DE1548609C - Method for determining the mean value of a plurality of variables and device for carrying out such a method - Google Patents

Description

The invention relates to a method for determining the mean value of a multiplicity over time Distance between successive physical quantities by converting them into corresponding electrical ones Measuring pulses, the amplitudes of which are proportional to the amount of the physical quantities that trigger them are and their time interval corresponds to the time interval in which the physical Sizes occur one after the other in such a way that the measuring impulses in a monopolar, histogram-like manner running signal converted with the signal sections corresponding to the individual measuring pulses whose levels correspond to the amplitudes of the pulses causing the signal segment are proportional and remain constant for a period of time with a level, its Duration is equal to the time between the occurrence of the triggering measuring pulse and the occurrence of a following measuring pulse passes.

Although the invention can be used advantageously for a number of applications, but for the sake of greater clarity and ease of understanding in the following reference should primarily be made to a particular situation with which the invention relates can be evaluated in a preferred manner. This is the investigation of small ones Particles, d. H. Particles of microscopic size, in particular according to a principle as it is fundamentally is described in German patent specification 964 810.

According to this principle, the particles to be examined are in an electrolytic liquid suspended, and the suspension itself is passed through a passage opening, the one has a very small diameter, on the order of the particle size, but of course larger than this to allow the particles to pass through. Located on both sides of this passage opening electrodes that form the exit of a sensitive measuring section. Every time a particle dies Passing through opening changes the resistance of the between the two electrodes Suspension section. This change in resistance is recorded from the measuring section by means of a suitable display or recording device. There this. Measure of the change in resistance of the amount of liquid displaced by the respective particle and so that the volume of the particle is proportional, it is not only possible according to the principle mentioned, simply make a count of the particles contained in the suspension, but beyond As already indicated, the volume of each particle can also be determined.

The particles themselves can be substances of the most varied types, for example particles of a biological nature such as blood cells, bacteria or the like, or microscopic particles, as found in industrial applications Find sector and for example in the form of powder, dust grains, emulsions or muddy Crowds occur.

Especially when examining particles of the aforementioned type, which are all different Having volumes, it is often very important and of great importance to know how big the volume is of such particles is a mean, this mean value also being of particular interest because he L'leich / .eithj represents the 5 () "/ o point of a characteristic curve, with which the integral over a certain number of measured quantities, in the example, to the Particular reference is made here, i.e. particles are formed. Half of the total mass of the examined Particle is then above and the other half of the total mass below this 50 ° / o point. So far, the determination of the mean value, if it actually seemed justified, a corresponding effort for it at all

ίο drift, roughly made in such a way that, for example quite elementary the values or amounts of the individual quantities of interest for the assumed So case the volumes of the passing particles, measured, summed and then divided by the number of sizes considered. It is obvious that such a procedure is cumbersome, lengthy and, moreover, also relatively imprecise. On top of that there is not permitted to provide a continuous overview of the values applicable to a continuous flow of particles Win relationships.

By plotting suitable curves, the mean value of a plurality of quantities can be determined if necessary also determine by drawing, with essentially the same disadvantages in purchase take are the same as with the immediate calculation. Also from the possibility of data processing equipment can be used in principle. For such use, these devices must however, they can be very complex and are so costly that their use is made economical Considerations out usually must be disregarded.

Because of the awkwardness or costliness of the solutions described above is therefore also Attempts have already been made to create a compromise with so-called electronic "windows". In order to one saves a time-consuming examination and evaluation of every single variable during the »Window installation itself does not entail particularly high costs. Under the electronic "windows" themselves are discriminators with a pass band between a lower and an upper one To understand limit voltage, so that only pulses are allowed through, the amplitude of which is above the lower limit voltage, but below the upper limit voltage. With a plurality windows of this type, each of which has a pass band on a different level, can have more definite impulses Size groups from the totality of the measurement pulses generated by the passing particles sorted out and given on counter, whereby a certain overview of the size distribution can get.

Apart from the fact, however, that such a procedure is obviously a comparatively large one Uncertainty area can exist because the window always only a certain area, but not a discrete value can be eliminated, it remains in practical application in spite of a certain value Simplification compared to the determination methods mentioned above is still cumbersome and time consuming. This is due, among other things, to the fact that when setting the pass bands the individual discriminators at least some knowledge of the pulse sizes is required. on the other hand if the sizes are to be determined by the measurement, so that an initial

An investigation must be carried out on the basis of which the windows can then be adjusted, etc. In particular, it is of course unknown where the 50 ° / o point of the measurement lies, above and below whose permeability values of the windows would have to be adjusted. Even with that in mind So it is necessary, even if one is already in the right one with regard to the passbands It is on the order of magnitude to take repeated measurements over the entire size spectrum and then calculate the average value from the results obtained therefrom.

The way in which the identified problems are actually reflected in practice is a determining factor also emerges from the following very special example: In medicine it has in many cases been found to be proved very valuable, the mean volume, for example of blood cells, according to this Introduced English term to determine the mean cell volume or MCV for short, because it is derived from it Very informative information for the diagnosis of diseases and changes in their symptoms include in the size of blood cells, especially those related to the treatment of Anemia.

The classic method for determining the MCV uses the plasticity of the red blood cells which enables a sample with the blood cells to be examined to be centrifuged in a vessel, so that the blood corpuscles themselves are deposited at the bottom of the vessel, the plasma on the other hand upwards push away. So there is a certain stratification of blood cells and plasma, which too called a hematocrit. The relative height of these layers can be determined and related with this result, after counting the number of blood cells accumulated at the bottom of the vessel, The mean blood body volume or MCV can be determined in a further measuring method. However, given the cumbersome nature of the procedure, one was always more at first failed to determine and evaluate the MCV, although its value as a guide for Diagnosis was undisputed.

In connection with the above-mentioned principle according to German patent specification 964 810, the Determination of the MCV, however, found a new and increasing prevalence because of it especially the counting of blood cells with the required and desired accuracy and before all speed could be made.

Although the MCV method has recently been used to a greater extent, there are still problems that definitely need remedial action. So is z. B. the amount of blood required for centrifugation is comparatively large and is at least 50 mm ³ . If one wishes to carry out experiments on small animals, then in order to create the conditions for an exact measurement, ^{an amount of 50 mm 3 of} blood must be taken which could lead to such an animal bleeding to death. Another problem is that the dividing line between the plasma layer, on the one hand, and the blood cells, on the other hand, must be precisely determined, which in practice leads to difficulties because the dividing line is not razor-sharp but has a more flowing course.

The high demands on the accuracy when determining the MCV are mainly due to this due to the fact that the differences between normal and abnormal values of the MCV are very small, that on the other hand it is precisely these small differences that are decisive in order to achieve the desired ones Obtain information about the condition of the patient to be diagnosed.

You can see how big the need is on the one hand

ίο after a suitable and easily applicable Method is by which the mean value of a sequence of quantities can be determined, and what difficulties on the other hand oppose this. It is also not possible, for example, by means of the sizes the particles to generate electrical impulses by means of a suitable transducer, which would occur temporally in the same rhythm as the sizes themselves, i.e. also in theirs Amplitude would be proportional to the values or amounts of the quantities, and from this, for example through integration, to take the mean. If namely a chain of such at irregular intervals and with different amplitudes occurring electrical measuring pulses in one of the known Arrangements for determining the average or Mean value is introduced, this represents from such an arrangement delivered the average value of the total area over the Measuring time distributed pulses, so that the remaining between the occurrence of the individual pulses Time intervals - which in themselves have nothing to do with the mean value - the measurement result also affect. On the other hand, since these intervals are not known, such a measurement result can cannot be used to provide any indication of the extent of the mean To obtain amplitude. On the basis of this result it can only be said that if there is a Current represents that this would be the current that would have generated a charge during the measurement time, the would correspond to the sum of the partial charges generated by the individual pulses in the pulse chain. The measured value itself is, especially if it is at least in the particle investigation The usual case is that the impulses only follow one another at fairly large intervals, quite considerably smaller than the value of the amplitudes of the detected pulses.

