons 90 which was the last to remain energized.

It should be mentioned that the system described hereinabove represents one of many possible embodiments of the correction pulse generator 35. Further, the numerical values of the components and the numerical examples illustrating the operation of the apparatus have been given merely for purposes of explanation. The same is true in respect of the correction curves shown in FIG. 3, since they depend both upon the uncorrected error produced by the detecting apparatus and the degree of accuracy desired.

What is desired to be protected by United States Letters Patent is: