DE3205552A1 - METHOD AND DEVICE FOR DIGITIZING ANALOG CURVES - Google Patents

Description

Int. Ref .: Case 1503 February 12, 1982

Hewlett-Packard Company

METHOD AND DEVICE FOR DIGITIZING ANALOGUE CURVES

The digitization or conversion of an analog voltage into a digital word is done by means of Arialog / digital converters. The transformation can be done in many different ways. The converter types described below are known, among others.

Subsequent converter

In the case of an analog / digital converter of the following or servo type, a binary up / down counter for generating a binary output signal uses »which in turn a digital / analog converter in a feedback loop around a subtracting · comparator controls »The corresponding scheme is shown in Figure 1A. The system counts the incoming clock pulses starting with zero, and the count value is converted into a by the digital / analog converter converted to analog voltage Vq. This analog voltage is with compared to the actual analog input voltage Vn. If the Analog voltages Vq and Vn are the same, the comparator switches over and stop counting. The digital output signal of the binary counter corresponds to the analog input voltage. This type of converter has the disadvantage of being slow if it is a digital one . The output signal should be obtained over a wide range, since (2-1) Clock cycles must be counted if N is the number of bits in the digital word »

Step converter · ■.

The step converter is a simplified version of the follow-up converter. Instead of an up / down counter, a simple binary counter is used used. The counter is reset to zero with each conversion step and starts counting clock pulses. at this method produces an output signal of the digital / analogue

Hewlett-Packard Company

Int. Ref .: Case 1503 ~ 5 -

Converter, which has a stair-shaped course. Similar to the ■

To! expenditure! the counting process in the step converter is stopped- · y _{ ' th if its output signal is equal to the analog input span-' ...

nung'is. Its last count therefore corresponds to the digital one. · Word for the analog input signal. This type of converter has the. ··; .. same disadvantageous slowness as the night! on converter. The . · ..

Slowness is intensified by the fact that after every step of the walk the counter is not reset to zero.

Successive approximation converter

- ■■■ -. ■ ■ 'f

With this type of converter, the analog input signal is converted to [..

successive trying out a digital "1" in all on- · L consecutive bits of a feedback digital / analog converter. 'I. approximated, starting with the bit with the highest significance. '·

will. A block diagram of such a system is shown in FIG. 1 B ί

shown. A "1" is tried in the first converter step .. .J If the actual analog voltage is greater than the converted j

Analog voltage and the output signal of the comparator does not change, the "1" is retained. However, if the actual analog voltage is smaller, the "1" is replaced by a "0" · '

replaced, and a "1" is tried in the next bit .. On this. ; The approximation process continues until all bits ■ '

are calculated. At the start of each conversion cycle, the Holding register · set to zero and the entire conversion sequence i

will be repeated. This type of converter is required for its implementation complicated peeling and also has the disadvantage

of slowness. '

• Converter with ramp comparator

FIG. 1 C shows an example of an indirect analog / digital converter, the ramp comparator or pulse width modulation converter. At the beginning of a conversion cycle, the cycle switch S ₁ opens, and a capacitor C. is set at a constant

3205S52

Hewlett-Packard Company "■
Int. Ref .: Case 1503. - 6 -

Current I ₁ charged, whereby a ramp voltage V _ß = (I ₁ ZC ₁ ) t is generated, where t is the charging time. When switch S ₁ is opened, the counter is activated and begins with. the counting of incoming pulses. The count continues until the ramp voltage Vn is equal to the analog input voltage V ". The comparator then stops the counting process and the last count in the binary counter corresponds to the digital word for the analog voltage.

With this type of converter (2 - 1) pulses must be counted in order to obtain an output signal for the full range with N bits will. Although this type of converter is simpler in structure than the other types mentioned, it has a very slow conversion speed.

Simultaneous converter

Simultaneous or parallel analog / digital wall! He is by far the fastest. With this type of converter, each quantization level in the digital word is from mull to full range uses its own analog comparator with a fixed reference signal. The outputs of these comparators then match each other connected so that a parallel digital output signal is created. Figure 1 D shows an example of such a system According to the state of the art. For a resolution of N bits (2 - 1) separate comparators and separate reference levels are required. Such a system therefore requires great complexity as the price for the high conversion speed.