It is known (W. Gruhle, »Electronic Aids des Physikers ”, 1960, pp. 141 to 145), a sequence of needle-shaped individual impulses, two antiparallel-connected To supply triodes, which are blocked in the idle state and to each reloading of a downstream storage capacity on the respective opened with a new value. The one applied to the storage capacity as a function of the amplitude of the input signals up and down increasing voltage in the form of a sequence of square-wave pulses is smoothed by a subsequent AC element.

In this way, an output voltage is created that is not constant over time and to a large extent is dependent on the signal amplitudes present at any given moment. A real Averaging in the sense of determining an average value from a series of individual Measuring pulses over a longer period of time cannot be realized in this way.

According to the invention, which is based on a

5 6

driving of the type mentioned, the invention has further set itself the goal of this problem is solved by the fact that the above-mentioned physical principles are based on the basic principle of the invention Size particles in a suspension so that it can also be used in such cases which allows an electrical scanning zone to pass through and leads to precise results, and generate said measuring pulses, and that 5 in which coincidence with a certain frequency the mean amplitude of signal sections which occurs.

make up the total time, is determined, so that at the end of another, an essential inventory of the measuring process a practically constant part of the invention forming embodiment Average value of the measured quantity is obtained. The present invention therefore provides the

Such a method actually makes it possible to continuously evaluate io individual signal sections of the histogram-like signal caused by the variables to be examined only at time intervals for measuring pulses and to allow them to follow one another, the intervals themselves directly increasing the mean value of their amplitude in each case by the occurrence of a new one Get the measurement without it being necessary to trigger the size impulses, so that the number of time intervals between them to 15 intervals know directly from the number of impulses or otherwise take into account. The display depends, while the length of the intervals in a self becomes independent of the number of during a certain percentage ratio to the measure of the pulses generated during a certain period of time. Amplitudes of the measurement pulses can be brought, which has the advantage that the measurement result can be expected regardless of the situation is.
from the concentration the suspension containing the particles can be obtained. are to be used below on the basis of execution

With the method according to the invention it will be explained examples in connection with the drawing

z. B. also possible, the pulses of the measured value wall 25 are. In it shows

lers, for example by means of a tape, to be stored F i g. 1, in a greatly simplified form, a graphical representation, and at any time and at a different display of the location on which the invention is based, the information on the tape according to the invention, which simultaneously illustrates how to evaluate. Furthermore, by means of the invention, it is basically possible to obtain from the measurement pulses provided by a suitable measurement device that not only the 50% - 30 value converter supplied measurement pulses artificially a point, but also a correspondingly different process histogram-like signal is obtained,
tual values can be determined. F i g. 2 in the form of a block diagram an output

A particularly favorable device for carrying out a device according to the invention for carrying out the method according to the invention is characterized according to the invention in that 35 successive variables,
a transducer for generating the to F i g. 3 shows a time diagram of the course in the electrical arrangement corresponding to the variables to be examined according to FIG. 2 occurring pulse trains measuring pulses one to hold a certain and the derivation of the histogram-like signal electrical charge on one of the amplitude of the resulting,

Measuring impulses corresponding level suitable Ge 40 F i g. 3 a is an isolated representation of the in FIG. 3 memory level is connected and that an electrolytic histogram-like signal,
Irish switching device to change the level F i g. 4 schematically shows the structure of a special transducer depending on the charge of the memory stage for generating electrical measurement of a new measuring pulse to a new level, which is proportional to the amplitude of the new pulse as a function of this measured value, 45 transducer-flowing particles,
with its input to the output of the measured value F i g. 5 in the form of a block diagram a counter-converter and with its output to the input via F i g. 2 a slightly modified embodiment of the memory stage connected and the output of a device according to the invention,
the memory level to the input of a suitable F i g. 6 shows a timing diagram of the arrangement for forming the mean value of 50 according to FIG. 5 occurring impulses as well as the memory level detected charges connected line of the desired histogram-like signal sen is. from it,

The already mentioned method and the F i g. 7 shows the circuit diagram in a greatly simplified form

Devices particularly suitable for implementation are one in connection with the structure according to FIG. 5

always then without special additional measures 55 related electronic switch,

can be used if the variables to be examined only F i g. 8 in the form of a block diagram a further

occur one after the other. In practice, there results tere embodiment of the invention with which it

However, there is also often the situation that it is possible to obtain mean values for which the

Variables occur at the same time or the effect of coincidences is already taken into account.

at least their appearance overlaps in time, 60 is shown, and which also has an integrating device

d. that is, coincidence occurs. owns to during a certain period of time

Especially to store the charges of the signal occurring through the narrow passage opening,

guided particles, coincidence occurs proportionally F i g. 9 a further in the form of a block diagram

often on, so that the amplitudes of the modified embodiment of the invention, the

Simultaneously passing particles generated pulses 65 essentially of the structure according to FIG. 8

superimpose and thus corresponds to a falsified image, but no special switching elements

deliver. for the compensation of the coincidence

With regard to these special conditions has a called influence on the measurement results,

Fig. 10 is a timing diagram of the various in the structure according to FIG. 9 occurring pulse sequences, which also shows how the desired histogram-like signal is formed,

11 partially in the form of circuit diagrams and partly in the form of block diagrams, a practical embodiment of the invention, their Structure basically F i g. 8 corresponds.

Before the invention is explained in detail in conjunction with the figures of the accompanying drawings is in view of the clearest possible understanding of the invention and what is under some expressions used in the description should be understood, first of all, the following to send in advance:

It was already stated at the beginning that, albeit predominantly in the context of the description reference is made to the investigation of particles, the invention generally to the Determination of the mean value of temporally successive quantities is simply used, as long as only the sizes, which in practice with suitable transducers at least in a Many cases can be converted into suitable electrical impulses, their amplitudes the amounts of the quantities and their intervals correspond to the time differences with which the quantities successive. In the special application of the investigation of particles, there is z. B. to determine the mean value of the "amount" of the particles, where "amount" is diameter, Cross-section or volume can mean. If furthermore from occurring in "any" sequence We are talking about impulses, it should be expressed that the occurrence of impulses does not obey any particular law, knowledge of which may otherwise result in a calculation of the Average value would also allow according to the previously known methods. The irregularity with which the pulses occur as well as the fact that it is of no importance for the invention, the Knowing the size of the time intervals between the pulses is a decisive factor Aspect of the invention.

If a "monopolar" signal is mentioned, it is meant that it is a signal acts in which amplitude changes only to one side of a reference line, which in general will correspond to zero potential. In the present context, however, under "histogram-like" is to understand that a signal consists of a sequence of signal sections which essentially the shape of a rectangle, i.e. a constant amplitude over a certain period of time have and either directly to one another or at least only slightly with the interposition of them Connect time intervals to one another.

The simplest example of such a histogram-like and also monopolar signal would be So for example a current or voltage curve that maintains its amplitude for a certain time, then within the time zero this amplitude is reduced or increased by a certain amount, thus suddenly assumes a new second amplitude value and then during a subsequent one certain second time segment maintains the new second amplitude value. After expiration of the second time segment, the new, second amplitude value becomes in the same way as beforehand, the initial amplitude value within the time zero to a further third amplitude value reduced or enlarged, which in turn is retained until the next sudden change remains, etc., so that overall a terrace-like offset amplitude contour results.

Following these general statements, reference is made to the drawings themselves. Included serves F i g. 1 first and foremost, the underlying theory is a little closer in principle to explain and thus to further simplify the understanding of the invention, with the in context However, the theory dealt with in no way restricts the invention should represent.

With the in F i g. 1, pulses P ₁ , P ₉ , P _3,. . .P ₁₃ is

a chain of impulses that follow one another at any time. The electrical quantity to which these pulses correspond can be an electrical voltage or also an electrical current, which is plotted over the time axis 10. If one assumes that the pulses P ₁ , P ₂ ... by means of a device for examining particles, as shown in FIG. 4 is shown schematically, their amplitudes are proportional to the "amount" of the examined particles, in the present case their volume. As far as other coincidences are concerned, for the sake of simplicity it is assumed in the following that such cases do not occur in order not to unnecessarily complicate the description of the invention.

