DE2720653A1 - PROCEDURE AND CIRCUIT ARRANGEMENT FOR CORRECTING THE VOLTAGE DEPENDENCE OF SEMI-CONDUCTIVE RESISTORS - Google Patents

Description

Boeblingen, the 5th Shark A YR ^) · $

• moe-bd / som

Applicant:

International Business Machines Corporation, Armonk, N.Y. 10504

Official file number: New registration

Applicant's file number:

FR 975 303

Representative:

Patent assessor
Dipl.-Ing. Anton Mönig
Sindelfinger Str. 49
7030 Boeblingen

Description:

Method and circuit arrangement for correcting the voltage dependency of semiconductor resistors

709848/0851

FR 975 JP3

The invention relates to a circuit or a train * Hearing working method of the type mentioned at the beginning. It is in the field of by a diffusion or ion implantation method in a single-crystal semiconductor substrate, which typically consists of silicon Resistors and, in the narrower sense, indicates a method for correcting the voltage dependency of such semiconductor resistors.

: Integrated circuit components are being used more and more in circuit design. These semiconductor structures have changed in recent years as a result of multiple improvements in the basic active and / or passive components. In addition to the so-called active components, including transistors,

Diodes etc. are understood, the passive components, especially resistors, play an essential role for many types of circuits, especially for circuit applications related to analog circuits. In order to provide resistors together with the active components in integrated circuits, essentially two Techniques are used: one is the so-called. Compatible hybrid technology, in which the resistances in the form thin layers are formed on the surface of an integrated semiconductor wafer. The other option is monolithic technology where the Resistors are formed in the semiconductor substrate itself in the same way as the active components. in the In the latter case, the resistors can be produced, for example, by a thermal diffusion process or by means of ion implantation.

Thin film resistors are generally made of a suitable metal such as a nickel-iron-chromium compound or tantalum. These types of resistors offer a number

709848/0851

FR 975 303

- each -

of advantages, including a sheet resistance value in the range of from 2OO to 10.0OO Ω / ο and a relatively low temperature coefficient for the resistance value (TCR) in the order of 3% between 0 ° and 75 ⁰ C. The disadvantage is however, that thin film resistors often require additional process steps (vacuum deposition, cathode sputtering), which are usually not compatible with the process steps used to manufacture the semiconductor wafer itself. This immediately degrades the yield of good circuits. Manufacturers of monolithic circuit concepts therefore consider thin-film technology to be relatively unattractive for | Purposes of inexpensive mass production. The mono- | In contrast, Iithisehe technology appears to be the most attractive solution. j

Diffused resistors are known per se. They generally consist of a concurrent with the base or the emitter of a bipolar transistor formed doping region. For example, a P-doped base region diffused into a weakly N-doped epitaxial layer, which extends over a weakly P-doped single crystal silicon substrate extends. The resistance value depends on a number of influencing factors e.g. the doping profile, the penetration depth of the dopants, the length / width ratio of the diffused zones, etc. At the ends of the diffusion area two ohmic contacts are then formed, which represent the output connections of the resistor. The most positive The potential of the circuit is generally applied to the area of the epitaxial layer which is within the area of the Isolation enclosed area and contains the resistance area. One is limited with such Resistances in terms of the low sheet resistance (400 Ω / σ), the relatively large fluctuation range of the

709848/0861

, PR._97i. 303

Nominal values (+ 20%) as well as with regard to the high temperature coefficient, due to which the resistance depending on the temperature between 25 ⁰ C and 75 ⁰ C ι can change considerably. On the other hand speaks for these resistors their simple and well-known manufacturing method.

Resistances produced by means of implantation processes are a particularly interesting alternative to diffused ones Resistances, because they offer considerable advantages in comparison. Boron-implanted resistors in particular show a large sheet resistance range (3 to 10,000 Ω / 0 and above) and a low temperature sensitivity

(Resistance change of 0.5% between 0 and 75 ⁰ C). ! Furthermore, such resistors can be produced relatively precisely (+ 2%). A more detailed presentation of this type of resistance

can be found in an article by K. Rosendal with the title ι "Ion Implanted Planar Resistors," in the journal

Radiation Effects, Volume 7, January 1971, pages 95-100.