The limited speed and / or great complexity the known analog / digital converters lead to an upper operating limit the systems in which they are used. This is particularly evident in known digital storage oscilloscopes, for example at high speed where the upper frequency limit is about 1 GHz. A digital storage oscilloscope

Hewlett-Packard Company

Int. Ref .: Case 1503 - 7 -

holds analog / digital wall! er for converting or digitizing the analog input curve into a digital signal that is stored e.g. in a semiconductor memory. This stored signal is then used for display on a conventional cathode ray tube converted back to analog form. This method is in Figure 5 shown. If, for example, the analog / digital converter cannot work with a sampling frequency above 25 MHz, this is it system containing it is limited to 10 MHz. All signals above 10 MHz can then no longer be processed.

The invention according to the independent claims has the object based on creating an analog / digital converter that can work above 10 MHz and is particularly suitable for use in oscilloscopes that process higher frequencies can. .

An analog-to-digital converter according to the preferred embodiment of the present invention has the disadvantage of being limited Bandwidth of the known converters is no longer available. A high-speed digital storage oscilloscope containing such an analog-to-digital converter therefore no longer has the frequency limitation of the known oscilloscopes, but can instead use curves of analog signals in the Record, digitize, process and display the GHz range, even if they are cyclically of short duration or only one appear only once. In addition, these curves can be stored indefinitely in digital form without changing over time worsen. The saved curves are to be used later Time available for processing, e.g. for calculation, the Fourier transformation, spectral power density, cross power spectrum, Cartesian and polar coordinates, correlation function ,! transfer function, coherent output power etc. After processing, the information can be conveniently displayed for visual analysis, e.g., on an ordinary Cathode ray tube, a television screen, etc.

. Ό β O

Hewlett-Packard Company

Int. Ref .: Case 1503 ~ 8 -

According to the preferred embodiment of the invention, the uses Analog / digital converter a semiconductor image converter, e.g. in Form of a charge coupled device with a recessed channel, creating an almost instantaneous quantization of an analog signal for digitization is possible. The use of a solid state imager in the present embodiment of the analog / digital converter specifically enables digitization of analog signals in the GHz range for later processing or display, even if these signals are only briefly cyclical Are permanent or occur only once. This means a significant improvement over the known analog / Digital converters and thus the well-known storage oscilloscopes.

A solid state imager typically operates as follows. When light representing an image hits an array of semiconductor elements of the imager falls, it is detected by the semiconductor elements as each element exposed to the light is charged will. The array of semiconductor elements forms an X-Y matrix. The mosaic formed in the matrix by the charged elements is therefore a representation of the incident image. By successive Scanning of the semiconductor elements and detection the presence or absence of charge in these elements gives an electrical signal representing the image.

Since the charge in the semiconductor element is also proportional to the intensity or the amount of incident light, the digitization of the charge information and the processing these data provide information about the center of the track which forms the picture. This way of determining the curve is more accurate than simply determining the number of charged elements in the X-Y matrix. In other words, the exact one analog waveform can be extrapolated from a series of points as a fine line, so that the (lack of) focus of the impinging image on the solid-state image converter is unimportant

Hewlett-Packard Company

Int. Ref .: Case 1503 - 9 -

will.

The intensity information can also be used to improve the quality of the Serve image signal. In a solid-state image converter, certain spots can be activated even when there is no incident light is. This can be due to possible defects in the semiconductor element. This means that if there is no image, the image converter nevertheless does not emit a signal that corresponds to complete darkness. This is referred to as the dark drawing of the device. the Dark drawing of the image can be obtained from the digitized intensity information be derived. By storing the dark drawing of the image converter and subtracting it from an image signal then to an improvement in the quality of the image signal, since the fixed Pattern noise due to the dark drawing is removed.

If the imager is clocked continuously without the Cycle is interrupted, the dark values for all image elements or spots in a line (or in each half of a line, when the split structure described below is used according to the preferred embodiment). So it is only required to store a dark value for each line (or half line).