Furthermore - which is also in accordance with the actual circumstances - the assumption is made that the duration of the occurrence of the pulses P ₁ , P ₂ ... in relation to the time between the occurrence

If two successive pulses P ₁ , P ₂ ... occur, the time that passes is so short that the width of the pulses P ₁ , P ₂ ... is neglected for the present consideration, ie it can be set equal to zero. This assumption is made e.g. B. with regard to the particle investigation, the more the more dilute the suspension containing the particles is,

d. H. that is, the fewer particles there are per unit volume of the electrolytic carrier in it.

Let the amplitudes of the pulses P ₁ , P ₂ ... be with

I _1, I ₂ ... referred, such that, taking into account measurements N or N variables for the average of all measured amplitudes

· _ ^Llunh

where i _{(Iliri; h} would correspond to the amplitude of the line 12 shown in FIG. 1 in dashed lines.
The difficulty of being able to automatically carry out a calculation process corresponding to relation (1) by means of an electrical circuit arrangement or otherwise with a device, and the fact that it has not yet been possible to develop such devices, is primarily due to the fact that the individual summands Z ₁ , I ₂ etc. do not all exist at the same time and are therefore not suitable for evaluation using conventional methods for adding currents. Generated

109 536/63

However, if a histogram-like signal is obtained by means of the pulses, this difficulty can be overcome by simply determining the mean value of this histogram-like signal and thus actually _{obtaining the value / through} as a direct measurement result without the need for intermediate calculations.

It can be seen in Fig. 1 that the individual signal sections of the histogram-like signal 18 have the levels L ₁ , L ₂ , L ₃ ..., which each have the amplitudes Z ₁ , i ₀ , z ₃ ... correspond to the _{pulses F 1} , F ₂ , F ₃ ... relevant for the individual signal sections. The change from one level to the other takes place suddenly, so that the connection

lines between the individual levels L ₁ , L ₂ , L ₃ ... have an infinitely steep rise, which can also be achieved in practice with suitable circuit elements in a largely approximate manner.

Once the histogram-like signal has been formed, which is designated generally by 18, the mean value can be determined. The average value of the current i _dmch , as indicated by the dashed line 12, can be referred to as the equivalent direct current which, if it were to _{flow up to the time i 13} , would deliver the same amount of electrical charge as the during the time i ₁₃ Sum of all currents Z ₁ to Z ₁₃ . This latter sum is

q =

z, (f ₂ - t _t ) + Z ₃ (i ₃ -t ₂ ) - \ z

_13th

(2)

The average value of the current z _through would then be equal to this sum divided by the time f ₁₃ or

'through

(3)

'13

q = Sidt, (4)

represents each of the addend / "(t _n -t _n-1) a charge contribution to the total amount of electric charge is. If one has, for example, a thousand pulses F _n with the amplitude /", which follow one another at any time intervals, the length of a time interval between successive pulses on average

'through

T _n N

(5)

where T _n corresponds to the time that has elapsed between the occurrence of the initial pulse and the iV-th pulse, while N, as in (1), denotes the number of pulses that appear during the time segment T _n . The average charge q is thus

ⁿ ih woe + "2 h ^f through +""" ¹ Vn ^ through>. (6)

where ZJ _{1 is} the number of pulses of a certain size group and i _{t is} the current corresponding to this amplitude, while n ₂ and z _{2 are the} number and current of the next size group, etc.

With these prerequisites, the average current level -12 of FIG. 1 is obtained

m im ^ through

'through -

m = l

• (7)

In order to form the histogram-like signal 18 in practice by means of the pulses F ₁ , F ₂ , F ₃ ..., One can use a simple circuit as shown in FIG. 2 is shown as a block diagram. F i g. 3 shows in comparison to FIG. 1 shows, on an enlarged scale, a sequence of pulses and processes taking place in connection therewith, as they are during the operation of the arrangement of FIGS. 2 occur.

In the block diagram of FIG. 2 the pulse source is formed by a transducer 22, which in practice can be a device for examining particles, as shown in FIG. 4 is shown schematically. This device has a collecting vessel 24 in which a part of the sample, d. H. that is, electrolytic liquid with particles suspended therein, is located as the amount of liquid 26. A tube 28 with a passage opening 30 is immersed in the vessel 24, the passage opening 30 is below the level of the amount of liquid 26. Between an electrode 32, which is located inside the tube 28, and one located outside the tube 28, into the vessel 24 immersed electrode 34, both of which are connected via conductors 40, 42 to a suitable power source 38 (FIG. 2) are, due to the electrolytic properties of the sample liquid, a current flow. The current source 38 ensures that the current between the electrodes 32 and 34 is constant remains as long as no particles migrate through the opening 30 and that, on the other hand, as soon as a particle passes through the opening 30, signal pulses can be generated which by means of a suitable, The measuring section connected to the electrodes 32, 34 can also be detected.

In the block diagram of FIG. 2, conductors 40 and 42 are shown only as channels 40, 42 in order to keep the circuit diagram simple. Channel 40 represents the connection of the transducer 22 to the current source 38, channel 42 the connection of the output of the transducer 22 to the input of an amplifier 50 (or a pulse height discriminator, etc.), the current source 38 and the amplifier 50 (or Discriminator) can be combined in a single housing. The in F i g. 1, pulses F ₁ , F ₂ ... then appear in relation to the block diagram of FIG. 2 at the output 52 of the amplifier 50.

During operation of the device, as shown in FIG. 4 As can be seen, a vacuum source can be connected to the tube 28 via a valve 54. When opening the Valve 54 is then a certain proportion of the amount of liquid 26 through the vacuum Opening 30 sucked into the interior of the tube 28. If necessary, an exact amount can be provided by the shown, known per se siphon arrangement are measured. In the U-shaped section between the pipe socket 58 on the one hand and the pipe section 60 on the other hand is a grounded at 68 Mercury column 56. In the piece 60, the pipe cross-section is only capillary-like and with a start contact 62 and a stop contact

11 12

64 provided. When the vacuum on the cavity signals via connection 96 a univibrator 94

of the tube 28 acts, it not only sucks liquid triggers. The univibrator 94 provides a sharp one

speed 26 in the pipe, but also pulls the narrow pulse 98, whose width t _mv by the

Mercury column 56 from its equilibrium position in the selection of the circuit constants of the univibrator 94

the area of the pipe extension 58 into it. After 5 it is set to a value that is much smaller

Closure of valve 54 reverses the mercury level as the average between occurrences

column 56 back into its equilibrium position. Two successive impulses pass away

it first reaches the start contact 62, whereby time t _through .

the counter connected to lines 40, 42. The output of the univibrator 94 is via the circuit is switched on. When stopping io channel 102 to a normally open electrical contact 64 the counting circuit is stopped. In this way, ironic switch 100 can be connected and the same achieve that in each case a very specific amount of liquid at the rear via a channel 104 is being investigated. detector 106, a further channel 106 and a

After in connection with Fig. 4 a prak- short-circuiting stage 110 via the channel 112 to the Table example of a transducer shown 15 memory stage 78 switched back. Purpose of the as it is preferred for the invention in question last described arrangement, by means of the comes, are now with F i g. 3 the further individual univibrator impulses 98 one in a second unit with respect to the memory stage 114 provided by the transducer 22 capacitor generated measurement pulses as well as the processing of this one channel 116 and the said electronic Measurement pulses in the arrangement of FIG. 2 switches 100 directly to the output of the wrote. first repeater too. connect and at the same time

