DD140187A1 - DIGITAL-ANALOG CONVERTER - Google Patents

Abstract

The invention relates to the field of computational and DigitalmeStechnik and aims to a digital-to-analog converter to provide a simple metrological control at simultaneous simplification of the circuit of the summing unit can be made. The technical task is to a digital-to-analog converter for conversion (sampling) of the To create an output circuit, one of resistors with one limited number of denominations composite summation unit having. The digital-to-analog converter includes a series circuit from a summation unit, a unit of switching elements, a sampling unit, wherein the output of the summation unit and the Input of the sampling unit the output or input of the digital analog converter represent. The scanning unit closes,, n " one-type function cells according to the number of positions of the item to be converted Codes. Each function cell contains a logical OR element and a flip-flop whose I output is one digit of the multi-digit output the scanning unit and one of the inputs of the logical OR element one digit of the multi-digit Represents information input of the digital-to-analog converter. The The output of the logical OR element is connected to the I input of the Flip flops connected. Further features characterize the invention. The invention may be applied to digital-to-analog converters and digital meters be evaluated. - Fig.1 -

Description

WW

Field of application of the invention

The invention relates to the field of computational and digital metrology and relates in particular to digital-to-analog converter. The invention can be applied in digital-to-analog converters and digital meters.

Characteristic of the known technical solutions

Digital-to-analog converters (E.I, Gitis, "Information Transducer for Electronic Digital Computing Machines", 1971) are well known which include a summation unit for the code value proportional to the normal values. The output of the summation unit represents the output of the digital-to-analog converter, which also includes a unit of switching elements whose number is equal to the number of code parts »The switching elements control the activation of corresponding normal values. The output of each switching element is connected to a corresponding input of the summation unit. The inputs of the switching elements represent the inputs of the digital-to-analog converter.

As is known, the digital-to-analog converters have errors in converting the code to an analog value. These

§ 2

54 488 /

Errors essentially consist of a deviation of the * analog value from the corresponding nominal value of the code to be converted.

The distribution law for the system error of the digital-to-analog converter has a very complicated character, which is due to deviations of sizes of the hormonal values of the characteristics.

The problem of metrological control of digital-to-analog converters in the operation of precise digital-to-analog converters is particularly noticeable. On the other hand, the precise digital-to-analogue converters for 16 to 18 binary digits are used as controllable cormals to test the highly accurate measuring apparatus the accuracy of such precise digital-to-analog converter must be higher than the accuracy of the existing measuring apparatus and the metrological control of the known precise digital-to-analog converter can only be realized by Außeninessungeη _{s it} is impossible with a certain degree of accuracy, the metrological control of the digital Analog converter.

There are also digital-to-analog converters (WG * Galualu, WoW, Bratkevich, AP, Stachow "Transducer of code in voltage for (i, k, s) ~ algorithms", USSR university news, device construction, Bd. XVIII, H. 4 _S 1975), which are similar to the digital-to-analog converter described above. The special feature of this device is that the position values of the a, n correspond to the inputs of the switching elements.

-3- ZU® p.92 26.4.1979

54 488 /

elements of incoming codes are chosen according to a recurrent Fibonacci relationship, i. the significance of the 1-th code is equal to the significance of the (l-1) -th and the (l-p-i) -th digit.

However, in this digital-to-analog converter, due to the considerations and causes described above, the metrological control is more complicated.

Moreover, with this digital-to-analog converter, it is impossible to realize a conversion of the "golden" p-code into an analogue value. The "golden" p-code becomes the representation of a real number A in the form of a polynomial

+ co

A = £ a ^ _p (I)

l = -CO

understood, being

a, £ f 0,1] - a binary digit in the first position of the "golden" p-code

and the significance of the first digit of the code, the one

P represents the 1th power of the "golden" proportion,

-. a "golden" p-proportion that is a positive root

an equation

P ⁺¹ _ _X P -1 = ο. (2)

represents (p - a given natural number), denote.

~ ⁴ ~ 20® 9 9Jf 26.4.1979

54 488 /

The exploitation of the "golden" p-code can lead to a simplification of the summation unit, namely to a reduction of the number of normal resistances of a grid, which converts the "golden" p-code into an analogue value.

Object of the invention.

The aim of the invention is to provide a digital-to-analog converter in which a simple metrological control is carried out with a simultaneous simplification of the circuit of the summation unit. ·

Statement of the nature of the invention

The invention has for its object to provide a digital-to-analog converter having a stepwise metrological control, an analog output value by converting (sampling) one of the output number corresponding output code in Kodekombination ^ the same number, in particular by using a "golden" p Codes, and further comprises a summation unit composed of resistors having a limited number of characteristics.

This is achieved by providing the output of the digital-to-analog converter, which has a summation unit for normal values whose magnitudes are proportional to the codestring values, and a unit of switching elements by codestring number whose multistage output is connected to the multi-digit input of the totalizer summation unit As the output of the digital-to-analog converter, according to the invention there is a multi-digit information input for inputting a code combination and a control input for inputting a signal for a code signal.

-5- 208 9 92

54 488 / Ϊ7

There is a code scanning unit having the multi-digit output connected to the multi-digit input of the unit of switching elements.

It is expedient that in the digital-to-analog converter, the code scanning unit according to the invention "n" puncturing, whose 1-th, the 1-th code correspondingly has a flip-flop whose 1- output the 1-th digit of the multi-ellenausganges the Sampling unit comprising a logical OR element whose first input occurs as the 1-th digit of the multi-digit information input of the sampling unit and whose output is connected to the I input of the flip-flop, in addition to the digits of the (p + 1) -ten to the (n-1) th corresponding function cells include a logical AND element with (p + 3) inputs whose output to the O input of the flip-flop of the same functional cell, whose first input to the 1 output connected to the flip-flop and the second inputs of the logical UlTD elements of all function cells are interconnected and act as Sfeiereingang the scanning unit, the inputs from the third to the (p + 3) -th of each l ogischen ÜND element of the 1-th function cell respectively to the 0 outputs of the flip-flops of the function cells from the (l-1) -th to the (1-ρ-1) -th, the second and the third inputs of the logical OR element of the 1-th function cell, respectively connected to the outputs of the logical AND elements of the (1 + 1) -th and the (1 + ρ + 1) -th function cell, wherein

1 = 0, 1,2,3, ..., (n - 1)

η - the number of digits of the p-code.

" ⁶ ~ 208 9 92 26.4.1979

54 488/17

Preferably, in the Digit al analog converter according to the invention the summation unit for normal values in the form of a first group of (n + 1) resistors and a second group of η resistors enclosing resistance matrix executed »the resistances of the first group © inen at the two ends a.' And the resistor of the second group is connected at one end to the connection point of each pair of the resistors of the first group and the connection point of the first resistor pair of the first group represents the output of the digital-to-analog converter, the first and second circuits the (n + 1) th resistor from the first resistor group a value "C-Jti * all the remaining resistors of the first group have a value oL""R and each of the resistors of the second group has a value <A, ^ ⁺ R, where C - is a real positive root of equation (2)

^ _X P "1, ο means. , ·

It is permissible for the connection points of each resistance pair of the first resistance group to form the multi-digit input of the summation unit in the digital resistor analogue according to the invention and for the free ends of all resistors of the second group to be connected to the hull rail.

It is also permissible that in the digital-to-analog converter according to the invention the multi-part input of the summation unit is formed by the free ends of the resistors of the second group _β

-7- ZUCS 9 yd 26.4.1979

54 488 /

embodiment

The invention will be explained in more detail by description of an exemplary embodiment. In the accompanying drawings show:.

Pig. Ii a punctuation content image of the digital-to-analog converter according to the invention

Pig. 2: a principle of the unit I.

Like from Pig. I, the digital-to-analogue converter includes a summing unit I for normal values whose magnitudes are selected in proportion to the numbers of a code series, for example the Pibonacci-p code. The output X of the unit I acts as the output of the digital-analog-Y / andlers, from which a p-code corresponding analog value is read. The multi-digit output of a unit 3 of switching elements is connected to the multi-digit input of the summation unit I. The number of switching elements is equal to "n" and corresponds to the number of digits of p-code "to the multi-point input of the unit 3 of the multi-point output of a scanner unit 4 is connected to a p-code of the same to implement the ^Ä usgangskodes a number in other code combinations Number is provided »

The Ausgangskodekombinationi-reaches the multi-digit input 5 of the digit al-analog V / andlers, which also has a control input 6 for inputting a control signal for the sampling of the ρ-code.

The circuit of the sampling unit 4 for a p-code is designed for a specific value ρ = 1 The sampling

54 488 /

In the general case (p-any given number), "4" contains n-type puncture cells 7 (n = code; -step number), where η = 8. The L-th, the 1-th digit of the p- Corresponding puncture cell 7 is provided for sampling the 1-digit I-value location for the (1-I) and (1-ρ-1) -th digit of the p-code having an O value. In Pig. I are the (1-I) and the (1-ρ-I) -th place, for example, for 1 = 4, the third and * second place «Each puncture ropes 7 contains a flip-flop 8. The I-output 9 of each flip-flop 8 represents the lth position of the multi-digit output of the unit 4 connected to the input of the 1 th switching element (not shown in FIG. 3) of the unit 3. Each functional part 7 also contains a logical OR element 10 whose one input represents a position of the multi-digit information input of the digital-to-analog converter and whose output is connected to the I input II of the flip-flop 8 of the same function cell 7 »

The dot lines 7 corresponding to the digits from the (p + I) th to the (n-I) th also include a logical AND element 12 which, in conjunction with the logical OR element 10, ensures sampling of the point concerned "

Each logical AND element 12 has (p + 3) * i. four inputs, on »The output 13 of the logical AND gate 12 is connected to the O input of the flip-flop 8 and the first input to the I output 9 of the flip-flop 8 * The other inputs of the logical UMD elements 12 of all function cells 7 are connected together and represent the control input 6 of the digital-to-analog converter «Between the Punktlonszellen 7 of the unit 4 exist

-9- ZU & 9 92 ²⁶ · 4.1979 • 54 488 /

following compounds. The inputs of the logical AND element 12 from the third to the (p + 3) th (ie the fourth) of the puncture cell 7 are respectively connected to the O outputs 14 of the flip-flops 8 of the functional parts 7 of the (1-I). The other inputs of the logical OR elements 10 of the 1-th function cell 7 are respectively connected to the outputs 13 of the logical UHD Elements 12 of the (1 + I) th (d * h of the fifth) and the (1 + ρ + I) th (ie, the sixth) functional cell 7 are connected.

FIG. 2 shows a basic circuit of the summing unit I, with the aid of which a conversion of the "golden" p-code into an electrical voltage equivalent thereto is performed. The summation unit I represents a matrix of resistors connected together to form two resistor groups 15 and 16 The one group 15 includes (n + I) resistors 15j to 15n ₊ j, that is, the stator terminals 15j to 15g, which connect one to the other at the two ends. Form zero bus 17 connected series circuit. The other resistor group 16 includes (n), ie, eight, resistors I6y to I6 _g a. One of the ends of each resistor 16j, 16g is coupled to one of the connection points 18j to 18g of each pair of resistors 15j to 15q of the first group 15.

If the digital-to-analog conversion is based on the principle of current addition, all other ends of the resistors 16-bia 16q are to be connected to the zero rail 17, and the entirety of the connection points 18 to 18g in this case will represent the multilevel input of the summation unit I, thereby is the input 18- an input for

~ ¹⁰ ~ 208 9 92 ²⁶ · 4.ΐ979

54 488 /

the most significant digit, while the input 18 _{R is} an input of the least significant digits and the "golden" p-code. - "

The free ends of the resistors I6j to I6 _ß form the multi-digit input of the summation I (this variant is not shown in Fig. 2), and the digital-to-analog conversion is in this case according to the principle of iäpannungssuinmierung »The special feature of FIG. matrix shown 2 consists in the selection of nominal values of the resistors I6j to I6 _g and 15j to 15 ", as in this case, the unit I is a transformation of the" golden "p-code implemented e the nominal values of resistors 15 _O to 15 _O to the first group 15 sindoC / ^F R, the nominal values of the resistors 15 _T and 15 <W ^ Pr and the nominal values of all resistors 16 _T to 16g of the second Gruppect £ R equal, where R - the dimension of electrical resistance undoL ^ - a real positive Root of equation (2)

~ x * - I = 0, which in this case (for ρ = I) has the form x ² - χ - I = 0,

assumes call, and the nominal values of said resistors 15j, 15g '15 ₂ *> ^is 15 and 16 to _{ß ß} I6 respectively equal _

1 + -fi _R. 5-1 _R and 3

R, V ' R ^WiX "' ^T ^ R

are _β

In this case, resistances of only three nominal values are used in the resistance matrix

-9- Zuq W 936.4.1979

54 488/17

Manufacturing technology for such matrices significantly simplified.

The digital-to-analog converter (Pig. I) works as follows. The converted into an analog value output code (for example, the Fibonacci-p code, or the "golden" p-code) reaches the multi-digit information input 5 of the sampling unit 4 in the form of a binary combination of electrical impulses via the corresponding logical OR elements 10 and the I inputs II of the flip-flops 8 initiate a reversal of the I-values of the output codes corresponding flip-flops 8 in the I state * at the I outputs 9 of these flip-flops 8 I signals are formed over the corresponding switching elements of the unit 3 control the corresponding normal values (resistors 15j to 15q, 16, to 16g (FIG. 2) in the unit I). As a result, at the output 2 of the unit I, an analog value corresponding to the predetermined p-code occurs. Let us also consider the work of the digital-to-analog converter by the example of the conversion of a "golden" p-code, because. Less attention is paid to the latter in the literature. The power of the "golden" p-proportion has the form

cL - is a real positive root of equation (2).

Is on the multi-digit input 5 of the digital-to-analog converter, for example, a "golden" p-code of the form

54 488 /

76543210 - Job number

IOOIOOIO .- "golden" p-code.

delivered j will be its numerical equivalent in the relevant

Case the number:

while at output 2 (Fig * I) an A proportional analogue value will be formed *

If additional normal values are entered into the unit I of the digital-to-analog converter (FIG. 2) whose magnitudes are negative powers of the "golden" I-proportion

»-I» -2 rJ -3

with the help of such a unit I, an analogue value can be converted into any natural number »-. ·

In the case of a conversion according to the principle of current addition (FIG. 2), the outputs of the unit of the switching elements which correspond to the I-digits of the "golden" p-code reach the corresponding input 18j to 18f> an ITormal current I * " golden "p-proportion has a following basic property

¹ -ρ-

which is valid for any integer 1 «,

For P = I the identical equality (h) has the form:

¹ ^ ² (5)

54 488/17

Taking into account the properties (4) and (5), it is not difficult to convince oneself that the equivalent resistance at any point 18 _T to 18 _Q against the zero rail 17 is constant and equal to β R, where - -

β = means and the voltage over-

The factor of transfer between two adjacent points (for example, between 18 "and 18 ^) is equal to. Therefore, when the Kormalstrom I _{Q is applied} to the point 18g representing one connection point of the resistors 15 ^ 1f> ₂ ^ ₁ ^ ₁ , it appears at the exit 2 an electrical voltage equal to ß IR, and at the supply of the Uormalstroms at the following points 18 ", 18g, ... at the output 2 each voltages

ß _p I _o R <^ ~ ^I , ß _p I ₀ Ro <^ ² , ... etc. occur »

The above-described property of the Pibonacci p-code and the "golden" value which connects the position of the 1-th digit with the values of the two preceding parts, the 0- I) and the (1-ρ-I) -th digit "p-code allows the metrological control of the digital-to-analog converter when using the scanning unit. 4 for the p-codes significantly simplify. If, for example, the multi-digit information input 5 is one of the seventh power oLj of identical equality (5)

= 10000000

-14- cuo yy £ - 26.4.1979

.54 488 /

sufficient "golden" I-proportion corresponding "golden" p-Kc-de is supplied, then, due to the identical Gleicheit (5) has a sampling in the unit 4 of the seventh, having an I-value location for the one O value sixth and fifth place, ie a following conversion

76543210 -. ^s hiring numbers ₇ JlOOOOOOO - point values

¹ Loiiooooo

is made _f - so does not change the number equivalent A of the p-output code. On the other hand, the value at the output 2 must not be changed in such a conversion * If then the fifth digit for the fourth and third and then the third scanned for the second and first digit, d * h »a subsequent conversion of the code«

Adjusting numbers ^S - 76543210

1101100,000 -.

OIOIIOOO pIOIOIIO

is undertaken, such Kodeumwandlungen must not cause any change in the output value at the output 2 · The deviation of the analog value from the default value by a metrological "Pehler the digital-analog converter border ^w ert just a feature of its non-conformity with the metrological characteristics.

The above idea of metrological control of the di-

-15- 2Q & 9 92 ²⁶ · ⁴ · ¹⁹⁷⁹

54 488 /

Gital-analog converter based on the operation of the scanning unit 4 in the operation of metrological control. In operation of the metrological control at the multi-digit input 5, a code corresponding to the normal value to be controlled may be present, for example a p-code t corresponding to the position of the seventh digit

765432IO - Position numbers ⁷ = 10000000 - Positions

arrive. In this case, the flip-flop 8 of the function cell 7 of the highest point goes into the I state, which is why

7 at the output 2 eino ^ j proportional analog value and at the first input of the logical AND element 12 of the function cell of the highest point form a release potential ·

First, a metrology control the size of the N _o rmalwertes the seventh point to which by means of a zero member (for example, a sensitive galvanometer) derö ^ -l corresponding analog output value is realized, with

7 is compared with a corresponding external normal value.

After that becomes a metrological control of the relationship. (4) between normal values. For this purpose, the control input 6 is supplied with a Kodeasbastsignal representing a short-term pulse whose duration is equal to the duration of the transition process in the logical AND element 12. This pulse leads to the appearance of an I signal at the output 13 of the logical AND element 12 of the function cell 7 of the highest point, which the flip-flop 8 of the function cell 7 of the highest point in

-16- diß® 992 04.26.1979

^ 54 488 /

allows the O state to be reversed, while via the ODBR logic elements 10 of the functional parts 7 of the two preceding digits, this pulse causes the flip-flops 8 of the function cells 7 of the two preceding points to tip over into the! state, i. Unit 4 goes into another

corresponding state:

76543210 - Line numbers

OIIOOOOO - Values of the "golden" p-code

In this state, the constancy of the 'analog value at the output 2 of the digital-to-analog converter is checked again

In such a state, release potentials appear in the unit 4 at the first, third, and fourth inputs of the UM logical element 12 of the fifth digit. After feeding the control input 6 of the next Abteat pulse, the unit 4 goes into a new djr correspond to the state:

76543210 - Job numbers

0I0II000 - "golden" p-code

and after the feed at the control input 6 of the third

7 pulses in the last state corresponding to %

76543210 - Position numbers 0I0I0II0 ~ "golden" p-code

over* .

In this case, the control input 6 is to be detected according to C © © ¹¹¹

-17- 208 992 ²⁶ · 4 · 1979

54 488 /

Sampling pulse, the constancy of the analog value at output 2 checked. Since at each step of the metrological control, the magnitudes of the normal values of the three-adjacent locations: the 1-th, the (1-I) -th and the (l-2) -th are compared, so in the case of mismatch of the digital Analogue converter with the metrological characteristics to easily determine the cause of this mismatch with an accuracy up to three digits.

The metrological control of such a digital-to-analog converter can also be carried out with the aid of a highly sensitive oscillograph, which separates the AC component and is connected to the output thereof * For this purpose, it is sufficient to multiply the same code, for example, a p- Code:

765432IO - Job numbers I0ÖOOOOO - "golden" p-code

to read the code after each recording by means of code sampling pulses arriving at the control input 6 "Here it is possible to judge about conformity of the digital-to-analogue converter with metrological characteristics on amplitude of the alternating component on the oscilloscope's screen»

Advantageously, in the present digital-to-analog converter, there is a considerable simplification in the procedure of setting the normal values of the unit. For ρ - I, it is sufficient in the n-digit digital-to-analogue converter proportional normal value of the highest (n - I) th digit and one

-ie % ß & 992 ²⁶ · ⁴ · ¹⁹⁷⁹

- 54 488 /

⁰ ^ / "proportional hormal value of the preceding (n - 2) th place, after which the normal value of the (n-3) th place is ^set in such a way that the relationship:

is satisfied, and after the setting of the Horrnalwertes ^ i - th 3) position is the normal value of (n - -th 4) _s location such that the relation:

(7)

is fulfilled

In addition, the code scan resulting in unit 4 can be effectively used to increase the accuracy of the digit-to-analog converter by reporting the analog value to various code combinations corresponding thereto. For this purpose, it suffices to connect an averaging device to its output and to supply several times a "golden" p-code to its input and then to convert it with the aid of the scanning unit 4. "In such an operating mode, it is in the digital» analog ~ converter an easier one to organize its continuous metrological control. For this purpose, it is sufficient to connect to the output of the digital-to-analog converter means for separating the alternating component, to couple the output to a control device for the amplitude of the alternating component

-19- £ U <9 WW ^ 26.4.1979

54 488 /

The aforementioned advantages of the present digital-to-analogue converter give rise to the assumption that it will find extensive application in digital measuring technology, in particular in the metrological control of high-precision digit-analog converters whose accuracy must exceed that of the existing measuring apparatus.

Claims (1)

-20- Ζΰ & fff 26.4.1979 ; 54 488/17 JE rf inda demand · 1e digit al-analogue Wandlex *, the one. Sampling unit for hormonal values, the magnitudes of which are proportional to the code value values whose first value is equal to a sum of the positions of the (1-I) -th and the (1-ρ-I) -th digit and a unit of switching elements according to the Code number contains _s whose multi-point output is connected to the multi-point input of the normal-value summing unit whose output appears as the output of the Bigital ~ Analogue V7andlera, characterized in that there is a code reading unit (4) whose multi-digit output is connected to the multi-digit input of the unit ( 3) is connected by switching elements and having a multi-digit information input (5) for inputting a code combination and a control input (6) for inputting a Kodeabtastsignals, wherein ρ - denotes a predetermined natural number. 2t digital-to-analog converter according to item 1, characterized in that the code reading unit (4) "n" puncture cells (7) $ whose 1-th, the 1-th Kodestelie includes a flip-flop (8) whose! - Output (9) represents the first digit of the multi-digit output of the sampling unit (4), a logical OR element (10) contains j its first input as l th place of the multi-digit information input (5) of the sampling -21-ZOB ffl "26.4.1979 54 488 / unit (4) occurs and the output of which is connected to the I input (11) of the flip-flop (8), as well as the function cells corresponding to the positions from the (p + I) -th to the (n-I) -th (7) include a logical AND element (12) with (p + I) inputs whose output (13) to the O input of the flip-flop (8) of the same functional cell (7) whose first input to the I -Output (9) of the flip-flop (8) connected and the second inputs of the logical UlTD elements (12) of all function cells. (7.0 are interconnected and act as the control input (6) of the scanning unit (4), the inputs from the third to to the (p + 3) -th of each logical UKD element (12) of the 1-th Funktbnszelle each to the O outputs (14) of the flip-flops (8) of the function cells (7) of the (1 - I) -ten to the (1-ρ-I) -th, the second and the third inputs of the logical OR element (10) of the 1-th function cell respectively to the outputs of the logical AND elements (12) of (1 + I) -th and the (1 + ρ + I) -th function cell (7) are connected, where η - the code number 1 = 0,1,2, oo., (N - I)
mean· Digital-to-analog converter according to items 1 and 2, characterized in that the summation unit (I) for normal values in the form of a first group (15) of (n + I) resistors and a second group (16) of η resistances enclosing resistor matrix is designed, the resistors (15j to 15 _{η +} τ) <* he first group (15) one at the two ends to a Hull rail (17) GE '-22- ^ Off ff 2 26.4.1979 · 54 488/17 connected series circuit and each of the resistors (1-6j to 16) of the second group (16) with one end to the Yerbindungspunkt (18j to 18) of each pair of resistors (15j biw 15 _{n +} j) d® ^ first group (15) _s is connected to the Yerbindungspunkt (18j) dea first pair of resistors (15j, 152) of the first group (15) represents the output (2) of the digital-to-analog converter, wherein the first resistor (15j) and the (n + I ) = te resistor (15 _{n +} j) from the first resistor group (15) a value ^ LR _t all other resistors (15 ₂ to 15 _n ) of the first group (15) a value «L -pR and each of the resistors (16 _T to 16) of the second group (16) have a value a ^ P + Ij>, where <= c - a real positive P
Root of the equation χΡ - I. 0
mean. 4 «digital-to-analog converter according to item 3, characterized in that the. Yerbindungspunkte (18j to 18) of each of the 7 / iderständ.e (15j to 15 _{n +} j) of ^the first resistor group (15) form the multi-digit input of Sununierungseinheit (I) and the free ends of all resistors (Ι6γ to 16) of the second group (16 ) are connected to the ITullschiene (17) * 5. A digital-to-analogue converter according to item 3, characterized in that the multiple input of the summing unit (I) is formed by the free ends of the resistors (16-16) of the second group (16) * O)