An article by H. H. Stellrecht provides a comparative presentation of the three types of resistance mentioned above entitled "Precision Ladder Networks Using Ion Implanted Resistors," in Wescon Technical Papers, Volume 15, 1971, ref. 8080-28 / 2 as well as an article by J. Den Boer et al with the title "The Termal Properties of High Value Gallium and Boron Implanted Resistors in Silicon ", in connection with the" European Conference on Ion Implementation " was published from September 7th to 9th, 1970.

However, little attention has been paid to this so far devoted to the increasingly important problem of the voltage dependence of such resistances. The current-voltage characteristic of a resistor should normally be linear. In fact, this can only be done for implanted resistors.

709846 / 08S1

FR 975 303

values with a very low sheet resistance in the order of magnitude of a few 100 Ω / _π , see for example an article by John McDougall et al with the title "High Value Implanted Resistors For Microcircuits", published in the journal Proceedings of the IEEE, Vol. 57, No. 9, September 1969, pages 1538 to 1542 and there in particular in Fig. 5 of this article. For implanted resistors with high resistance values, which require high sheet resistances (eg 20 ΚΩ / α), the current-voltage characteristic becomes non-linear and drops, so that there is signal distortion of an input signal applied to it. This is because the actual resistance for a given voltage is higher than expected. In this case, the current-voltage characteristic corresponds almost entirely to that of a junction field effect transistor (JFET). It should be emphasized that the value of the resistance is primarily dependent on the voltage applied to the transition between the resistance region and the epitaxial layer. To some extent, the value is still related to the voltage V _n across the resistor. The above-mentioned distortion arises from the fact that the depletion layer on both sides of the PN junction expands with the voltage applied to the PN junction, the effective cross section of the resistor track being reduced and thus ultimately its resistance value being increased. This is a phenomenon known under the name pinch-off effect, especially with JFETs, but it is particularly critical with implanted resistors because these generally have higher specific resistance values than diffused resistors and consequently the resistance is achieved in thinner layers .

A solution to the problem mentioned can be found in the aforementioned article by John McDougall. After that, will

709846/0851

FR 975 303

-β-

the specific resistance of the epitaxial layer is increased to values greater than 1Ω * cm. However, this allows the electrical properties of neighboring active components are adversely affected.

Another possibility could be the formation of the depletion region within the resistance region limit and thus increase the nominal resistance value. This can be achieved by increasing the ion dosage, such as that in the article by J. W. Hanson entitled "Ion Implanted N Type Resistors on High Resistivity Substrates ", in the journal J. Vac. Sei. Techn., Vol. 10, No. 6, November / December 1973, pages 944 to 947 and in particular in Fig. 6 is described.

Another possibility is given in the French patent specification 71 44 227. Though not a particular reference on the voltage dependence of the resistors is given, a correction of the voltage dependence shines through Implanting neutral ions into the resistance region, pre-; preferably near the PN transition, accessible. The main disadvantage associated with this is in that additional required process step to see.

It is the object of the invention to be able to better correct the voltage dependency of implanted resistors which allow the use of epitaxial material with relatively low resistivity values. Depending on the effort, that one is willing to invest should be one of the

; Voltage at the semiconductor junction between the resistance area j and the epitaxial layer of completely independent resistance! be achievable without additional process steps j

I are necessary in the formation of the resistance structures. j

The features that are important for solving this problem can be found in the claims. In summary, takes place after

709846/0851

FR 975 303

the invention a special setting of the changing potential difference between the resistance range and the surrounding epitaxial layer in order to either avoid the signal distortions caused by such changes in potential to keep them small or to compensate for them by means of equally large opposing effects. To the epitaxial layer a potential defined in more detail below is applied, which is preferably subject to the same changes as the resistance whose voltage dependency is to be corrected. In other words, if the average value of the potential difference between the resistance region and the Epitaxial layer remains constant over the length of the resistor, the resistance value also remains constant, and the mentioned interference can be neglected.

The invention is explained in more detail below on the basis of exemplary embodiments with the aid of the drawings.

Show it:

Fig. 1 is a partial cross-section through

a typical ion implantation resistor with its equivalent electrical circuit diagram;

i Fig. 2 shows a first embodiment of the invention,)

in which the corrective measures are independent of the resistance to be corrected with regard to its voltage dependency are;

Figs. 3A-D further exemplary embodiments of the invention,

where the corrective action depends on the resistance to be corrected;

709848/0851

FR 975 303

- Vd -

Flg. 4 shows the application of the first exemplary embodiment for correcting the voltage dependency in a digital-to-analog converter;

5 shows an illustration of the effects that can be achieved by means of the correction of the voltage dependency in digital-to-analog converters with different capacities, and FIG

Figs. 6A and 6B are illustrations of the use of the other exemplary embodiments for correcting the Voltage dependence of implanted resistors in an operational amplifier.

1 shows a conventional, monolithically integrated resistor structure 10. On a single crystal A semiconductor substrate, which typically consists of a silicon body 11 of a first conductivity type, e.g. j relatively thin (on the order of a few micrometers) I epitaxial layer 12 applied, which is relatively weakly doped (average specific resistance 1Ω * cm) arranged. In this epitaxial layer of the opposite conductivity type to the substrate, an isolated area is formed, the insulation by the heavily doped and down to the substrate isolation zones 13 is formed. By means of ion implantation of a dopant of the first conductivity type, an elongated region 14 is then produced which is intended to serve as a resistance region. This manufacturing step is generally carried out following the Formation of the P-doped connection regions 15 carried out, which were mostly provided simultaneously with the diffusion of the base regions of the bipolar transistors. the heavily doped relative to the doping of the resistance area

709848/0851

FR 975 303

-vr-

th connection areas ensure that the resistance value defined by the resistance area 14 is determined only by the sheet resistance of this area. At the same time as the emitter regions of the bipolar transistors, a second heavily doped region 16 of the second conductivity type is generally provided as a connection region for the epitaxial layer. For the resistance area, the two connections marked 17 and 18 are also formed. The second doping region 16 is also equipped with an ohmic contact 19. The pros associated with making implanted resistors; Blerne are known per se, cf., in addition to the prior art already mentioned, for example US patent 3 902 926. The potentials V ₁ and V _{2 are} applied to the connections 17 and 18 of the resistor. The connector 19 to the epitaxial layer can either remain open (in general

i mean undesirable) or with a suitable voltage

source are connected while the most negative potential

of the circuit is applied to the substrate to generate the PN- j

transition 20 between the substrate 11 and the Isolationsge j

offer 13 on one side and the epitaxial layer on j

the other side to lock. Through this reverse voltage j

the electrical insulation of the j

Zones framed semiconductor area accomplished. If necessary, the most positive circuit potential can the connection 19 are placed, whereby the electrical insulation of the resistance area from the epitaxial

! layer is guaranteed. This should be used if several resistors are to be formed in the same epitaxial layer area. Via the external connection can at least one depending on the epitaxial layer

In the circumstances, a suitable potential must be applied.

In Fig. 1 (right) is still the electrical replacement

709848/0851

FR 975 303

A circuit diagram of such an implanted resistor 10 is shown. The reference symbols given there correspond to those from the cross-sectional illustration.

The resistance value R of such a resistor 10 can be expressed by:

R - Ro (1 + λ V _eff ) (1)

Mean:

Ro the actual resistance value determined by the type and size of the semiconductor material,

λ is the coefficient of the voltage dependence and

V _ff a voltage resulting from the empirical relationship

^V eff - ^V e _P i - ^V RP ^{+ k V} R ⁽²⁾ can be derived.

In the latter equation:

V. the potential applied to the epitaxial layer, V _DD the most positive potential at the resistor (in this

Example is V-. = V. assumed),

V is the voltage drop across the resistor, i.e. variable potentials), and

V _R = V ₁ - V ₂ (V ₁ and V ₂ are essentially

a coefficient which essentially depends on the sheet resistance of the resistor region; the value of k has been found to vary between about 0.4 and 0.6. An approximate value of 0.5 can usually be used. V _ff and therefore the value R of the resistance are essentially dependent on the variable potentials V ₁ and V.

709848/0851

FR 975 303

If V _{ff is} constant, R is independent of the applied potentials V. and V ₂ . Such a result can be obtained by applying a potential to the epitaxial layer with the resistance, which is to be corrected for its voltage dependence, which satisfies the following relationship:

^v epi - \ <V ^V 2> ^{+ V} O ⁽³⁾

V _Q is the voltage required to block the PN junction 20 * between the P resistor region and the N epitaxial layer and the junction 20 already mentioned. If one replaces V. according to equation (3), and one sets

in equation (2) ' ^V RP = V ₁ , and k = 0.5, results themselves: ^V eff V ₁ + V 2> ^{+ V} 0 - V ₁ + O , 5 (\ and Consequently ^V eff ^{= v} o (4)

This shows that with this particular value of j

V. the voltage V-- independent of V_, namely that across j

voltage applied to the resistor. For the cons |

The stand value R then finally results: I.