With 71, 72 and 73 three positive voltage the memory stage 114 inevitably referred to the chip pulse, which at the point 52 of the FIG. 2, the voltage level of the output 84 of the first amplifier, i.e. practically to increase or decrease at the output of the measured value 80,
represent the signals occurring in the converter. If, when the pulse 71 occurs, the base line of the slide pulse 98 shown in FIG. 3, the electronic switch 100 in the grams should correspond to zero potential, so that the conductive state is transferred so that the memory pulses have only one polarity or, in order to connect stage 114 abruptly with the output 84 of the Verden term already introduced above to amplifier 80, the capacitor will use, are monopolar. The amplitudes of this pulse of the memory stage 114 caused by an earlier pulse are _{denoted by V 1} , V ₂ and V ₃ and are those in FIG. 3 pulse, not shown, has already increased to a volume of the three particles that the opening 30 pas value v _{p was} charged, now to one of the Amsiert and thereby generated the pulses 71, 72, 73, amplitude V ₁ corresponding level raised in each case proportional. The pulses themselves will then generally be much further apart from one another by the capacitor of the memory stage 114 than is kept. The output voltage of what is thought in the drawing appears to be, but for nis stage 114 is fed via a channel 118 into a second intermediate amplifier 120 which is illustrated in the present explanation from the drawing, the output signals of which appear at 122 in comparatively close succession. As far as has been asked so far. described, the signal 122 provides the horizontal

A memory stage 78 with a diode, a level L ₁ , the capacitor following the multivibrator pulse 98 and an amplifier with an extremely high rising edge R ₁ as well as the level L ₂ , which holds the input impedance one after the other, _{corresponds to the peak voltage V 1 of} the pulse 71,
As already mentioned, the univibrator pulse will correspond to the pulses arriving at point 52. Also in a trailing edge detector 106 a-The amplifier of the memory stage is fed as an intermediate 45, which in connection with the short amplifier 80 referred to and receives its signals closing stage 110 the capacitor of the first Gevof the memory stage 78 via the channel 82. The memory stage 78 are short-circuited and the output voltage of the amplifier occurring at 84 can accordingly discharge to zero. Here, kers 80 is shown in FIG. 3 is shown as dashed line 86 (86 ', the discharge follows the falling, dashed-86 "). This dashed line 86 follows 50 line 86'. line 76 has reached, the memory level 78 is devalued V ₁ and then remains on this load and for the storage of a new pulse level V ₁ , while the pulse 71 itself prepares again for 72.

the zero level 76 drops. The high input resistance This next pulse 72 has an amplitude V ₂ , the amplifier 80 is required to have a 55, which in the example assumed here is quite a bit

Discharge of the capacitor of the memory stage 78 is smaller than the amplitude V _{1 of} the previous one

to prevent the impulse applied by the impulse 71. The memory level 78 follows - as as

Saves charge. Continuation of the dashed line 86 'shown -

At the same time that this stores the pulse 72 back up to its peak value. Amplitude V ₁ takes place, the pulse 71 also acts 60 While the pulse 72 decays again, however, holds

Via channel 88 to a limiter level 87 with the memory level 78 the maximum voltage at 86 '

comparatively low limit voltage, as it is fixed until the trailing edge of pulse 72 exceeds the limit

indicated by the horizontal line v _th , a. voltage v _jh and the univibrator 94 of

As soon as the trailing edge of the pulse 71 triggers the level again. At this point in time, the span of this limit voltage has been reached, and at 90 the second memory stage 114 is recessed, still V ₁ , and

output signal with which a trailing edge detector if now their capacitor suddenly with the

92 is applied, which is connected by means of the rear edge amplifier 80, the voltage of which is

of the output supplied by the limiter stage at a level corresponding to _{the value v 2}

finds, then the charge level of the memory stage 114 and thus the voltage of the amplifier 120 falls in accordance with the line i? "of FIG. 3 down to level L ₃ , which is retained until the next pulse arrives. The next pulse 73 has a peak voltage V ₃ , and in the same way as previously described, the level L ₃ then rises along the curve segment R ₃ to the level L ₄ , and so on.

As a result, the monopolar, histogram-like signal 18 is obtained in the desired manner of Fig. 1, which is shown separately in Fig. 3a would. In the block diagram of FIG. 2, this histogram-like signal 18 is on the channel 122 as the output signal of the amplifier 120.

The in F i g. 3 time intervals denoted by t ", t"', t _n " and t _n '" , during which the individual levels L ₁ , L ₂ , L ₃ are retained, are essentially the same as the time it takes from a certain point of an impulse passes to the corresponding point of a following impulse. At the F i g. 2 corresponding arrangement lasts, as shown in FIG. 3 shows the time interval from the moment when the trailing edge of a pulse reaches the level of the limit voltage v _th to the moment when the trailing edge of the following pulse again reaches the same limit voltage v _th . However, one is not bound to choose precisely these particular points of the impulses as reference points for the running time of an interval. For example, in an arrangement to be described below and representing a modified exemplary embodiment of the invention, the peak values of the individual pulses are selected as reference points for the intervals of the individual levels L ₁ , L ₂ , L ₃ .

The histogram-like signal 18 itself can then simply be passed through a low-pass filter 124 and then be evaluated by an arrangement that is suitable for determining the mean value. In Fig. 2, a moving-coil galvanometer 126 is provided for averaging, which measures the mean value of the Forms current, which the histogram-like signal 18 allows to flow in a suitable resistor. Means a schematically indicated potentiometer 128, the instrument 126 can be calibrated so that the mean amplitude 12 can be read immediately. Provided that they are supplied by the transducer Measurement impulses are proportional to the volume of the examined particles, corresponds to that on the Instrument 126 read amplitude value already the mean volume of the particles.

Instead of a moving-coil instrument 126, an IC element could also be used to calculate the mean value of the signal, and then the output value of this i? C element into an electronic Display device or a servo system with a dial driven thereby or with a feed in another scale calibrated in volume units. The inertia of the moving coil instrument or the time constant of an electronic filter circuit can, however, possibly be too large to to make the particular arrangement chosen sensitive enough so that fluctuations in the average value the sizes of an examined series of measurements are not recorded, or too much time passes before after the start of a measurement, a steady state or a state of rest can be achieved. on the other hand can reduce the time constantc of the device for detecting the mean value to one Vibrate the pointer lead. In such cases it has proven to be advantageous to use the signal 18 to store in an integrating stage, for example, has a large capacitor that is in the Is able to hold on to an electrical charge over a period of several seconds, and thereby the histogram-like signal into the integration stage for a precisely defined period of time

ίο to be fed. The resulting reading delivers, if it is calibrated so that the time interval is also taken into account, a definitive one Value for the searched mean amplitude.

In Fig. 5 shows a further device according to the invention for carrying out the method on which the invention is based, in the form of a block diagram. The main difference compared to the arrangement of FIG. 2 consists in the fact that the start and end points of the intervals during which a level L ₁ , L ₂ ... is kept constant are determined here in a somewhat different way.

The measuring transducer 130 can be of any type. At its output 132 there are again pulses released, the with fed into the transducer 130 and to be examined variables with regard to Time and "amount" correspond. An amplifier 134 with the output 136 amplifies the measuring pulses, without, however, fundamentally in terms of their size or time proportionality to the sizes examined to change something.

In Fig. 6 is the output voltage of the amplifier 134 again indicated by three pulses 71, 72 and 73, which for the sake of simplicity are in the same Were represented like the impulses of the F i g. 3 and accordingly also the same reference numerals carry.

The output 136 is here via an electronic switch 142 and a channel 146 with a Memory stage 144 connected, the output 148 of which is in turn connected to an amplifier 150 with a high input resistance connected. In addition, the structure of the arrangement of the F i g. 5 no further from that of FIG. 2, so that again the same reference numerals for the low-pass filter 124 etc. were used.

The electronic switch 142 is normally in the open state, so that of the Amplifier 134 cannot get a signal to the memory stage 144 before the electronic one Switch 142 is brought into its closed state. The output 136 is also about a channel 152 connected to a limiter stage 154, with the help of which the baseline of the pulses is increased beyond the amplitudes of a noise voltage 140. That way will prevents the univibrator 156 from being exposed to noise pulses or other glitches from the examination should remain excluded, is initiated.

The protective measures mentioned with regard to the noise or interference voltage 140 are here - in In contrast to FIG. 2 - required because as a criterion for the intervals, as is to be shown in a moment, not points of the trailing edges but rather the peaks of the measuring pulses are evaluated. In this respect, could without the protective measures taken, it may happen that the apex as well

a glitch has the same effect during the evaluation how the apex of a measuring pulse would exert, which of course is undesirable.

The impulse 71 thus appears with its portion protruding above the level determined by the limiter stage 154 as a new impulse 158 which starts from the level of the limit voltage v _th. The signal 158 is in FIG. 6 is shown by means of a dashed line, which follows the actual course of the pulses 71, 72, 73 above the limit voltage v ". As soon as the signal 158 reaches its peak value, normally at the highest point of the pulse 71 (72, 73), a switching pulse is generated here by means of a peak value detector 160 - instead of the trailing edge detector 92 as before - which is sent via the channel 162 to a univibrator 156 got. The univibrator 156 then responds and delivers a narrow pulse at its output 164, which is designated by 164 'in FIG. 6 and lies within the large measuring pulses 71, 72, 73 in each case. This narrow pulse now converts the electronic switch 142 into its conductive state, so that the memory stage 144 is connected to the amplifier 134 during the duration of the pulse 164 '. Since this connection takes place at the point in time at which the pulses 71, 72, 73 reach their respective peak values, the capacitor of the memory stage 144 is charged (or discharged) to the level of the peak value of a pulse appearing at the output 136. The time constants are chosen so that the capacitor is recharged from its old to the new level within a time in which the univibrator keeps the electronic switch 142 closed.

In Fig. 6 it can be seen that an earlier, not shown pulse has brought the potential of the capacitor of the memory stage 144 to the level L ₁ . When the electronic switch 142 closes, the voltage of the memory stage 144 rises according to the line, R ₁ to the level L ₂ , the height of which corresponds essentially to the peak amplitude of the pulse 71. Subsequently, as a result of the occurrence of the following pulse 72, the level L ₀ falls back to the level L ₃ according to the line R ₉ , the height of which in turn corresponds to the peak amplitude of the pulse 72. The pulse 73 lets the signal 18 along the line 2? ₃ rise again to a level L ₄ , etc.

The histogram-like signal 18 is then available again at the output 122 of the intermediate amplifier 150 . The time during which a level L ₁ , L, is maintained is in turn equal to the distance between a certain point of a first pulse and a corresponding point of a subsequent pulse, in which case these points are the apexes of the pulses acts. The time intervals t ", t"', t "" ... are therefore essentially the same in each case in FIGS. 3 and 6, only their starting and end points, based on the same pulse chain, are somewhat offset from one another in both cases, which, however, is of no significance in the context of the present invention . ■■ '.' ·.

It goes without saying that every suitable point of an impulse can be used as the beginning or end of a Zeitihtervalls can be used.

To further explain the invention is. FIG. 7 shows another possible circuit structure of the electronic switch 142 of FIG. 5. The nature of the components used and the type of their mutual connection is clearly evident from FIG. 7 and in this respect does not require any further explanation. When the narrow pulse 164 'of the univibrator 156 acts on the base of the transistor Tr ₁ , this transistor changes to the conductive state, so that current can flow through its emitter-collector branch. The

ίο in turn leads to the fact that the base-emitter circuit of the normally non-conductive transistor Tr.- is conductive, so that during this time current from the terminal 136 via the diodes D ₁ , D ₃ to the terminal 146 or vice versa from the Terminal 146 can flow to terminal 136 via diodes Z) ₂ , D ₄ , depending on which terminal is at the higher potential. As soon as the transistor Tr ₉ stops conducting because the signal at the terminal 146 has disappeared again, the terminals 136 and 146

zo info'ge the polarity of the diodes D ₁ , D ₂ , D ₃ , D ₄ again separated from each other.

Both the arrangement according to FIG. 2 as well as that according to FIG. 5 can be made out of circuits be constructed of known design. For the sake of completeness, however, the following is a brief one Description of at least one possible implementation of the different, each one block of the Block diagrams corresponding to FIGS. 2, 5 Circuits given:

In Fig. 2, the limiter stage 87 can be any circuit that has an output voltage if the level of their input voltage exceeds a predetermined value, but no output voltage or the output voltage is zero if the level of its input voltage is less than said predetermined value, such as in the case of an overloaded differential amplifier or with a Schmitt trigger circuit.

The trailing edge detector 92 can be a differentiating network, in the simplest case a capacitor followed by a resistor in series, the time constant of this RC combination being small in relation to the duration of the pulses involved. A diode ensures that output pulses of undesired polarity are rejected.

The univibrator 94 can be any circuit which, when subjected to an input switching pulse on the output side emits a pulse of predetermined amplitude, duration and polarity.

The short-circuit 110 can be formed by a transistor whose base bias is selected so that it normally blocks, but which can then be brought into saturation and thus into the conductive state by means of an input pulse acting on its base.

The memory or pulse stretching stages 78 and 114 or 144 can be formed by a diode and a capacitor which are connected to one another that the capacitor is coupled to the input of the memory stage and is accordingly charged by the input signal when the size of the Signal increases, which, however, is separated from the input when the size of the signal decreases.

The amplifiers 80 and 120 can in both cases be formed by a simple, resistor-coupled amplifier with a high input impedance,

109 536/63

which does not load the feeding stage, but can follow its output voltage.

The limiter stage 154 in FIG. 5 one can move quickly be regenerative amplifier whose bias is chosen so that it is normally locked State.

As the peak detector 160 of FIG. 5 a differentiator can be used, which each time if the differentiated signal goes negative and thereby reaches the value zero, an output signal supplies.

The previous considerations had always assumed that the Examination of the sizes no coincidences occur, d. i.e. that at a certain point in time always only one single variable triggers an impulse and that a further impulse is only generated then, when the pulse generated by the previous magnitude has already completely died down. In the In practice, however, coincidences occur in many situations with too great a frequency for you Influence in the evaluation of the measurement result could always be neglected. As in the beginning already indicated, the occurrence of coincidence leads to a reduction in the number of pulses counted, d. That is, fewer pulses are detected in the count than actually occurred. On the other hand, the coincidence leads to an increase in the amplitudes of the coincidence situation immediately involved pulses, so that the value of the average amplitude is greater than it corresponds to the actual circumstances. The present invention is concerned with the actual Find the average value or average amount of a particular system. In cases where for whom the influence of coincidence cannot be neglected, it is therefore necessary that measurement result obtained by the above-described simplified embodiments of the invention reduce to a certain extent to the desired actual mean in place of one To obtain a value that is too large for a proportion due to coincidence.

As already indicated at the beginning, this is achieved according to the invention in that the histogram-like signal is loaded and thus _{a certain amount of charge is removed from each signal section with a level L 1} , L ₂ ... of constant amplitude, the value of which depends on the effective volume of the The touch zone of the transducer passed through the examined variables and the number of pulses generated depends. As will be shown in more detail below, the load is carried out during a specific and always the same time interval, namely once per time interval or per signal section, but the amount of the load is always in a specific percentage ratio to the height of the individual levels and is therefore proportional to the amplitudes of the measuring pulses or the corresponding levels of the histogram.

This type of load represents a fairly simplified measure, but should be practical this way all due to the change in the average amplitude of the histogram-like signal Coincidence factors involved are appropriately taken into account. The following is an attempt will be able to demonstrate the correctness of this view to a certain extent.

In doing so, certain assumptions have to be made. So it is initially assumed that all pulses have the same amplitudes and that the amplitude of coincidence pulses is in each case on the Double of a single pulse grows, which means that the cases are not taken into account in which to more than two impulses or two at the same time

ίο pulses only during a fraction of their respective Duration occur together. During the investigation, the amount of electrical charge is used as a reference value based on that of the signal during a certain predetermined period of time is transmitted. The histogram-like signal is accordingly as a current signal or as Consider charge per time unit signal.

