BACKGROUND OF THE INVENTION

The invention relates to a regenerator circuit for CCD arrangements operating in accordance with the CCD principle (charge-coupled-device principle).

Several types of regenerator circuits for CCD arrangements are known. As described in the publication, W. F. Kosonocky, J. E. Carnes, "Charge Coupled Digital Circuits," IEEE Journal of Solid-State Circuits, Vol. SC-6, No. 5, October 1971, p. 314-322 in the one type of regenerator circuits, in addition to the charge representing the item of information, the so-called basic charge is also input. In digital application, for example, a charge Q.sub.1 corresponds to the binary "1," and a charge Q.sub.0 corresponds to the binary "0." In this operating mode long regeneration intervals can be achieved as a result of reduced transmission losses.

A disadvantage consists, however, in that a considerable circuitry amount is required for the determination of the two different quantities of charge Q.sub.0 and Q.sub.1 which are both unequal to 0, in the regenerator circuits. Therefore, only a comparatively coarse CCD draft raster can be achieved with this process.

Another type of known regenerator circuit is particularly suitable for digital operation. In this case, only with one of the two binary signals is charge input into the following CCD stage, so that, for example, the input charge Q.sub.1 corresponds to a binary "1," and the charge Q.sub.0 = 0 corresponds to a binary "0." Although with this principle the higher transmission losses means that only short regeneration intervals can be achieved, the smaller circuitry amount means, however, that a considerably higher packing density of the overall circuit can be achieved.

A disadvantage of these circuits consists, however, in the fact that the component straying on the semiconductor chip considerably affects the function.

SUMMARY OF THE INVENTION

The aim of the present invention consists, accordingly, in providing a regenerator circuit as described above, for short regeneration intervals, in which only the component straying of the components which are adjacent on the semiconductor chip in a regenerator circuit affects the function.

This aim is realized by a regenerator circuit for CCD arrangements comprising a first and second data storage device; a first capacitance and means to charge said first capacitance to a known potential; a second capacitance and means to charge said second capacitance to a known potential; means to detect the absence of charge at the output terminal of said first data storage device and upon detecting said absence of charge not discharge said first capacitance; gate means at the input to said second storage device detecting the presence of charge on said first capacitance and the injection of charge from said second capacitance into the input terminal of said second data storage device being blocked; means to detect the presence of charge at said output terminal of said first data storage device and upon detecting such presence said first capacitance being discharged upon said gate means detecting the absence of charge on said first capacitance and the injection of charge from said second capacitance into said second storage device being blocked.

An essential advantage of the invention consists in that the reference potentials which are essential to the function are in each case produced by the regenerator circuit itself.

In the following, the invention will be explained in detail, making reference to the description and the Figures.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 schematically illustrates a known regenerator circuit, in which not only the charge representing the information, but also the so-called basic charge is input;

FIG. 2 schematically illustrates a regenerator circuit in accordance with the invention;

FIG. 3 shows the pulse train program for the circuit corresponding to FIG. 2;

FIG. 4 shows the circuit corresponding to FIG. 2 and the associated potential well model;

FIG. 5 illustrates a further development of the regenerator circuit in accordance with the invention;

FIG. 6 shows the pulse train program relating to the circuit in FIG. 5;

FIG. 7 illustrates a further development of the regenerator circuit in accordance with the invention with two pulse train supply lines;

FIG. 8 illustrates a further development of the regenerator circuit in accordance with the invention with four drive lines;

FIG. 9 illustrates another further development of the regenerator circuit in accordance with the invention;

FIG. 10 shows the pulse train program relating to the circuit in FIG. 9;

FIG. 11 also shows another further development of the regenerator circuit in accordance with the invention; and

FIGS. 12 to 17 each show an interconnection of a plurality of regenerator circuits in accordance with the invention.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

In FIG. 1, the CCD output stage is referenced 1, and the CCD input stage is referenced 3. The known regenerator amplifier 2 is arranged between the two stages. In this arrangement, the output diffusion zone 11 of the output stage 1 is connected via the line 21 to the gate electrode 31 of the input stage 3. The electrode 31 represents the gate electrode of the input transistor T.sub.E of the input stage 3. Here, this input transistor T.sub.E consists of the diffusion zone 32 acting as source and the gate electrode 31. The potential sink beneath the first CCD shift electrode 35 acts as drain of the transistor T.sub.E, where the electrodes 31 and 35 possess the spacing which is characteristic of a CCD arrangement. 12 designates the overall capacitance of the output diffusion zone 11 of the output stage 1, of the line 21 and of the gate electrode 31. The capacitance of the diffusion zone 32 is referenced 33. Before the information transmission from the output CCD arrangement 1 to the input CCD arrangement 3, the capacitor 12 is pre-charged to a reference potential U.sub.A = U.sub.1 and the input diffusion zone 32 of the input stage 3 with the capacitance 33 is pre-charged to the reference potential U.sub.E = U.sub.2. For this purpose the transistor 22 is switched conductive via the potential .phi. 22 at its gate terminal 222, so that the potential U.sub.1 present across its terminal 221 reaches the diffusion zone 11. In addition, the transistor 23 is switched conductive by the potential .phi. 23 which is present at its gate terminal 232, so that the potential U.sub.2 present at its terminal 231 passes to the diffusion zone 32. Here the potentials U.sub.1 and U.sub.2 obey the equation:

U.sub.1 - U.sub.2 = U.sub.A - U.sub.E = U.sub.TE + .delta.U.

here U.sub.TE is the start voltage of the input transistor T.sub.E of the input stage 3. .delta.U is advantageously contrived to be such that it is equal to half the signal amplitude .DELTA.U.sub.A /2 which arises when a charge Q.sub.1 is transmitted to the output diffusion zone 11, where

.DELTA.U.sub.A = Q.sub.1 /C.sub.12 = 2 .delta. U (1)

when the charge Q.sub.1 arrives at the output diffusion zone 11, the input transistor T.sub.E of the input stage 3 is not switched conductive so that no charge Q.sub.3 is input in the input CCD circuit 3. If a charge Q.sub.1 fails to arrive in the output diffusion zone 11 of the output arrangement 1, the input transistor T.sub.E of the input stage 3 remains switched conductive on the other hand, so that the charge Q.sub.3 is input in the CCD input stage 3. Thus, the circuit has an inverting operation.

The essential disadvantage of this circuit consists in that the potentials U.sub.1 and U.sub.2 for the plurality of regenerator circuits arranged on a semiconductor chip are only provided once. Due to the component straying on the semiconductor chip which, for example, contains a fluctuation in start voltage of .DELTA.U.sub.TE which can easily reach the magnitude of the voltage difference .delta.U, the equation (1) is not fulfilled at the location of various regenerator circuits on the semiconductor chip, which as a rule means a breakdown of the overall circuit.

FIG. 2 illustrates a regenerator circuit in accordance with the invention. Here details of the Figure which have already been described in association with the known circuit shown in FIG. 1 bear the corresponding references. The regenerator circuit in accordance with the invention is referenced 4. In this circuit, the reference potentials U.sub.1 and U.sub.2 are produced in accordance with the equation (1) in each case directly at the location of the regenerator circuit 4. The advantage is thus achieved that only the component straying of the directly adjacent transistors in the regenerator circuit affects the function. In the circuit shown in FIG. 2, the reference voltages U.sub.1 and U.sub.2 are produced with the aid of the transistors 42 to 45. Here the transistor 45 is connected in the manner shown in the Figure on the one hand to the diffusion zone 32 of the input stage 3 and on the other hand to the line 451 which preferably is connected to ground. The gate terminal 452 of the transistor 45 is connected to the line 453, to which the potential .phi..sub.K can be connected. The transistor 42 is connected on the one hand to the diffusion zone 11 of the output stage 1 and on the other hand to a line 421. This line can be connected to the reference potential .phi..sub.R. This line 421 is also connected to the gate terminal 422 of the transistor 42. This causes this transistor to act as load element. Similarly, the transistor 43 is connected on the one hand simultaneously by its gate terminal 431 to the line 421 and on the other hand to the transistor 44 which is likewise connected as load element. This transistor 44 is on the one hand, together with its gate terminal 441, connected to the transistor 43 and on the other hand is connected to the input diffusion zone 32 of the input stage 3. The output diffusion zone 11 is connected via the connection line 41 to the input diffusion zone 32.

Between the line 421 and the diffusion zone 32 is arranged the coupling capacitance 46, the dimensioning of which will be described in detail further in the description.

In the following, the function of the regenerator circuit corresponding to FIG. 2 will now be explained in association with the pulse train program corresponding to FIG. 3 and the potential well model corresponding to FIG. 4. Here in the potential well model shown in FIG. 4, for the individual times the potentials .phi..sub.S are given for the various locations of the input stage and output stage.