A charge coupled device imager is a suitable imager in the preferred embodiment of the invention. It is with ■ recessed channel in NMOS technology with two levels of clock lines made of polysilicon and with lighting from the front executed. This imager technology represents a good one A compromise for a solid-state imager in which light sensitivity, low signal capacity and simplicity at the same time should be present.

For generating and projecting the image onto the shark conductor image converter a small cathode ray tube is used. Of the

Hewlett-Packard Company

Int. Ref .: Case 1503 - 10 -

Electron beam within the cathode ray tube draws an image on a fluorescent F ³ ""': e of the cathode ray tube. The fluorescent surface besteh, * ^"A · he fiberglass plate with a layer of phosphors ^ \, ·>_a" one face of the plate.

. ■ * ·> .% ,
Therefore, if the analog sigfi "··. '.. · ** ■>', - ^ htstoff-surface is drawn as a picture, wiro, ^s >, ^e outside of the fiberglass board transferred. On, / * r '^'" ^: the transmitted image remains relatively sharp and clear and not blurred. When the solid-state image converter is brought into contact with the fiber optic plate in such a way that the transmitted image is incident on the image converter field, the image converter can receive the projected image. In addition, a further fiber optic plate can be arranged directly on the surface of the image converter, to be precise as part of the image converter arrangement. The fiber optic plate of the imager can then be brought into direct contact with the fiber optic plate of the cathode ray tube so that the transmitted image can be easily and effectively transferred to the. "'Image converter. The fiber optic plates lead to a good coupling efficiency of the image transmitted from the fluorescent plate of the cathode ray tube. Since the solid-state image converter is light-sensitive, the projected image is captured by the individual light-sensitive elements contained in the image converter field In conventional circuits, the individual elements of the field can be read or interrogated one after the other, so that a digital signal is obtained which corresponds to the supplied analog signal and can be stored and further processed.

Further advantageous embodiments or developments of Invention are characterized in the subclaims.

The invention is explained below using exemplary embodiments in conjunction with the associated drawing. In the Show drawing

Hewlett-Packard Company

Int. Ref .: Case 1503 -IV-

FIG. 1 A, B, C and D examples of analog / digital converters according to FIG the state of the art;

Figure 2 shows the optical interface according to the present invention between a solid state imager and a cathode ray tube;

Figure 3 shows a semiconductor imager chip with associated registers and amplifiers on the wafer;

FIG. 4 shows a cross section through an NMOS charge-coupled picture transducer device with recessed channel;

FIG. 5 shows a conventional digital storage oscilloscope according to FIG State of the art; and

Figures 6 and 7 preferred embodiments of a high-speed digital Storage oscilloscope in which an analog / digital converter is used in accordance with the present invention.

As shown in Figure 2, a cathode ray tube 2 with a thin electron beam 4 is used to create a curve the surface of a wafer with a semiconductor imager array 6 to write. The plate 6 is outside the wall the cathode ray tube 2 arranged, and a fiber optic Plate 10 for the cathode ray tube 2 and a fiber optic "window" 8 for the semiconductor wafer 6 are used to the output light of a phosphor layer 11 of the cathode ray tube 2 to be transferred to the image converter plate 6.

As shown in FIGS. 3 and 4, the image converter 6 is designed as an NMOS charge-coupled device with a recessed channel and has two levels of polysilicon clock lines 12 and lighting from the front 14. These two levels

Hewlett-Packard Company

Int. Ref .: Case 1053 - 12 -

of polysilicon lines I and II are in the present example denoted by 12A and 12B for differentiation. This technology is a good compromise between photosensitivity and processability small signals and simplicity. With these characteristics, the charge coupled device imager can use as a semiconductor image converter 6 to build a high-speed analog / digital converter. This fast converter can in turn can be used to build a fast storage oscilloscope with high writing speed. A two-phase Clock system 11 or 13 is used to clock out the charge signal, as it is compatible with the double polysilicon process and because the corresponding clock signals can easily be generated. Recessed channel charge coupled devices have one higher transmission efficiency and can operate at higher frequencies than surface channel devices. One self aligned boron implant under the polysilicon layer II 12B forms a barrier layer area. This increases the treatment option a single charge hole formed under the polysilicon layer I 12A.

