JPS6141170B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a charge transfer device in which the DC level of a signal detected at a signal output terminal is kept constant.

Recently, there has been active development of so-called charge transfer type transversal films that use charge transfer devices (CTDs) such as charge coupled devices (CCDs) and bucket brigade devices (BBDs). It is being done. Among these, a programmable filter whose weighting coefficient can be varied by an external electrical signal is extremely useful as an automatic equalizer. Patent application filed in 1973 by the same person as the present applicant as a transversal filter suitable for this purpose.
-100413 (charge transfer type transversal filter). This transversal filter controls the gain of an input signal by an external electrical signal, and injects or drains a charge proportional to the signal into a charge transfer device to generate a positive,
Negative arbitrary weighting is performed.

However, this type of transversal filter has the disadvantage that the amount of charge transferred in the transfer channel changes depending on the state of the weighting coefficient of each input stage, and the DC level of the signal detected at the output end changes. had.

This invention was made in view of the above circumstances.
The purpose is to provide a charge transfer device that can obtain an output signal with good linearity without changing the DC level of the signal detected at the output end even if the weighting coefficient of each input stage changes. be.

Hereinafter, one embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing the basic structure of a charge transfer device, such as a transversal filter. In the figure, 1 is a charge transfer channel that delays, adds, and subtracts signals, and a DC charge injection means 2 for injecting DC charges into this channel 1 is provided at the front stage of this charge transfer channel 1. There is. Further, each stage of the charge transfer channel 1 is provided with weighting circuits 3a to 3d. The weighting circuits 3a to 3d are for weighting the input signal inputted through the terminal 4 with a positive weighting coefficient according to the control voltage inputted through the terminals 5a to 5d. Furthermore, charge transfer channel 1
Weighting circuits 6a to 6d are provided at each stage. The weighting circuits 6a to 6d are for weighting the input signal inputted through the terminal 4 with a negative weighting coefficient according to the control voltage inputted through the terminals 7a to 7d. That is, the input signal is weighted by each of the weighting circuits 3a to 3d and 6a to 6d, and is injected into or flowed out from the charge transfer channel 1 as a signal charge. After the charges in the charge transfer channel 1 are sequentially transferred through each stage, they enter the output detection means 8 and are detected as an output signal from the output terminal 9. The output detection means 8 is also provided with an output terminal 10 for detecting the DC level of the output signal. A signal from this output terminal 10 is inputted to the DC charge injection means 2 by a feedback means 11. That is, the DC charge injection means 2 injects DC charges into the charge transfer channel 1 in accordance with the DC level of the output signal detected by the output detection means 8.

FIG. 2 is a schematic diagram for explaining the operation of the transversal filter shown in FIG. 1 above.
That is, positively weighted signal charges qo(N) + qs(N) are injected into each stage of charge transfer channel 1, or negatively weighted signal charges are injected into each stage of charge transfer channel 1.
qo'(N)+qs'(N) is transferred to the output detection means 8 while flowing out. Here, qo
(N), qo′(N) are DC components, qs(N), qs′(N)
represents the AC component. At this time, the charge Qo + Qs detected as an output signal by the output detection means 8 is becomes. Therefore, from the output terminal 10 of this output detection means 8, a charge proportional to the above charge Qo + Qs is generated.
Qo′+Qs′ is output. This charge Qo' + Qs' is compared with the reference charge Q _c in the comparator 12, converted to a DC signal in the low-pass filter (LPF) 13, and the DC charge Q _I1 is input from the front stage of the charge transfer channel 1. be done. That is, this DC charge Q _I is controlled by the feedback means 11 so that the DC component Q _p of the charge Qo+Q _s outputted from the output terminal 9 has a value proportional to the reference charge Q _c .

Therefore, in the transversal filter having the above configuration, the DC component of the signal charge at the output stage is constant regardless of the magnitude of the weighting coefficient of each stage, and therefore the DC level of the output signal is also constant.