In the time interval between t0 and t1, charge is displaced from the electrode 14 to the electrode 13 as a result of the connection of the pulse train .phi. 13 which can be connected to the electrode 13 of the output stage, and by disconnecting the pulse train .phi. 14 which is connected to the electrode 14 of the output stage 1. At the time t1 the charge .phi..sub.1 is localized in the potential well located beneath the electrode 14 (time t1 in FIG. 4).

At the times t1 and t2 the pulse train .phi..sub.K is switched on and subsequently switched off again. This causes the capacitor 33 to be connected to ground via the transistor 45 and thus to be discharged.

At the following times t3 and t4, the pulse train .phi..sub.R with the amplitude U.sub.R is switched on and off again. This means that at the time t4 a voltage

U.sub.A = U.sub.1 = U.sub.R - U.sub.T42 (2)

drops across the capacitor 12, where U.sub.T42 is the start voltage of the transistor 42. At the same time the voltage

U.sub.E = U.sub.2 = U.sub.R - U.sub.T43 - U.sub.T44 - .delta.U (3)

is produced across the capacitor 33, via the transistors 43 and 44, where U.sub.T43 is the start voltage of the transistor 43, and U.sub.T44 is the start voltage of the transistor 44. The voltage change .delta.U is capacitively input-coupled via the capacitor 46 at the time t4 on the disconnection of the pulse train .phi..sub.R, and here the magnitude of the coupling capacitance of the capacitor 46 can be determined in accordance with the equation ##EQU1## In this equation (4), U.sub.R signifies the amplitude of the pulse train .phi..sub.R. The start voltages of the directly adjacent transistors 42, 43, 44 and T.sub.E in the circuit are equal in first approximation, so that from the equations (2) and (3) one obtains:

U.sub.2 = U.sub.R - U.sub.T43 - U.sub.T44 - .delta.U = U.sub.1 - U.sub.TE - .delta.U (5)

accordingly, the circuit operates independently of the magnitude of the amplitude U.sub.R and thus also independently of fluctuations in supply voltage.

At the time t4 the circuit is in the reference state and the input transistor T.sub.E of the input stage 3 is conductive.

If, then, at the time t5, the pulse train .phi. 13 is switched off, the charge Q.sub.1 passes to the output diffusion zone 11 and gives rise to a negative voltage change .vertline..DELTA.U.sub.A .vertline., so that the input transistor T.sub.E of the input stage 3 blocks. The potential well formed beneath the electrode 34 as a result of the connection of the pulse train .phi. 34 at the time t6 remains empty, even when the pulse train .phi. 35 is connected to the electrode 35. If, on the other hand, no charge Q.sub.1 reaches the output diffusion zone 11 at the time t5, the voltage U.sub.12 across the capacitor 12 remains unchanged and the transistor T.sub.E of the input stage 3 remains conductive. This results in the fact that at the time t7, charge flows into the potential well located beneath the electrode 34. This is schematically illustrated in FIG. 4 for a time t7'.

At the following times t8 and t9, on the one hand the charge is transported on to the electrode 36, and on the other hand the next reference state is prepared, where in FIG. 3 the times t0', t1', t2', etc. correspond to the times t0, t1, t2, etc.

In the further development of the regenerator circuit of the invention illustrated in FIG. 5, one transistor less is required than in the regenerator circuit of the invention shown in FIG. 2. Details of the circuit in FIG. 5 which have already been mentioned in association with the other Figures bear the corresponding references. In FIG. 5, the regenerator circuit is referenced 5 and in contrast to the regenerator circuit 4 offers the advantage of a smaller space requirement.

The transistor 52 is connected on the one hand to the output diffusion zone 11 and on the other hand to the line 521. The gate electrode 522 is operable via the line 523. The transistor 53 is on the one hand connected with its gate terminal 531 to the line 521 and on the other hand to the input diffusion zone 32. The transistor 55 is connected on the one hand to the input diffusion zone 32 and on the other hand to the line 521. The gate terminal 551 is operable via the line 552. The coupling capacitance 56 is arranged between the line 521 and the input diffusion zone 32. The diffusion zone 11 and the electrode 31 are connected to one another via the line 51.