With these simplifying assumptions, the result for the signal during the predetermined Time total amount of charge delivered the printout

q = 2i- (T / N _R ) N _c + i (T / N _R ) ■ (N _R - AQ. (8) where (

q the total charge,

i is the amplitude of a single current pulse

(with the assumption that all impulses are equal),

T is the total time of the measurement,
N _R the uncorrected, when counting the
Sizes received numerical value or the raw

count,

N _{c is} the number of double pulses
(Coincidence pulses).

According to expression (8), the total amount of charge q is made up of a portion of the charge due to the double pulses and a further portion due to the single pulses. With the above definitions, N,

ie the actual number of examined sizes (particles) is equal to the sum N _c + N _R , or N _R is equal to N - N _c . This results in the simple relationship for expression (8):

q = TiNIN _R. (9)

As already stated, simplifying assumptions were made here, which still deviate considerably from the actual circumstances. The consideration of some other factors besides 2 could lead to better approximations in order to be able to estimate the proportion of the impulses caused by coincidence. If one does not commit oneself to a certain numerical factor 2 as above, but rather generally starts from a factor which represents the increase in the pulse amplitudes as a result of coincidence, then one obtains for expression (8) if one simultaneously runs the current / through expresses the ratio of voltage e to resistance R ,

q = uelR · T / N _R ■ N _c + elR ■ T / N _R ■ (N _R - AQ.

(10)

Thereafter, the total amount of charge transferred during the measurement period T is again made up of a portion that is caused by the coincidence pulses and a further portion that goes back to the individual pulses. Each of these two proportions depends on the intensity of the

individual impulses. The portion caused by the coincidence pulses is it times the Partial amount of a single average quantity times the average duration of a level times the number the frequency of occurrence of such impulses. The part that goes back to the individual impulses is on the other hand, equal to the partial amount of a single average variable times the average duration of a Levels times the difference between the raw count and the frequency of coincidence.

Expression (10) can be further simplified

T- [l + KN _R (ul)], (11)

where K is a constant which can be determined in a statistical way through series of tests for a certain type of transducer, with the transducer again being predominantly connected to a device for examining particles according to FIG. 4 is intended. In this case, the constant is actually by a known expression

F ₁ . = KN

(12)

defines where F _{c is} a correction factor proportional to the raw count. Expression (12) is used when examining particles in order to take into account the influence of the coincidence on the measurement result and to be able to correct the result itself accordingly.

Expression (11) was obtained by deriving a new value for N _c from the ratio of the actual numerical value N to the other values. It is

in addition, however, N = N _R + N _c also applies, so that results for N _c

N _c = KN _R *. (13)

Using this expression in (10) results in relation (11).

According to the relation (11), the sets during a certain measuring time or a certain Time unit transferred total charge now from a representing the actual conditions, correct portion and an error portion

KN _R (ul)

together, the error portion being determined by the coincident pulses and by the loss of counting that occurred as a result of the coincidence. Compensation for errors caused by coincidence must therefore be able to be achieved by separating an amount corresponding to the error component from the signal.
If, accordingly, a certain amount of charge is derived from it in the desired manner, q '= e / RT

must, as can be seen by rearranging relationship (14), apply

i _c = TK (ul).

(15)

q _c = elR-t _c

is destroyed per pulse, where t _c corresponds to the duration during which the charge is destroyed, then during a certain time T thus N _R such charge components or a total amount of charge N _R · e / R · t _c Coulombs are destroyed. The remaining amount of charge is then

q ' = _L. Γ [1 + KN _R (u - I)] -N _R ■ e / R ■ t _c . R.

(14) This means, however, that the duration during which charge is destroyed by loading the signal must, is proportional to the constant («- 1) for each impulse, i.e. is proportional to the amount, by which an average coincidence pulse exceeds a single pulse. So there is one linear equation before so that the duration of the charge practically in a simple manner, such as by setting a potentiometer in a timer circuit.

As far as the amount of the load or its percentage of the total load is concerned, information about this can also be obtained from the considerations made above, and in itself a somewhat rough approximation, but in principle it does justice to the circumstances it can be said

Total percent load = N _R K (u - 1) x 100.

(16)

The essence of this derived relationship is to be seen in the fact that, according to this, the percentage load of the raw count N _{R is} proportional. Here, too, the value K represents a constant that depends on the special features of the transducer and is known for the particular application. The expression («- 1) corresponds to the amount by which the average coincidence pulse exceeds a single pulse and corresponds to a fixed number. Since this equates to a percentage, the charge fraction destroyed in a signal section also clearly represents a certain fixed percentage of the charge contribution provided by a pulse.

The load on the overall signal thus represents a fixed percentage of this signal, and each charge contribution corresponding to a pulse can be reduced by a fixed percentage, so that in this way compensation for the influences on the signal caused by coincidence can be achieved. Since the amount of charge destroyed is proportional to the raw count N _R , a certain amount of charge can be subtracted from each increase in charge caused by a pulse. The actual percentage can be determined on the basis of the constant K and by varying the circuit constants until changes in the concentration of the suspension carrying the particles have little or no influence on the average amplitudes determined by the measuring arrangement.

The discharge or destruction of charge can, as is evident from what has been said so far, on done different ways. With the one sent ahead Consideration was z. B. assume that the current for a certain duration of each signal section cut from the signal down to the baseline so that the amplitude of the cut out part was proportional to the amplitudes of the affected pulse parts. Instead of

Parts of this could also be cut out of the signal, not necessarily from the vertex of the impulse must reach its base line and their amounts do not necessarily apply to everyone Impulse must be the same. According to an exemplary embodiment of the invention described below a continuous reduction of the signal z. B. made that initially one of the perfect shape such as that shown in FIG. 3 a, a slightly different histogram is generated where the deviations correspond to the desired charge loss and where, however, additionally charge portions were cut out in order to satisfy the condition that a change the concentration only to negligible changes in the output of the measuring device delivered measured value should lead.

In Fig. 8 is a block diagram of a practical embodiment of a device according to the invention reproduced, with the help of which the histogram-like signal are additionally loaded and so the influence the coincidence can be compensated according to the previous considerations. The main kitchen Difference between this device and the one shown in FIGS. 2, 5 therefore consists in that a special load level 200 is provided here, which, controlled by the univibrator 156, the output signal of the amplifier ISO is loaded for a fraction of each signal segment. The so The signal sections obtained are stored in the integrating stage 202 via the “And” gate 204, with the gate 204 in turn is controlled by a start signal via a control circuit 206 and a timer 208 is. The triggering of the start signal is usually due to the mode of operation of the transducer 130 influenced. The integrating stage 202 stores charge fed in for a predetermined time and then delivers the accumulated amount of charge to the meter 126. After completing a Measurement can be carried out by the integration stage 202 by hand a line 210 is discharged and thus again prepared for a new measuring process.

A comparison of the block diagrams according to FIGS. 2 and 5 on the one hand and according to FIG. 8 on the other hand shows that both arrangements are largely similar, but that they are not only differ with regard to load level 200. The arrangement according to FIG. 8 is missing electronic switch 100 (Fig. 2) or 142 (Fig. 5), with the help of which the level of the output signal could be retained in the downstream memory level. Accordingly, the Voltage at the output of the intermediate amplifier 150 drop before a new pulse arrives, so that the histogram of this arrangement actually loses a certain amount of charge per pulse, at the same time, however, at another point with regard to certain impulses again about a certain one Amount increases, but the losses are greater than the gains. Thus it can be in a very simple way and economically compensate the influence of the coincidence in any case in a rough approximation. As the amount of charge that, apart from the loss as a result of the deviation of the histogram is separated from the ideal shape, additionally by the load level, the The course of the histogram obtained with the circuit structure according to FIG. 8 is also determined the explanation first on the basis of the circuit according to FIG. 9 made, which has a stress level of 200 is missing. As in FIG. 8, the Arrangement of the F i g. 9 an electronic switch.