The image converter shown does not need any light shielded. Transport register to avoid image blurring. The light falls on the array of imager elements only for a short time Duration. During this time the clocks are stopped or the exposure is so brief that the clock does not change seems to be moving. The exposure can be controlled by varying the current of the electron beam. As a result, there are no against Light-shielded transport registers are required, and the structure of the light-sensitive area can be based on an array of parallel Registers 14 are reduced from charge coupled devices.

Figure 3 shows an array of charge coupled devices that is divided in half along a vertical center line. The polysilicon lines are connected so that the charge

Hewlett-Packard Company

Int. Ref .: Case 1053 - 13 -

is clocked horizontally, starting from the center of the P-plate and progressing to the sides. The charge will column for Columns are transferred to high speed serial registers 16 located on either side of the array of semiconductor imager elements run along. The signal charge representing an image is therefore subject to a transfer from parallel to serial. For each slow parallel clock cycle, there are many fast ones serial clock cycles.

In the embodiment shown in FIG. 3, two detection amplifiers 18 are located on the image converter plate 6. One is located at the end of each serial register 16. This split structure reduces the time required for clocking out the accumulation of charges from all semiconductor imager elements. The dark charge of each charge accumulation is reduced by half, and so is the standard deviation of the charges becomes strong, reduced.

When an analog waveform is written on the image converter surface the time axis is horizontal and parallel to the channels of the charge coupled device. Hence represents each column of spots represents a time sample. The analog accumulations of charge are clocked out serially, each time occurs simultaneously for one column. While processing of the charge pattern of a column can position the beam on the vertical axis, which is proportional to the signal voltage, can be determined very precisely.

By the construction of the center line shown in FIGS. 3 and 4 the imager is split in half, with the. Periodicity of the field is retained. The middle polysilicon line II 15 is with the two adjacent polysilicon lines I 17 and 19 connected. The remaining polysilicon lines I and II 12A and 12B are arranged in pairs so that the load is moved away from the center line.

Hewlett-Packard Company

Int. Ref .: Case 1503 - 14 -

The horizontal channels of the charge coupled device intersect the vertical serial registers at right angles. the Transferring a column of charges from the parallel field to the serial register is done by expanding each other Polysilicon-II gate of the serial register until it's the last polysilicon line I of the parallel field overlaps. This structure directs the accumulation of charge around the corner and prevents that charge in the serial register can reenter the parallel field.

At the output of the serial shift register, the charge flows into a resettable free-floating detection distribution. the last cell is from the detection distribution through a polysilicon-II gate separated at a fixed potential. This reduces the capacitive coupling between the serial clocks and the Amplifier input.

In Figure 6 is a high speed digital storage oscilloscope shown, as an example of an oscilloscope that the invention Analog / digital converter used. Analog input signals become input terminals Y1, Y2, Y3 or combinations thereof fed. The signals control the electron beam in a cathode ray tube 2, which takes an image on a semiconductor imager 6 is generated, which forms part of the analog / digital converter according to the invention. The charge signal from the picture elements of the semiconductor image converter 6 runs in two detection amplifiers 18, of which only one is shown in FIG. 6 for the sake of simplicity is. The amplitude of the output signal from the detection amplifier 18 is digitized by an analog / digital converter 19. That The digital output signal is fed to a digital subtraction circuit 20 in which the dark drawing of the image converter 6 is subtracted. The result of the subtraction is fed to a circuit 21, which the center of the illuminated track on the Image converter surface is determined with a resolution that is finer than the size of a light spot. An optimal midpoint value

Hewlett-Packard Company

Int. Ref .: Case 1503 - 15 -

is then stored in a memory 29 for each image converter column.

A microprocessor 31 controls the data transmission and obeys commands from external sources, e.g. a keypad, control buttons or a rotary pulse generator 33. It outputs control signals for sensitivity. Zero offset, scanning speed, Trigger control, etc. The microprocessor 31 can also handle data processing and data extraction from the memory 29 for display run on a cathode ray tube 37.

Because all signals are stored in memory before they are displayed the display portion of the oscilloscope can operate at its own refresh rate, which is just high enough must that flickering is avoided. The cathode ray tube 37 therefore has enough time to draw its own grid, including the. