Digital-to-Analogue Converter Arrangement

The present invention relates to a digital-to-analogue converter arrangement formed on the basis of a digital-to-analogue converter of a type producing at its output an output voltage V'(N') such that: ##EQU2## where k is a constant, E is a voltage applied to a reference voltage input, n' is a binary number applied to an input for binary signals and M' is the maximum value n' can attain.

This type of converter arrangement is very advantageous for signal processing applications.

Because to the linearity of equation (1), the steps by which the voltage V develops result in an imprecision which is relatively greater for small amplitudes of the value V than for the large amplitudes. If a high precision for weak amplitudes is required, a digital-to-analogue converter circuit must be used in which "n'" is formed from many binary elements.

The invention has for an object to provide a digital-to-analogue converter arrangement providing a high precision for a low-amplitude output voltage without utilizing a binary number "n'" formed from many binary elements. To that end, a digital-to-analogue converter arrangement of the above-defined type is characterized in that is comprises a feedback network arranged between the reference voltage output and input.

Such a network may produce non-linearities between the output voltage and the number "n'". These non-linearities are not always desirable. In accordance with an important characteristic of the invention, a digital-to-analogue converter arrangement according to the invention is further characterized in that transcoder means are provided which, starting from a number "n" whose maximum value can reach M, adjust such a number "n'" that a "pseudo-linear" relationship is established between the number "n" and the voltage V': ##EQU3## where K is a constant and n.sub.p are specific values of n.

A "pseudo-linear" relationship must be understood to mean a relation which is only linear for certain values n.sub.p of "n". This notion will be explained in greater detail in the sequel of this description.

The following description, which is given by way of non-limiting example, with reference to the accompanying drawings will make it better understood how the invention can be put into effect.

FIG. 1 shows a circuit diagram of a digital-to-analogue converter arrangement according to the invention.

FIG. 2 illustrates the shape of the "staircase" curve which represents the voltage V as a function of the number "n".

FIG. 1 shows the digital-to-analogue converter arrangement according to the invention. It is formed around a digital-to-analogue converter circuit 10, for example an integrated circuit of the type NE 5020 produced by Signetics Corp. With a voltage V'at output 13, i.e. pin 2(1) of NE 5020 there must correspond a binary number n' applied to input 12 according to: ##EQU4## where k is a constant and E is the voltage applied to its reference-signal input 14, i.e. pin (17) of NE 5020, M' is the maximum value the number n' can assume, M' has the form:

M'=2.sup.b -1

where b represents the number of binary elements of the number n'.

By way of example, let it be assumed that such a converter arrangement is used in radio-navigation to apply a control voltage to an altimeter display derived from a digital information processed by a radio altimeter. The object is to supply a voltage between 0 and 25 Volts for altitudes in the range between 0 and 5000 ft and with a resolution better than 0.5 ft near 0, that is to say that at this level the voltage must evolve in steps of 2.5 mV. According to the prior art this requires the use of a 14-element binary converter. The invention makes it possible to produce a voltage which is accurate within 2.5 mV for low values of V using a digital-to-analogue converter circuit processing 10-binary element numbers n' (M'=1023).

Therefore, the digital-to-analogue converter is characterized in that is comprises a feedback network (20) arranged between the output where the voltage V' appears and the reference-signal input. This network (20) is constituted by a series arrangement of two resistors R1 and R2. A voltage U of 10 Volts is applied to the first end of this series arrangement and the output voltage V' is applied to the second end. The junction of these two resistors R1 and R2 is connected to the input 14. The designers have constructed the digital-to-analogue converter circuit in such a manner that at this input 14 the value of the voltage must be limited so that, to obtain the required 25 Volts, a d.c. amplifier 35 having a gain equal to 2.5 is used. The output voltage V of the amplifier 30 is then V=2.5 V'.

The voltage E is then written as: ##EQU5## Assuming ##EQU6## then: ##EQU7## For the sake of simplicity, let it be assumed for the further course of the description that k=1; assuming .alpha.=0.1 it will be obvious that V evolves for n'=0, 1, 2 by steps of 2.4 mV. It will be obvious that this precision is only reached for low values of n'; for n'=1020, 1021 the steps become of the order of 290 mV.

Formula (4) demonstrates that the parameter V is not a linear function of the variable n'. In the practical example, calculating the altitude is obviously estimated by means of the number n with the highest possible precision, which provides that the number "n" is defined by 14 binary elements (M=16 383). A transcoding circuit 30 is provided by means of which it is possible to obtain a voltage V which is proportional to the number "n" while having the precision required for low amplitudes. The transcoding circuit may be a read-only memory which causes a code "n'" to correspond to the whole code "n" by means of the formula: ##EQU8##

The integral value which is nearest to the value given by (5) is taken for "n". In practice, a value of "n'" will correspond to each n for low values of "n". This will not be the case anymore for high values or for several values of "n" to which a sole value of "n'"corresponds. Thus, it is said that there is a pseudo-linear relationship between V or V' and the value "n". FIG. 2 shows schematically the shape of the staircase curve which causes V to correspond to the value n. It is shown that the absolute errors of the low values of n and V are smaller than the absolute errors of the high values of of V and n for which the steps are larger. The plateaus are wider because of the fact that the transcoding circuit produces the same value (n') for several values of "n". Thus, in FIG. 2, when n=n1, n2, n3 the same voltage value V.sub.K is obtained, while for n=0, 1, 2 there are different values of V. The feedback network 20 causes deviations between the plateaus which vary from .DELTA.V.sub.m to .DELTA.V.sub.M. The specific values n.sub.p1, n.sub.p2 . . . of formula (2) are taken approximately in the centre of the plateaus so that a straight line D can be drawn through these points and the point of origin O. This straight line D represents the ideal linear function between "n" and "V".

It should be noted that the use of feedback networks other than those described are of course also within the scope of the invention.