R - R ₀ (1 + λ V ₀ ) - h R ₀ (5)

I where h means a constant that is independent of V. j

andV is ₂ . !

The voltage to be applied to the epitaxial layer must therefore follow the relationship according to equation (3). That is, the epitaxial layer must be connected to a voltage which changes in the same way as the average voltage across the resistor to be corrected. Since the potentials V ₁ and V _{2 are} applied to the resistor, there are two

709848/0851

FR 975 303

-H-

Procedures possible, depending on whether or not the correction of the voltage dependency is independent of the Resistance or, more specifically, is independent of the voltage across the resistor.

The first solution is shown in the first exemplary embodiment according to FIG. 2, where the circuit means 21 for correcting the voltage dependency of the resistor 10 are related to the latter. The circuit means 21 for said correction essentially comprise a resistor 22 and the bias voltage V_. To satisfy equation (3), the voltage drop across the resistor must be 22 volts. The voltages V_ + V ₁ and V _Q + V-, ie the potential difference V ₁ - V_, is therefore applied to the resistor 22. These voltages V ₁ and V »can either be provided independently of the identically designated voltages at the resistor to be corrected or else can be derived therefrom by suitable electrical circuits. For the correction of the voltage dependency, it is ultimately only necessary that a direct voltage of the corresponding polarity is applied to one end of the resistor 22. The resistor 22 is in turn equipped with a tap, for example with a center tap, assuming that the value k is approximately 0.5. j The resistor 22 does not have to be on the same half

printed circuit board be accommodated, it can for example and advantageously also as simultaneously with the resistor 10 implanted resistor can be arranged in an area isolated therefrom.

The further solution is provided by the in FIGS. 3A, 3B and 3D illustrated embodiments shown.

In the arrangement according to FIG. 3A, the voltages across the resistor 10 are passed through a pair of amplifiers 33, 34 which are in the

709848/0851

FR 975 303

in the simplest case can consist of emitter follower stages, applied to the resistor 22 '. These amplifiers have a voltage gain of 1 and practically do not mean any additional current flow. The circuit device 21 * for correcting the voltage dependency also contains a direct voltage source 35 for V _Q which is connected between the center tap of the resistor 22 * and the ohmic contact 19.

Mainly because of the amplifiers 23 and 24 are with this embodiment, however, because of the greater complexity and switching speed reduction are associated with some disadvantages.

A simplified embodiment is shown in Fig. 3B. As in the past, the task here is also to reduce the fluctuations in the potential difference between the resistance region and the epitaxial layer, but in contrast to the solutions shown in FIGS. 2 and 3A, a compromise is made here. As can be seen from Fig. 3B, the terminal 19 is connected to the most positive voltage of V ₁ and V-. In other words, the voltage V is applied accordingly. = V ₁ instead of the voltage according to equation (3). With Vpp »V ₁ and k ■ 0.5 we get according to equation (2)

V as O 5 V (6)

^v eff ^{υ>: 3 V} R ^l '

The resistance R is now defined by the relationship:

R »R ₀ (1 + x | r) (7)

R thus increases with the voltage across the resistor. In this case, the correction of the voltage dependency is not complete, but only improved. Such a solution is entirely acceptable if the potential difference V- at the resistor is low and the value

709848/0851

FR 975 303

(V ₁ + V _) / 2 large 1st. This solution is particularly attractive due to its simplicity. It can be combined in an advantageous manner with the solution according to FIG. 3C described below.

In the exemplary embodiment according to FIG. 3C, the resistor 10 with the value R is divided into two components 36 and 37, which are electrically isolated from one another and whose partial values are, for example, identical. The resistor 36 is connected with its terminal 19 * to the voltage V _{1 , while the voltage (V 1} + V ₂ ) / 2 is applied to the resistor part 37 and its ohmic contact 19 ″. The new value of the resistor 10 is now calculated as follows The following relationships can be established for the partial resistance 36 with the value R ^1:

^V eff - ° ' ⁵ < ^V 1 - ^VJ ^ T ¹⁾ '

^{Rl =} T ~

In the same way, the following applies to the partial resistance 37 with its value R ¹ ':

i as

* Ο λ ^V R

_R .. = _ο (ι + I -5)

The following applies in each case to V _R = V ₁ - V ₃ . The resistance value R results in the end as follows:

R = R ¹ + R = R ₀ (1 + I ^) (8)

If you compare this value with that according to Flg. 3B or

709fU8 / 0851

V + V V - V!

