Signal Transfer System of Charge Transfer Device With Charge Retaining Clocking Providing Fixed Transfer Time Within Variable Trigger Pulse Time Period

This invention relates to a signal transfer system for a charge transfer device (hereinafter abbreviated CTD).

As an example of such a CTD, a charge coupled device (hereinbelow abbreviated CCD) employing three-phase or four-phase drive is known.

The CCD is constructed such that an input source electrode and an output drain electrode are provided in a substrate, a large number of transfer electrodes are arranged on the substrate between the source and drain electrodes through an insulating film, and three or four of the transfer electrodes constitute one stage. By successively applying three-phase or four-phase clock pulses to the transfer electrodes, charges in capacitors formed under the respective transfer electrodes are successively shifted and are fed to subsequent stages.

Such a CTD becomes a variable delay line by varying the clock frequency. The variable delay line can be used for compensating the time axis fluctuations of a recording appliance or equipment having a mechanical rotation system, such as video tape recorder and magnetic disc. When the CTD is operated by making the clock frequency during writing different than during reading, it can subject a signal to time compression or time expansion. Various applications of the CTD are considered in this manner, and for the time axis transforming operation as stated above, the clock must be frequency-modulated. When the clock frequency changes in this way, the transfer time of charges similarly changes. A change of the transfer time leads, in itself, to a change of the transfer efficiency. During the time axis transforming operation, the frequency characteristic changes with time. The operation has thus become inconvenient in the prior art.

The prior art has also had the disadvantage that a circuit for generating clock pulses corresponding to the respective clock frequencies becomes very complicated.

The principal object of this invention is to provide a signal transfer system for a charge transfer device which always maintains the transfer efficiency constant and makes a stable operation possible.

Another object of this invention is to provide a signal transfer system which is constructed of simple means.

In addition to the foregoing feature of transferring a signal, the CTD has the capability of storing the signal. That is, when a state, in which, among the plurality of transfer electrodes constituting one stage, a certain electrode has a predetermined high potential applied thereto and the other electrodes have a different low potential applied thereto, is maintained for a fixed time, charges remain beneath the electrode with the predetermined potential applied and the signal is accumulated for the fixed time.

Upon taking note of such an accumulative function, this invention is characterized in that clock pulses for transferring charges at a fixed transfer time are sequentially applied to respective transfer electrodes for a fixed period within a variable period of time corresponding to a delay time. A clock pulse for retaining transferred charges beneath a predetermined transfer electrode is applied to the transfer electrode for the remaining period within said variable period of time.

FIG. 1 is a diagram showing the structure of a CCD having a three-phase drive.

FIG. 2 is a waveform diagram for explaining the operation of the device in FIG. 1.

FIGS. 3a, 3b and 3c are diagrams for explaining the states of signal transfer of the device in FIG. 1.

FIG. 4 is a diagam showing the structure of a CCD having four-phase drive.

FIG. 5 is a waveform diagram for explaining the states of the operation of the device in FIG. 4.

FIGS. 6a, 6b, 6c and 6d are diagrams for explaining the states of signal transfer of the device in FIG. 4.

FIGS. 7 and 8 are waveform diagrams of clock pulses in the prior art as applied to the devices in FIGS. 1 and 4, respectively.

FIG. 9 is a block diagram of an embodiment of a circuit which generates clock pulses for use in the signal transfer system of this invention.

FIGS. 10a -10e are diagrams of waveforms at various parts in FIG. 9.

FIG. 11 is a block diagram of another embodiment of the circuit which generates clock pulses for use in the signal transfer system of this invention.

FIGS. 12a-12d are diagrams of waveforms at various parts in FIG. 11.

FIG. 13 is a block diagram of still another embodiment of the circuit which generates clock pulses for use in the signal transfer system of this invention.

FIG. 14 is a diagram of waveforms at various parts in FIG. 13.

FIG. 15 is a block diagram of yet another embodiment of the circuit which generates clock pulses for use in the signal transfer system of this invention.

FIG. 16 is a connection diagram of an example of a D.C. shift circuit for use in the circuits in FIGS. 13 and 15.

FIG. 1 shows the structure of a CCD of the three-phase drive. A P-type semiconductor substrate 1 is provided with a source electrode for input semiconductor region 2 and a drain electrode for output semiconductor region 3, while the first-, second- and third-phase transfer electrodes 5, 6 and 7 are arranged on the P-type substrate 1 through an insulating film 4. One stage is constructed of the three transfer electrodes 5, 6 and 7, and a large number of such stages are connected in cascade. Regions 2 and 3 are of N-type conductivity.

In FIG. 2, symbols .phi..sub.1, .phi..sub.2 and .phi..sub.3 indicate the waveforms of clock pulses which are respectively applied to the first-. second- and third- phase transfer electrodes 5, 6 and 7 in FIG. 1.

By sequentially impressing such clock pulses .phi..sub.1, .phi..sub.2 and .phi..sub.3 on the transfer electrodes 5, 6 and 7 in FIG. 1, input charges supplied to the source electrode 2 are sequentially transferred to capacitors under the respective transfer electrodes, and they are delivered as an output from the drain electrode 3.

FIGS. 3a, 3b and 3c illustrate the states of signal transfer of the device in FIG. 1, which states correspond to potential wells at times t.sub.1, t.sub.2 and t.sub.3 in FIG. 2, respectively.

More specifically, at the time t.sub.1, the potential of the transfer electrode 5 is the highest and charges are accumulated in the capacitor beneath it as illustrated in FIG. 3a. At time t.sub.2, the potential of the transfer electrode 6 becomes the highest and the charges under the transfer electrode 5 are shifted to the capacitor beneath the transfer electrode 6, as illustrated in FIG. 3b. At time t.sub.3, the potential of the transfer electrode 7 becomes the highest and the charges are shifted beneath the electrode as illustrated in FIG. 3c. Consequently, after the time of one period of the clock pulses, the charges are transferred in the direction of arrows in FIG. 3 to the amount of three electrodes or one stage. In this way, the charges injected into the source electrode 2 are transferred by the electric fields along the interface between the insulating film 4 and the P-type substrate 1, and they are sequentially extracted as an output signal from the drain electrode 3.

FIG. 4 shows the structure of a CCD having four-phase drive. A P-type substrate 1 is provided with an input source electrode 2 being an N-type diffused layer and an output drain electrode 3 being an N-type diffused layer. On the P-type substrate, a second-phase transfer electrode 6 and a fourth-phase transfer electrode 8 are formed through a thin insulating film 4. Further, in the insulating film 4, a first-phase transfer electrode 5 and a third-phase transfer electrode 7 are formed in a manner to correspond to the interspaces between second- and fourth-phase transfer electrodes 6 and 8.

.phi..sub.1, .phi..sub.2, .phi..sub.3 and .phi..sub.4 in FIG. 5 denote the waveforms of the first-, second-, third- and fourth-phase clock pulses which are respectively impressed on the electrodes 5, 6, 7 and 8 in FIG. 4. The clock pulses are rectangular waves and, letting the phase of the first-phase clock pulse be 0 radians, the second phase is shifted by .pi./2 radians, the third phase by .pi. radians and the fourth phase by 3/2 .pi. radians. In addition, the first-phase clock pulse and the third-phase clock pulse change from O to +V.sub.c, while the second-phase clock pulse and the fourth- phase clock pulse change from +.DELTA. V to +.DELTA. V + V.sub.c. In other words, the second-phase clock pulse and the fourth-phase clock pulse undergo a D.C. level shift by .DELTA. V as compared with the first-phase clock pulse and the third-phase clock pulse. The quantity of the D.C. shift (.DELTA.V) can be arbitrarily selected in the range O .ltoreq. .DELTA. V < + V.sub.c. The amplitudes of the clock pulses of the respective phases may differ to some extent, and need not coincide for all phases.

The transfer operation of the four-phase drive CCD in FIG. 4 will now be explained in detail with reference to FIGS. 6a to 6d. Consider the state of a period t.sub.1 in FIG. 5. Then, as illustrated in FIG. 6a, the potential wells are created in correspondence with the applied potentials, and electrons (signal charges) exist beneath the fourth-phase electrodes 8 to which the clock pulse .phi..sub.4 of the highest potential is impressed. In the subsequent period t.sub.2, the potential of the second-phase electrodes 6 with the clock pulse .phi..sub.2 impressed thereon and that of the fourth-phase electrodes 8 with the clock pulse .phi..sub.4 impressed thereon are reversed (FIG. 6b), so that the signal charges pass through the first-phase electrodes 5 with the clock pulse .phi..sub.1 impressed thereon and they are transferred beneath the second-phase electrodes with the clock pulse .phi..sub.2 impressed thereon. During period t.sub.3, the potential of the first-phase electrodes 5 with the clock pulse .phi..sub.1 impressed thereon and that of the third-phase electrodes 7 with the clock pulse .phi..sub.3 impressed thereon are reversed, and the position at which each channel is cut off moves from beneath the third-phase electrode 7 with the clock pulse .phi..sub.2 impressed thereon to beneath the first-phase electrode 5 with the clock pulse .phi..sub.1 impressed thereon. The signal charges, however, remain beneath the second-phase electrodes 6 to which the clock pulse .phi..sub.2 of the deepest potential well is impressed, and are not transferred (FIG. 6c). During period t.sub.4, as illustrated in FIG. 6d, the potential of the second-phase electrodes 6 with the clock pulse .phi..sub.2 impressed thereon and that of the fourth-phase electrodes 8 with the clock pulse .phi..sub.4 impressed thereon are reversed again, and the signal charges are transferred beneath the fourth-phase electrodes 8 with the clock pulse .phi..sub.4 impressed thereon. The same operation is thereafter repeated, and the signal charges are successively transferred from the source electrode 2 to the drain electrode 3.