In the following, the function of the regenerator circuit 5 in accordance with the invention will be explained making reference to FIGS. 5 and 6. The reference state is set up in that similarly to the circuit explained above in association with FIGS. 2, 3 and 4, first at the time t1, the capacitor 33 is discharged via the transistor 55. For this purpose, the latter is switched conductive by the pulse train .phi..sub.K connected to the line 552, and at the time 52 is switched nonconductive again.

In the time interval from t3 to t6, the reference potentials U.sub.1 and U.sub.2 are set up, where U.sub.1 = U.sub.R2 (5) and U.sub.2 = U.sub.R2 - U.sub.T53 - .delta.U (6). In the equation (5) U.sub.R2 is the amplitude of the pulse train .phi..sub.R2 which is connected to the line 521. In operation it must be ensured that the amplitude U.sub.R1 of the pulse train .phi..sub.R1 connected to the line 523 fulfills the equation U.sub.R1 .gtoreq. U.sub.R2 + U.sub.T52 (7), in order that the equation (5) is fulfilled.

FIG. 7 illustrates a further development 6 of the invention. Details of FIG. 7 which have already been described in association with the other Figures bear the corresponding references. In the case of the regenerator circuit 6 shown in FIG. 7, advantageously it is unnecessary to pay attention to the equation (7). Only the two drive lines 651 and 652 are provided. The transistors 62, 63 and 64 are arranged in the same way as the transistors 42, 43 and 44 already described in association with FIG. 2. The capacitor 66 is connected on the one hand to the line 651 and on the other hand to the diffusion zone 32. The transistor 65 is connected on the one hand to the line 651 and on the other hand to the diffusion zone 32. It is operable via the line 652 which is connected to its gate electrode 653. The diffusion zone 11 is connected via the line 61 to the electrode 31. The function of this circuit corresponds to the pulse train program represented in FIG. 3, the reference state being first set up in accordance with the equation (5).

FIG. 8 illustrates a further development 7 with four drive lines. Details of FIG. 8 which have already been described in association with the other Figures bear the corresponding references. The transistors 72, 73 and 74 again correspond to the arrangement of the transistors 42, 43 and 44 in FIG. 2. The capacitor 75 is connected on the one hand to the diffusion zone 32 and on the other hand to the drive line 761. The diffusion zone 11 is connected via the line 71 to the electrode 31. The transistor 75 is connected on the one hand to the diffusion zone 32 and on the other hand to the drive line 751, which is preferably connected to ground. Its gate 752 is operable via the line 753. The reference potential .phi..sub.R can be connected via a separate line 721. The function is again similar to the pulse train program already represented in FIG. 3, the reference state in each case being first set up in accordance with the equation (5).

The pulse train .phi..sub.C is connected via the line 761 to the capacitor 76, although this line does not lead to a further point of the circuit. It is thus possible to individually match the coupling voltage .delta.U (equations 1, 3, 4, 5, 6 and 7), which is of significance, for example, for experimental purposes.

With the aid of the circuit variants 4 and 5 (FIGS. 2 and 5), 6 (FIG. 7) and 7 (FIG. 8), it has been shown that the regenerator circuits in accordance with the invention can be optionally operated with three, two or also four discrete drive lines.

FIG. 9 represents a further development 8 of the regenerator circuit in accordance with the invention. Details of FIG. 9 which have already been described in association with the other Figures, bear the corresponding references. The transistor 82 is operable via a separate line 821 which is also connected to the latter's gate terminal 822, with the pulse train .phi..sub.L1. The transistors 83 and 84 are connected in the same manner as the transistors 43 and 44 in FIG. 2 and are operable via the line 831 by the pulse train .phi..sub.L2. The transistor 85 is connected on the one hand to the diffusion zone 32 and on the other hand to a line 851 which is preferably connected to ground. The gate 852 of this transistor is controllable via the line 853 by the pulse train .phi..sub.K. The amplitudes U.sub.L1 and U.sub.L2 of the pulse trains .phi..sub.L1 and .phi..sub.L2 are determined in accordance with the equation

U.sub.L2 = U.sub.L1 - .delta.U = U.sub.L1 - (.DELTA.U.sub.A /2) (8)

instead of the capacitive input-coupling of the voltage .delta.U, which is of decisive importance for the setting-up of the reference state, in this further development 8, the voltage .delta.U corresponding to equation (8) is set up by the difference between the amplitudes U.sub.L1 and U.sub.L2. Advantageously, therefore, in the further development 8, the dimensioning of the coupling capacitance 46, 57 and 66 of the circuits corresponding to FIGS. 2, 5 and 7 is avoided. Instead, the pulse generator which, for example, is integrated on the semiconductor chip and which supplies the pulse trains .phi..sub.L1 and .phi..sub.L2 for all the regenerator circuits is dimensioned in accordance with equation (8).

FIG. 10 illustrates the pulse train program relating to the circuit 8 corresponding to FIG. 9. The function, in detail, is that in the interval of time between t1 and t6 the reference state is set up by the switching on and off of the pulse trains .phi..sub.K, .phi..sub.L1 and .phi..sub.L2. The ground line 851 can, similarly to the circuit corresponding to FIG. 3, be spared if the gate 852 of the transistor 85 is connected to the pulse train line 821 or 831.

Similarly, to the circuit shown in FIG. 5, the possibility exists of sparing the transistor 84. This leads to the circuit variant 9 corresponding to FIG. 11. Details of this Figure which have already been explained in association with the other Figures bear the corresponding references.

The transistor 92 is connected on the one hand to the diffusion zone 11 and on the other hand to the drive line 922 (potential .phi..sub.L2). Its gate terminal 923 is connected to the line 921 (potential .phi..sub.L3). The transistor 93 is connected on the one hand together with its gate terminal 931 to the line 932 (potential .phi..sub.L1) and on the other hand to the diffusion zone 32. The transistor 95 is connected on the one side to the line 951 which preferably carries ground potential and on the other side to the diffusion zone 32. The gate terminal 952 of the transistor 952 is connected to the line 953 (potential .phi..sub.K). The diffusion zone 11 is connected via the line 91 to the electrode 31.

On operation, to ensure the satisfactory functioning of the circuit, the fundamental equation U.sub.L3 = U.sub.L1 + U.sub.T1 must be taken into consideration. Here U.sub.L1 is the amplitude of the pulse train .phi..sub.L1 connected to the line 921 and U.sub.L3 is the amplitude of the pulse train .phi..sub.L3 across the line 922.

In the described exemplary embodiments, in the field effect transistors connected as load elements, the gate electrode is in each case connected to the drain electrode. The field effect transistors are transistors of the enhancement type. An advantage of such a circuit consists in the omission of a line for drive purposes.

The gate electrodes of the field effect transistors which serve as load elements can also be pulsed. Such a method of operation offers the advantage of a lower power loss and a lower drive energy.

Finally, the load elements can also be field effect transistors of the depletion type. In this case, the gate electrode is connected to the source electrode. Such a circuit provides the advantage of shorter switching times.

Advantageously, for the purpose of space saving, it is possible to in each case provide the transistors 93 and 95 just once for a number, for example, N = 5, 10, 20, etc., CCD circuits, as schematically illustrated in FIG. 12. This is conditional upon the parameter fluctuations on the chip so permitting. In FIG. 12, this is illustrated for the circuit in accordance with the invention corresponding to FIG. 11. Details of FIG. 12 which have already been described in association with the other Figures bear the corresponding references. The CCD arrangements CCD 1, CCD 3, . . . to CCD 9, designated with odd-numbered numerals are arranged on the output side. The CCD circuits CCD 2, CCD 4, . . . CCD 10 designated with even numerals are arranged on the input side. The diffusion zones 11 of the output-CCD-arrangements with the odd numerals are connected in the manner shown in the Figure via a transistor 92 to the lines 921 and 922. To save space, the transistors 93 and 95 are provided only once and connected via a line 954 to the diffusion zones 32 of the input-CCD arrangements bearing the even numbers.

The centralization which has been represented for the circuit 9 in FIG. 11 is also both possible and effective for the other regenerator circuits in accordance with the invention.

FIGS. 13 to 17 represent the centralizations of the regenerator circuits 4 to 8 shown in FIGS. 2, 5, 7, 8 and 9. In these Figures, for the sake of simplicity, the CCD arrangements have not been shown. Details of FIGS. 13 to 17 which have already been explained in association with one of the FIGS. 2, 5, 7, 8 and 9, have in each case been referenced accordingly.

All the circuit variants of the new regenerator circuit in accordance with the invention can be used both for two-phase, three-phase and four-phase CCD arrangements, the drive means represented in FIG. 3 being retained accordingly.

Preferably, the regenerator circuits in accordance with the invention are constructed, together with the CCD arrangements, in a silicon-gate technique.

It will be apparent to those skilled in the art that many modifications and variations may be effected without departing from the spirit and scope of the novel concepts of the present invention.