12 " _% _ ^ r. , ^V 1 ^V 2

FR 975 303

- Yl - I.

Speaking of equation (7), it can be seen that by; Dividing the resistance into two equal parts can halve _{the value of the coefficient of V R.} In general it can be said that if the resistance R is divided into η equal proportions according to the relation:

η
R » I R ₁

the coefficient for the voltage dependence only with the factor 1 / n comes into effect. It should be noted that this technique can be applied with advantage to the embodiments according to FIGS. 2 and 3A.

The embodiments according to FIGS. 3B and 3C are of interest for circuit applications with unipolar signals in which, for example, the potential V _{1 is} always greater than V ₂ .

If V _{1 is} both positive and negative relative to V_, ie in the case of bipolar signals, the potential difference V ₁ - V- = v_ must be kept very small, for example to a few hundred mV or less, so that the transition then tensioned in the forward direction conducts no significant current between the implanted: resistance region and the epitaxial layer.

An embodiment example which is even more advantageous for the case of bipolar signals is shown in FIG. 3D. The resistance 10 is again designed with a center tap, the potential of which is connected to the epitaxial layer via the connection 19 ′ ″ can be created. As a result, the following advantages can be achieved for operation with bipolar signals. On the one hand, there is a symmetrical arrangement for the positive and negative voltage periods, and secondly a complete correction of the influence of the chip

709848/0851

FR 975 303

Achievement dependency, since the following applies again:

Third, the value of the voltage that can be applied is doubled.

In other words, a perfect correction of the influence of the stress dependency is achieved in this case, because the distortion caused by the resistor 38 is compensated for by an equal but opposite one compensating distortion caused by resistor 39.

In the following, special applications are dealt with for a better understanding of the invention and the advantages that can be achieved therewith. First is the application of the first Embodiment described on a digital-to-analog converter (Fig. 4 and 5); Applications of the other exemplary embodiments on an operational amplifier are described in Connection with FIGS. 6A and 6B explained.

Fig. 4 shows a pair of conventional parallel connected digital-to-analog converters of the type in which the binary Gradation takes place through the use of an R-2R chain ladder network. For the sake of simplicity, the number of stages is each translator limited to 3; every converter is therefore able to convert 2 three-bit combinations from OOO to 111 to process. Such digital-to-analog converters offer considerable advantages over the types that work with weighted currents. They are fast (switching time approx 200 ns), very precise (all current sources are identical) and take up relatively little space on a semiconductor wafer a. Their main disadvantage, on the other hand, is that the linearity of their output signals is through the sensitivity to the voltage dependence of the resistors in the R-2R ladder network when in use

709848/0851

FR 975 303

adversely affected by ion implanted resistors; can be. This is because the sheet resistance of such ion-implanted resistors increases with the voltage drop across the resistors, which causes distortions j (Nonlinearities) are generated in the output signal of the relevant j converter.

You can find more detailed information on both types of converters for example in the article by G. Kelson et al with the title "A Monolithic 10-Bit Digital-to-Analog Converter Using Ion Implantation", published in the IEEE Journal j of Solid-state Circuits, Vol. SC-8, No. 6, December 1973, pages 396 to 403. j

A typical value for the extent of the voltage-dependent ι The speed in this case is 6400 ppm / V, with the epitaxial layer voltage being approximately 1 to 5 V. The regarding The digital-to-analog converter 40 corrected for its voltage dependence of the resistances consists of two iden- j table (partial) converters 41 and 43, each R-2R |

Use ladder networks. The first converter 41 acts as the main or main implementer. Its output connection is labeled 42. The second converter does the job the correction of the voltage dependence. The R-2R network of the main converter is connected to the reference potential 44 and that of the converter 43 are connected to a positive DC voltage at terminal 45. This positive DC voltage allows the corresponding biasing of the PN junctions (e.g. 20 and 20 ') for each resistor in the R-2R ladder network of converter 41.