As is apparent from FIGS. 5 and 6, the signal charges are transferred only during the period t.sub.2 and the period t.sub.4, that is, only when the potentials of the second-phase electrodes 6 and the fourth-phase electrodes 8 are reversed. For period t.sub.1 or period t.sub.3, the charges are accumulated and retained beneath either the second-phase electrodes 6 or the fourth-phase electrodes 8 and are not transferred. The signal charges are never accumulated and retained beneath the first-phase electrodes 5 or the third-phase electrodes 7, and both electrodes 5 and 7 merely effect a switching operation.

In the waveforms in FIG. 5, the transfer time of charges is the period t.sub.2 or t.sub.4. The charge transfer time t.sub.s is represented by: ##EQU1## where f.sub.c denotes the clock frequency.

A rotary recording appliance or equipment such as video tape recorder and magnetic disc has a mechanical rotation system, and therefore gives rise to a time axis fluctuation (jitter). In order to compensate for this, the time axis transformation can be made by the use of the CTD.

When the CTD is operated in such a manner that the clock frequency during writing and the clock frequency during reading are caused to differ, it can subject a signal to time compression or time expansion.

In order to perform such a time axis transforming operation, it has hitherto been contrived that, as illustrated in FIGS. 7 and 8, the frequencies of the clock pulses are varied in correspondence with delay states.

However, when the clock frequency is thus varied to change the charge transfer time, there are drawbacks in that the transmitting efficiency changes and that the frequency characteristic changes. A further disadvantage is that circuits for generating the clock pulses corresponding to the respective trigger pulse periods become complicated.

FIG. 9 shows an embodiment of a circuit for generating the clock pulses for use in the present invention. Numeral 9 designates an input terminal for trigger pulses which have pulse intervals corresponding to delay states and which pulse intervals change continuously. Numeral 10 indicates a pulse distribution circuit which distributes each of the trigger pulses into two parts. Shown at 11 and 12 are trapezoidal wave generators. Numeral 13 represents an inverter, while numeral 14 is a delay line. Numerals 15-17 denote D.C. amplifiers, and 18-20 output terminals.

In FIG. 9, the trigger pulses which, as illustrated in FIG. 10a, vary continuously in correspondence with delay states are delivered from the input terminal 9 to the pulse distribution circuit 10. The trigger pulse is distributed by the pulse distribution circuit 10 into two parts, one of which is supplied to one trapezoidal wave generator 11 and the other of which is supplied to the other trapezoidal wave generator 12. An output waveform as shown in FIG. 10b is generated from the trapezoidal wave generator 11, while an output waveform as shown in FIG. 10d is generated from the trapezoidal wave generator 12.

In these waveforms, the time width of a flat portion, a rise portion or a fall portion is so set as to become one-third of a clock period at the maximum set frequency of the clock pulses.

The output of the trapezoidal wave generator 11 is inverted by the inverter 13 and is thereafter amplified by the D.C. amplifier 15, so that the first-phase clock pulses as shown in FIG. 10c are provided from the output terminal 18. The output of the trapezoidal wave generator 12 is directly amplified by the D.C. amplifier 16, and is delivered to the output terminal 19 as the second-phase clock pulses which, as illustrated in FIG. 10d, are delayed in phase by one-third of the clock period at the maximum set frequency of the clock pulses. Simultaneously therewith, the output of the trapezoidal wave generator 12 has its phase delayed by one-third of the clock period by means of the delay line 14, and is thereafter amplified by means of the D.C. amplifier 17, to be delivered to the output terminal 20 as the third-phase clock pulses as shown in FIG. 10e. The clock pulses in FIGS. 10c, 10d and 10e as thus obtained are impressed on the transfer electrodes 5, 6 and 7 in FIG. 1.

In this manner, the clock pulses of the respective phases are impressed on the transfer electrodes once by the single trigger pulse, and the respective phases of the clock pulses are always constant without depending on the period of the trigger pulses. When the period of the trigger pulse becomes long, the driving pulse of the first phase is maintained at a positive potential and the driving pulses of the second and third phases are fixed at 0 or a negative potential, so that charges are retained in the capacitor under the transfer electrode with the first-phase drive pulse impressed thereon.

That is, each time the trigger pulse arrives, the input signal is transferred to the next stage at a fixed time being determined by the maximum set frequency of the clock pulses. It is accumulated and retained until the next trigger pulse arrives.