FIG. 3 specifically shows the output detection means 8, feedback means 11, and DC charge injection means 2 in the transversal filter. In addition, in the figure, the weighting circuits 3a to 3d and 6a
~6d etc. are omitted.

The output detection means 8 has channels 14 and 15 formed by branching the output side of the charge transfer channel 1 into two;
Floating diffusion regions 16 and 17 provided in the channels 14 and 15, respectively, gate electrodes 18 provided adjacent to the floating diffusion regions 16 and 17, and gate electrodes 18 provided adjacent to the gate electrodes 18 on the channel 14 side. Provided gate electrode 1
9, 20 and drain region 21, and channel 15
It consists of a drain region 22 provided adjacent to the gate electrode 18 on the side, and a floating diffusion amplifier 25 consisting of transistors 23 and 24 connected to the floating diffusion region 17 and the drain regions 21 and 22. . That is, the signal charges transferred through transfer channel 1 are branched to channels 14 and 15. The charges branched into one channel 14 enter the floating diffusion region 17 and the floating diffusion amplifier 2
5 is detected as an output signal. The operation of this floating spreading amplifier 25 is a method well known to those skilled in the art, so the details will be omitted.
The charges branched to the other channel 14 enter the floating diffusion region 16 and
discharges the capacitor 26 connected to the . Also,
From this floating diffusion region 16, gate electrodes 18 and 19 to which clock pulse _φ1D is applied are connected.
A constant charge is drained to the drain region 21 every clock by the gate electrode 20 to which the clock pulse φ _2D is applied, and the capacitor 26 is charged. Therefore, the potential of the floating diffusion region 16 becomes a constant potential when the charges flowing in from the channel 14 and the charges flowing out to the drain region 21 become equal.

Further, the feedback means 11 includes the capacitor 26,
Inverter 27, resistor 28 and capacitor 29
and the low frequency amplifier 13 consisting of.

If the amount of charge flowing into the floating diffusion region 16 is greater than the amount of charge flowing out, the capacitor 26 is relatively discharged, so that the input voltage of the inverter 27 decreases. Therefore, the output voltage of inverter 27 increases. This potential change is inputted to the input gate electrode 30 of the DC charge injection means 2 after the AC signal component is removed by the low frequency filter 13 .

The DC charge injection means 2 is based on the well-known fill-and-spill method and is composed of a source region 31, the input gate 30, a storage gate electrode 32, and a sampling gate electrode 33. That is, from the source region 31, the pulse φ _s1
Charges are injected into the source region 31 by
The others are accumulated under the storage gate electrode 32. The amount of charge stored under this storage gate electrode 32 is inversely proportional to the voltage of the input gate electrode 30. In this case, since the voltage on the input gate electrode 30 increases, the charge under the storage gate electrode 32 decreases. This charge is transferred under the transfer gate electrode 34 by the sampling pulse φ _SH applied to the sampling gate electrode 33, and after being transferred through each stage, enters the floating diffusion region 16 through the channel 14. At this time, the amount of charge flowing into the floating diffusion region 16 is reduced compared to the initial state.

Conversely, if the charges initially flowing into the floating diffusion region 16 are smaller than the charges flowing out,
The feedback means 11 operates with the opposite polarity to that in the above case,
The amount of charge flowing in increases compared to the initial state. That is, the feedback means 11 operates so that the charge flowing into the floating diffusion region 16 and the charge flowing out from the floating diffusion region 16 are equal.

Next, the mechanism of charge injection and outflow will be explained in more detail using FIGS. 4 and 5. Fourth
The figure shows a cross-sectional view along the line AB of the charge transfer device having the structure shown in FIG. 3, and a potential diagram formed under each electrode, and FIG. 5 shows a pulse waveform diagram used here.

First, the input section will be explained with reference to the left side of FIG. 4. At time t=t ₁ , the source region 31
Since is at a low level, charge is injected from the source region 31, passes under the input gate electrode 30, and is stored under the storage gate electrode 32. At time t=t ₂ , the source region 31 becomes at a high level, so the excess charge under the storage gate electrode 32 returns to the source region 31 .