Conventionally, in addition to the R-2R network, the converter 41 includes switches S1, S2 and S3 which are operated by the represent digital input signals controlled switching transistors. For an input signal with the binary access

709848/0851

FR 975 303

- 2O -

In order "1" the switch is open and accordingly at closed to an "O" input signal state. Furthermore are Power sources 11, 12 and 13 with the same power feeds j provided. The analog output current is weighted in accordance with the binary input signals. Any resistance in the network shown, whether it is implemented as a column resistance, e.g. r., or as a row resistance, e.g. r ...

! IK JK

electrically isolated from the other circuit components in a suitable manner, for example by the isolation zones of the type shown in FIG. 1. The electrical connection to the epitaxial layer is again designated by 19. Structure and The operation of this type of digital-to-analog converter is well known per se, see e.g. the aforementioned article by G. Kelson.

The R-2R network generates the weighted output current, which is then converted into an analog output current, for example by means of an operational amplifier (not shown).

! voltage can be converted.

The circuit of the converter 43 is identical to the converter 41; the corresponding current sources are ΙΊ, 1 * 2 and 1 * 3, which, together with the currents of the converter 41, _{follow the relationship I 4} = IJ. Correspondingly, the respective column and row resistances of the R-2R network apply

Correlation: ^r 4j, * ^r i _K and rv ^ lj. The current switches S'1, Is ¹ 2 and S'3 are activated simultaneously with switches S1, S2 j

and S3 actuated. In other words, it will be the same ! digital input signal to both converters 41 and 43 ^;

lays. j

In equation (2) above, a value of approximately k = 0.5 was chosen. In this particular application, this is j The value for k is the most advantageous from the point of view of symmetry. Depending on the binary input signal, the potential V- am

709848/0851

FR 975 303

Circuit point A in FIG. 4 may be higher or lower than the potential V _n at point B. This makes it necessary to divide the resistances, for example r ', into two equal partial resistances. In the same way, the column resistances, such as r ',, are also subdivided into two equal partial resistances. The resistances r | formed by means of ion implantation _k and r! _k with the values R or 2R are equipped with center taps which can each be connected to the terminal 19 of the corresponding resistor r .. or *. ,, whose voltage dependency is to be corrected. On,

this way becomes substantially the same potential difference is maintained between the epitaxial layer and each point of each resistor of the converter 41 so that the relationship

ν = ν + - t ν + v) epi ^V O 2 ^lv r ^v 2 ^;

I is met.

The properties achieved by applying the present invention to a high-precision 10-bit digital-to-analog converter are to be illustrated with the aid of the illustration in FIG. The difference between the ideal output voltage ^V OAi ^{below the} measured output voltage V _0AM ^{is shown in the} ordinate direction in relation to the least significant bit position (LSB). The maximum voltage deviations and their proportions are plotted in the direction of the abscissa. The output characteristic of an uncontrolled digital-to-analog converter is denoted by 50. A considerable improvement can be achieved if one applies the described correction of the voltage dependency, for example to a 4-bit converter (curve 51), to a 6-bit converter (curve 52) or to a 10-bit converter (curve 53). In the latter case, the influence of the voltage dependence can be seen to be one tenth

709848/0851

FR 975 303

to reduce. The corrected bit positions are in each case those with the highest significance ( denoted by MSB in the figures). In contrast, in the case of a resistor that has not been corrected with regard to its voltage dependency, the most positive voltage (+ V in FIG. 4) i
are on the epitaxial layer.

On the basis of FIGS. Finally, FIGS. 6A and 6B show how the invention is used to correct the voltage dependency a linear amplifier can be used, the two produced by ion implantation Contains resistors R1 and R2, which are designated in the drawing with 61 and 62, respectively.

In Fig. 6A, resistors 63 and 64 are used, the values of which are much larger than those of R1 and R2, so that the current drift in resistors 63 and 64 can be neglected. The actual operational amplifier is at 65 designated. There is a potential difference of V. across the resistor 61, since the potential of the circuit node E, which represents an input of the amplifier 65, is very close to the ground potential. To correct the influence of the voltage dependence of the resistor 61 it is necessary to to the epitaxial layer the potential according to the relationship:

V ₁ » j V _± + constant to be applied.

This is done via resistor 63, whose value 2R is much greater than the input impedance Z. of the (overall) amplifier 60. In addition, resistor 63 allows a direct voltage to be applied to the epitaxial layer with the value V_ / 2, for example 6V various PN transitions 20 and 20 * according to FIG

7098A8 / 0851

FR 975 303

corrective (semiconductor) resistors must be biased accordingly. In the same way is the influence of The voltage dependency of the resistor R2 is taken into account by the resistor 64 provided for this purpose.

Based on FIG. 6B, a modified method for correcting the voltage dependency of the Resistors R1 and R2 of the same linear amplifier 60 are explained, namely by using the j hang with Fig. 3C described embodiment. By Division of the resistors into three equal parts 61 ', j

61 "and 61" · or 62 ', 62 ", 62"' as well as by applying!

ι of the respective epitaxial layer areas to the most positive j Potential that occurs at the respective resistor, the influence of the voltage dependence of the resistors R1 and R2 can be reduced to a third. In the latter case | V. must be less than V _ref . j

! j

In the case of the FIGS. 6A and 6B are circuits shown j the resistors R1 and R2 have been corrected with regard to their voltage dependence. The fact that such a Input resistance R1 is present, is advantageous for some circuit applications. This results in the total

reinforcement of such a circuit from the ratio R2 to R1; this relationship is based on the one described Way to make it independent of the voltage sensitivity of the resistors used, if one applies the measures according to the invention to it.

7098A8 / 0851

Blank page

Claims (10)

FR 975 3O3 PATENT CLAIMS A method for correcting the voltage dependency of semiconductor resistors, which are formed as doping regions of a first conductivity type, preferably produced by means of diffusion or ion implantation, within isolated regions of an epitaxial layer of the opposite second conductivity type and the associated potentials V ₁ and V_ are applied to their outer connections, characterized in that a potential is applied to the epitaxial layer which represents a function of at least one of the potentials V ₁ or V ₂ . 2. The method according to claim 1, characterized in that the potential applied to the epitaxial layer V. by function
epi ^V epi - \ ⁽ VV ^{+ V} O is determined, where V is a constant DC bias means. 3.Circuit arrangement for correcting the voltage dependence of semiconductor resistors, which are formed as doping regions of a first conductivity type, preferably produced by means of diffusion or ion implantation, within isolated areas of an epitaxial layer of the opposite second conductivity type and the associated potentials V ₁ and V- are applied to their outer connections by a potential derived via further circuit means and applied to the epitaxial layer, which potential represents a function of at least one of the potentials V ₁ or V ₂ . 4. Circuit arrangement according to claim 3, characterized by a potential applied to the epitaxial layer ORIGINAL INSPECTED PR 975 303 - 2 -V. according to the relationship ^v epi 2 ^{l V} 1 V ^v 0 where V _{Q means} a constant DC bias. 5. Circuit arrangement according to at least one of claims 3 and 4, characterized in that the further circuit means biasing devices and at least one further resistor with a Include center tap. 6. Circuit arrangement according to at least one of claims 3 to 5, characterized in that one direct connection between the outer connection of the epitaxial layer and the resistor connection with the most positive circuit potential is provided. 7. Circuit arrangement according to at least one of claims 3 to 5, characterized in that the further circuit means a direct connection to the external connection of the epitaxial layer and an approximately average tap of the cons to be corrected with regard to its voltage dependency embrace. 8. Circuit arrangement according to at least one of the preceding claims, characterized in that the resistance R to be corrected with regard to its voltage dependence is structurally equal in η Partial resistances R. according to the relationship i = n R = R ₁ is divided. ORIGINAL INSPECTED FR 975 3O3
/ 1 U 6 L) J 9. Use of the method or use of the circuit arrangement according to one of the preceding claims in a digital-to-analog converter of the type in which the binary gradation is carried out by using an R-2R ladder network, the resistances of which increase with regard to their voltage dependency are corrected, with at least two circuit-wise identical converter circuits parallel to one another are provided and received the same digital input signals supplied, and wherein the resistors in the ladder network of one converter are provided with a center tap, each with the connection for the epitaxial layer area of the associated resistance of the ladder network of the other converter are connected. 10. Use of the method or use of the circuit arrangement according to one of claims 1 to 8 for Correction of the voltage dependency of the external resistance circuit of an operational amplifier.