In this way, a variable delay operation or a time axis transforming operation can be conducted. Since the charge transfer time is constant in this case, the transmitting efficiency is invariable. A further feature is that the circuit arrangement becomes very simple.

FIG. 11 shows another embodiment of the circuit which generates the clock pulses for use in the system of the present invention. In the embodiment, rectangular waves are employed for the clock pulses instead of the foregoing trapezoidal waves.

In the figure, number 21 designates an input terminal of trigger pulses; 22 a buffer; 23, 27, and 30 output amplifiers; 24, 28 and 31 output terminals; 25 an inverter, and 26 and 29 delay lines.

When trigger pulses shown in FIG. 12a are delivered from the input terminal 21 to the buffer 22, each is distributed into three parts on the output side of the buffer. One of the three outputs is directly applied to the outpt amplifier 23, so that the second-phase clock pulses shown in FIG. 12c are supplied from the output terminal 24.

Another output is inverted by the inverter 25, and is thereafter delayed by one-sixth of a clock period at the maximum set frequency of the clock pulses by means of the delay line 26. Thus, through the output amplifier 27, the first-phase clock pulses shown in FIG. 12b are delivered from the output terminal 28. The remaining output is delayed by a time of one-third of the above clock period by means of the delay line 29, and is thereafter passed through the output amplifier 30. Thus, the third-pulse clock pulses shown in FIG. 12d are delivered from the output terminal 31.

By impressing the clock pulses of the respective phases thus obtained on the transfer electrodes in FIG. 1, signal transfer is carried out.

FIG. 13 shows still another embodiment of the circuit which generates clock pulses for use in the signal transfer system of the present invention. The embodiment is employed for four-phase drive.

In the figure, numeral 60 indicates an input terminal to which is supplied a modulating signal whose level varies continuously in correspondence with delay states. Shown at 61 is a frequency modulator which effects frequency modulation with the modulating signal from the input termnal 60. Element 62 is a flip-flop, and numerals 63 and 64 represent the affirmation output and negation output of the flip-flop 62. Numerals 65 and 66 denote delay circuits, 67-70 pulse amplifier circuits, and 71-74 output terminals.

First of all, the frequency modulator 61 generates pulses at a frequency which is twice as high as the clock frequency as frequency-modulated by the modulating signal from the input terminal 60. The output pulses of the frequency modulator 61 have the frequency divided by 2 by means of the flip-flop circuit 62. Each of the affirmation output 63 and the negation output 64 is brought into two branches. The branched affirmation and negation output constituting one set are respectively supplied to the pulse amplifier circuits 68 and 70. Thus, the second-phase clock pulse .phi..sub.2 and the fourth-phase clock pulse .phi..sub.4 as illustrated in FIG. 14 are respectively delivered from the output terminals 72 and 74. The negation and affirmation outputs constituting the other set are delayed by a time equivalent to the transfer time t.sub.s by means of the delay circuits 65 and 66, and are delivered to the pulse amplifier circuits 67 and 69, respectively. Thus, the first-phase clock pulse .phi..sub.1 and the third-phase clock pulse .phi..sub.3, as illustrated in FIG. 14, are delivered from the output terminals 71 and 73. These clock pulses are applied to the transfer electrodes 5, 6, 7 and 8 in FIG. 4.

With a circuit arrangement in which a frequency modulator can directly generate rectangular waves, a circuit configuration shown in FIG. 15 can also generate similar waveforms. In this case, the rectangular wave output of the frequency modulator 75 need not have a 50 percent duty cycle. The rectangular wave pulse generated from the frequency modulator 75 is divided into two pulses, one of which is inverted by an inverter 76. The inverter output 78 and the direct output 77 of the frequency modulator correspond to the negation output 64 and affirmation output 63 of the flip-flop circuit 62 is FIG. 13. With the same succeeding stages as in FIG. 13, similar waveforms are obtained. In the case of FIG. 15, the frequency modulator 75 may generate the same frequency as the clock frequency.

Where the outputs of the circuit in FIG. 15 are used as clock pulses, the clock pulses .phi..sub.2 and .phi..sub.4 must be subjected to a D.C. shift as has been explained in connection with FIG. 5.

FIG. 16 shows an example of a D.C. shift circuit therefor. It is a diode clamp circuit which is composed of a capacitor C, a diode D, a resistor R and a D.C. power source E. With this circuit, the shift can, of course, be any potential by changing the clamp potential of the power source E. Such D.C. shift circuits are provided as parts of the pulse amplifier circuits 68 and 70 in FIGS. 13 and 15 and in the interior or exterior of the circuits.

Needless to say, the waveforms of the clock pulses are not restricted to the foregoing examples, but various modifications are considered.

The time width and the interval of the clock pulses of the respective phases are not restricted to the foregoing examples, but they can take various values.

Further, not only the three-phase and four-phase clock pulses, but also clock pulses of more phases can be adopted.