FIG. 10 again shows a pulse train 71, 72, 73 which corresponds to the pulse train of FIG. 3 and 6 is essentially identical, so that the relationships can be easily compared. The mode of operation of the arrangement according to FIG. 9 will be explained. At the beginning of the present consideration, the voltage at the output 118 of the memory stage 144 is again at a level L ₁ , which corresponds to the peak value of a previous pulse. This level L ₁ remains until the pulse 71 appears. As soon as this pulse 71 exceeds the lower limit voltage v _th specified by the limiter stage 87, the leading edge of the signal 90 'emitted by the stage 87 in conjunction with the leading edge detector 93 generates a sharp switching pulse 108' which triggers the short circuit 110 via the channel 108 and thus causes a brief discharge of the capacitor of the memory stage 144 via the connection 112. The level L ₁ accordingly drops according to the vertical line R ₁ . From then on, the pulse 71 passes the memory level 144 according to its original course and accordingly also occurs in this form at the output 118 of the memory level 144 until it has reached its peak value. At this moment, the memory stage 144 holds the new charge level L _{9 reached} , while the pulse 71 decays again. Later, when the next pulse 72 retracts, a switching pulse is generated 108 'again and the memory stage 144 discharged so that the level L ₂ along the line R ₂ decreases, whereupon the pulse 72, the memory stage of new charges. This leads to the creation of the new level L ₃ and, similarly, the creation of the lines R ₃ and the level L ₄ .

The signal 218 of FIG. 10, which can be viewed as a modified histogram, already has certain loss zones if you compare its shape with that of the ideal signal 18 of the F i g. 3 a compares. These loss zones are labeled 221, 223 and 224 and hatched in one direction while gain zones 222 and 225, showing an increase in charge over signal 18 of the Fig. 3a represent, are hatched in the opposite direction.

If the areas of the loss and profit zones were each of the same size, then the histogram 218 would be practically equivalent to the histogram 18. The areas of the loss zones, however, are larger than the areas of the profit zones, which is to be regarded as a particular advantage of the modified histogram 218 with regard to the problem of coincidence compensation, because in this way compensation for coincidences which, like Required in the above relationships, is a function of the raw count N _R. It is then relatively easy to adjust the fraction separated out per occurrence of a pulse so that it represents a specific percentage of the individual pulses and just reduces the signal again by the coincidence component. One only needs to provide a corresponding loading device, which a loading in the order of magnitude of the coincidence

23 24

proportion allows, and then the percentage on the circle 110 also only the rising edges stung by means of this device and the duration, during which the pulse 90 'can act,
The memory stage 144 has a charging circuit with which the desired value is to be set in such a way that the desired value has a capacitor C which, once the average amplitude has been reached. If it has been charged 5 to a certain potential, which is about the investigation of particles, its potential as a result of the diode arrangement D ₅ , this setting can be checked on the basis of changes that are associated with a high positive potential, concentration of the suspension holds on. When actuated, the short-circuiting circuit delivers the correctness and accuracy of the setting to 110 at its output 112 a sharp switching guarantee, ίο pulse similar to the pulse 108 ' of FIG. 10, the

A practically implemented device for which the capacitor C is discharged, except for the delivery of the desired display at 136, was primarily geared towards the potential of the input signal being discharged, the mean cell volume, whereby only the voltage drop (MCV) of red blood cells remains to be determined , the corresponding difference at the diode chain D ₅ had already been referred to at the beginning. At 15 is to be added. After the switching of this type of investigation has disappeared, the pulses of interest at the output of the short-circuit 110 values are in most cases very similar to one another. the memory stage 144 charges itself in accordance with the further increase in the measuring pulse in the practical embodiment mentioned and applies the amplitudes at the output of the main amplifier to the input 118 of the intermediate amplifier kers in the order of magnitude of 10 V; those between 20 150 with the value of their respective charge,
The intermediate amplifier 150 of Fig. 11 has a concentration ratio of 1:50 000 passing its input a so-called field effect time was about 250 microseconds, while the transistor TR ₈ , which is due to its high duration of the pulses In the order of magnitude of 30 loss resistance, it was particularly suitable for a Bekrosecond and could essentially be viewed as a constant load on the memory level 144 through which it remained in between. Prevent from the uni-amplifier 150 . The transistor TR ₈ vibrators delivered pulses were included in a source follower circuit lasting less than 2 or 3 microseconds. The other transistors of the intermediate position, which thus remained far below the duration of the one amplifier 150 cause a phase reversal with individual pulses. 30 feedback, the output 84 being an emitter

Fig. 11 shows an electronic circuit which is the follower circuit.

relevant components of the arrangement according to Voltage present at output 84 beauf-F i g. 8 contains, but in some details of it beats the low-pass filter 124 and acts on the deviating from there. As far as it makes sense, for the identification “and” gate 204, which in the present case is formed by a connection of the components of FIG. 11, however, the same simple switch that has been selected by a given reference symbol as in the previous relays of the timer 208 are closed figures. In detail is the output of the measuring can. As in Fig. 8, the timer 208 is actuated again by the value converter 130 at the input of the main control circuit 206 , so that the switch amplifier 134 is switched. At the main amplifier of the “and” gate 204 only during a predetermined period of time, a further intermediate 40th time is closed. Here, too, the amplifier 260 is switched on, which prevents an inadmissible load on the output voltage of the “And” gate 204 again in the amplifier 134- . The input voltage of the integrator stage 202 is stored and then drives that at 261 can be 15 V, for example. The Tran servo system 265 , which in turn has the mechanical sistor TR ₃ as well as the other transistors used in the circuit display disk 267 opposite a window, are of known type. Da 45 cut twisted.

to the values of the resistors and capacitors in order, the signal of the repeater 150 each are now not specified, the transistors burden, by means of the output 96 in Fig. 11 provided with their reference numbers of Anstiegsflankendetektors93 switching Worden occurring, based on which the here The impulses in question are triggered by the univibrator 156 , which an average person skilled in the art can derive the appropriate values for a narrow impulse of precisely defined width for the components. Input 102 of the load level 200 is there. Respect-

The point 136 corresponds to the point 136 of the load level 200 it depends, Fig. 8 and represents the input to the memory that the load pulse emitted by it both level 144 and to the limiter level 87. If it is sharp and wide enough, a measuring pulse 55 acting on the stage 87 by the diode present in the load stage exceeds the limit voltage level v _th (FIG. 10) either in saturation or in the blocked state, inevitably occurs at the output of stage 87 keep. In the event of a negative signal at the output, the square-wave pulse 90 'appears (Fig. 10). During the course 84 , the deflection of the load pulse arrangement of FIG. 8 only has to be sufficient to block the diode, so that the rising edge detector 93 shows that the uni 60 load can once be triggered in the desired manner instead of the vibrator 156 on the other hand, the short finds. The resistor R ₅ determines the extent to which the closing circuit 110 is to be actuated, the circuit has a load and can be adjusted accordingly, according to FIG. 11, two rising edge detectors. If the correct setting has been found, the block designated by 93, as the rising edge detector, which is the compensation of the coincidence influences, ensures that only the rise 65 takes place to the required extent stood to be fixed in this setting. In Fig. 11 Univibrator 156 can get, while the input is the output 212 of the load stage 200 straight transition of the short circuit 110 is designed so that connected before the "and" stage, instead

however, it could be coupled to the input of the integrator stage 202 or also to the intermediate amplifier 150.

In the arrangement according to FIG. 11, as in the other embodiments of the invention described above, the basic idea of achieving a reduction in the charge of the histogram-like signal is that the signal is loaded each time in the same rhythm with which the pulses on Appear entrance of the arrangement. This means, however, that the load is proportional to the raw count N _R , which in this respect can be directly called the pulse load. As stated above, the percentage of the load required to compensate for the coincidence represents a constant value, which makes the construction of an arrangement that is suitable for the desired load very easy, as has already been shown (Fig. 10), that at least a partial coincidence compensation can be achieved solely by generating the modified histogram 218 instead of the ideal histogram ^ 18.