Next, at time t= _t3 , the sampling gate electrode 33 becomes high level, so the storage gate electrode 32
A portion of the charge below is transferred below the transfer gate electrode 34. The amount of charge at this time is equal to the product of the difference between the potential under the input gate electrode 30 and the potential under the sampling gate electrode 33 and the capacitance under the storage gate electrode 32. Here, the potential of the sampling gate electrode 33 and the capacitance under the storage gate electrode 32 are constant. Therefore, the amount of charge injected and transferred is
It is proportional to the input voltage V ₁ applied to the input gate electrode 30.

Next, the output section will be explained with reference to the right side of FIG. 4. Since the transfer gate electrode 34 is at a low level at time t=t ₁ , the charge in the transfer channel 1 passes under the output gate electrode 35 and enters the floating diffusion region 16 . At time t= _t2 , since the gate electrodes 18 and 19 are at a high level, floating diffusion regions 16 are formed below the gate electrodes 18 and 19.
More charge flows in. Next, at time t= _t3 , since the gate electrodes 18 and 19 are at a low level, a charge equal to the product of the potential difference under these electrodes 18 and 19 and the capacitance under the gate electrode 19 is generated by the gate electrodes 18 and 19. is scooped up. Time t=t ₁ again
Returning to , since the gate electrode 20 is at a high level at this time, the gate electrode 20 opens and the scooped up charges are discharged to the drain region 21 . That is, this series of operations causes a certain amount of charge to flow out from the floating diffusion region 16 every clock.

As described in detail above, according to the present invention, the DC level at the output end of the charge transfer device remains constant regardless of the operating state, and an output signal with good linearity can be obtained. Further, it becomes easy to design a floating diffusion amplifier, a subsequent source follower amplifier, and the like.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing the basic configuration of a transversal filter according to an embodiment of the present invention, FIG. 2 is a schematic diagram for explaining the operation of the filter, and FIG. 3 is the same as that shown in FIG. 1. FIGS. 4 and 5, which are plan views specifically showing block diagrams, are for explaining the operation of each of the above embodiments in more detail. FIG. 4 is a plan view taken along line A-B in FIG. FIG. 5 is a diagram showing the state of potentials formed under each electrode, and FIG. 5 is a waveform diagram of a clock pulse. DESCRIPTION OF SYMBOLS 1... Charge transfer channel, 2... DC charge injection means, 3a-3d, 6a-6d... Weighting circuit, 8... Output detection means, 11... Feedback means, 1
2... Comparator, 13... Low frequency wave generator, 16, 17
...Floating diffusion region, 18, 19, 20
... Gate electrode, 21, 22 ... Drain region,
23, 24...Transistor, 25...Floating diffusion amplifier, 26...Capacitor, 27...
...Inverter, 30...Input gate electrode, 31...
...Source region, 32...Storage gate electrode, 33...
...Sampling gate electrode, 34...Transfer gate electrode, 35...Output gate electrode.

Claims (1)

[Scope of Claims] 1. A charge transfer means for delaying, adding and subtracting signals, a DC charge injection means for injecting DC charges into the first stage of the charge transfer means, and a predetermined charge transfer means from each stage of the charge transfer means. a plurality of input means for injecting signal charges; a means for detecting the DC level of the delayed, added, and subtracted signal charges at the output stage of the charge transfer means; and the DC level detected by the means is always constant. A charge transfer device comprising: means for feeding back the DC level information to the DC charge injection means to keep the DC level of the output stage constant. 2. The means for detecting the DC level branches a channel at the output stage of the charge transfer means, inputs the signal charge from one channel into a storage means consisting of a floating diffusion layer, and outputs the signal charge from the storage means every clock. 2. A charge transfer device according to claim 1, wherein a constant signal charge is caused to flow out. 3. The charge transfer device according to claim 1, wherein the charge transfer means is an input weighted charge transfer type transversal film.