Claims (23)

. . Patent claims: 1. Method for determining the mean value of a large number of temporally spaced successive ones physical quantities by means of conversion into corresponding electrical measuring pulses, whose amplitudes are proportional to the amount of the physical quantities that trigger them and their time interval corresponds to the time interval in which the physical quantities successively T. occur in such a way that the measuring pulses in a monopolar, histogram-like signal are converted with the individual measuring pulses corresponding signal sections, whose Levels correspond to the amplitudes of the signal - Section causing pulses are proportional and each remain constant for a period of time with the level, the duration of which is equal to the time that elapses between the occurrence of the triggering measuring pulse and the occurrence of a following measuring pulse, characterized in that the physical sizes' particles mentioned in a suspension are obtained by running an electrical scanning zone and generate measurement pulses mentioned, and that the average "amplitude of signal sections 0γ * ο · •"^'i B ~ ⁱ ni)>' ^{r 'we} l ^c h ^{e is} the total time (T _n ) form, is determined so that a practically constant average value (12) of the measured variable is obtained at the end of the measuring process. 2. The method according to claim 1, characterized in that that the individual signal sections are formed immediately adjacent to one another will. 3. The method according to claim 1, characterized in that the individual signal sections are connected to one another _{with levels (L 1} , L., ...) Lying between their areas of constant levels (L 1, L., ...) Lying time intervals (221, 222 ", 223, 224, 225) .. 4. The method according to claim 3, characterized / that the intervals (221, 222, 223, 224, 225) are each dependent on the occurrence of a measuring pulse (71, 72, 73) by means of the starting range of the leading edge of the measuring pulse (71, 72, 73) are triggered and last until the peak value of the measuring pulse (71, 72, 73) is reached (FIG. 10). 5. The method according to claim 3 or 4, characterized in that the intervals are selected as a function of the frequency with which the variables to be examined coincide in time and thus brought into a certain percentage ratio to the total value of the monopolar signal (218) . 6. The method according to claim 1 to 5, characterized in that a new signal section is triggered by reaching the peak value of a measuring pulse (71, 72, 73) (Fig. 6). 7. The method according to claim 1 to 5, characterized in that a new signal section is triggered by the decay of a measuring pulse (71, 72, 73) by means of its rear edge (Fig. '3).' 8. The method according to claim 1 to 7, characterized in that the measuring pulses (P _1x P _2. .-. ·; 71, 72 ...) are generated by particles contained in .einer electrolytic suspension which pass through a passage opening (30 ). , are passed through, which between two measuring electrodes (32, 34) one for the detection of changes in resistance which the particles cause when passing through the passage opening (30). suitable measuring section is arranged. 9. Device for carrying out a method according to claim 1 to 8, characterized in that one serves to generate electrical measuring pulses (F ₁ , F ₂ ...; 71,; 72 .....) corresponding to the quantities to be examined transducer (22; 130) (. 71, 72 ..) (.. L _1, L ₂ .. ") a corresponding level for holding a predetermined electrical charge on one of the amplitude of the measurement pulses · appropriate memory stage (78, 114; 144 ) is connected and: that an ... electrical switching device (94, 110, 100; 156, 142) for changing the level (L _n ) of the charge of the memory stage (78, 114; 144) as a function of a new measuring pulse (71, 72 -) to a new level (L _{n + 1} ), which is proportional to the amplitude (v ") of the new pulse, with its input to the output of the transducer (22; 130) and . its output is connected to the input of the memory stage (78, 114; 144) and the output of the memory stage (114; 144) is connected to the input of a suitable arrangement for forming the mean value of the charges detected by the memory stage (114; 144) (Figures 2, 5, 8, 9). 10. The device according to claim 9, characterized in that for actuating the electrical switching device (94, 110, 100; 156, 142) between the output of the transducer (130) and the input of the switching device (94, 110, 100; 156, 142) a discriminatory test stage (87, 92), which is preferably at the end of the trailing edge of a measuring pulse (71, 72...), Is connected to the time value of the measuring pulse (FIGS. 2, 3). V 11. The device according to claim, characterized in that for actuating the switching device (94, 110, 100; 156, 142) between the output of the transducer (130) and the input of the switching device (94, 110, 100; 156, 142) a discriminator stage (160) responding to the peak value of a measuring pulse (71, 72 ...) is connected (FIG. 5 , 6). 12. The device according to claim 9, characterized in that for actuating the electrical switching device (94, 110, 100; 156, 142) between the output of the transducer (130) and the input of the switching device (94, 110, 100; 156, 142 ) a discriminator stage (87, 92) responding to a time value of the rising edge of a measuring pulse (71, 72), preferably in the vicinity of the zero line (10) (FIGS. 8, 9, 10). 13. Apparatus according to claim 9 to 12, characterized in that the arrangement for averaging is a moving coil instrument (126) . 14. The device according to claim 9 to 12, characterized in that the arrangement for averaging an integrator stage (202) connected to the memory stage (144 ) and an electrical circuit arrangement (206, 204) with a timer (208) for connecting the integrator stage ( 202) to the memory stage (144) during a time predetermined by the timer (208) and a device for determining and displaying the amount of charge stored in the integrator stage (202) during the predetermined time (FIG. 8). 15. The device according to claim 14, characterized in that to display the amount of charge of the integrator stage (202) is connected to this a servo device (265) with a movable dial (267) . 16. The device according to claim 9 to 15, characterized in that the transducer (22; 130) is a device for examining particles, in which a first container (28) with a second container (24) via a passage opening (30) in There is a connection and on both sides of the passage opening (30) measuring electrodes (32, 34) an electrical measuring section suitable for detecting changes in the conductivity of a suspension with the particles that occur as a result of particles passing through the passage opening (30) are arranged (Fig. 4). 17. The device according to claim 9 to 16, characterized in that one of the measuring pulses (71, 72...) Controlled second electrical switching device for reducing the mean value of the histogram-like signal (18; 218) by an amount that corresponds to the increase in amplitudes corresponds to individual measuring pulses due to coincidence, can be switched to the input of the arrangement for averaging (126, 202, 203, 265, 267) . 18. The device according to claim 17, characterized in that the second switching device has a load stage (200) and a switching stage (156) for switching on the load stage (200) to the input of the arrangement for averaging (126, 202, 208, 265, 267 ) in the rhythm of the pulses (71, 72 ...) supplied by the transducer (130) and during time intervals lying within a pulse interval (FIG. 8). 19. The device according to claim 17 and 18, characterized in that the load level (200) is adjustable to desired load values, depending on the frequency of expected coincidences and depending on the respective measurement conditions and the transducer (22; 130) in one certain percentage relation to the height of the levels (L ₁ , L ₂ ...). 20. Apparatus according to claim 18 and 19, characterized in that the load stage (200) has a resistor (R ₅ ) switchable by means of a transistor (Tr ₁₃ ) (Fig. 11). 21. The device according to claim 20, characterized in that the resistor (R ₅ ) is variable (Fig. 11). 22. The device according to claim 9 to 21, characterized in that the memory stage (78) is connected to the memory calls (114) via a normally interrupted connection channel (84, 116) via which the memory stage (78) is charged in the uninterrupted state to the memory stage (114) and that a second, from the measuring pulses (71, 72 ...) controlled switching stage (100) for the temporary closure of the connecting channel (84, 116) depending on a certain, respectively the same height or the apex of the measuring pulses (71, 72 ...) and only at least partial discharge of the memory stage (78) is provided (FIG. 2). 23. The device according to claim 9 to 22, characterized in that the memory stage (144) is connected to the output of the transducer (130) via a continuously conductive connection channel (136) and that one of the rising edges of the individual measuring pulses (71, 72, 73) controlled switching device (87, 92, 110) is provided for the immediate discharge of the memory stage (144) to an essentially discharged state (FIG. 8). For this purpose 3 sheets of drawings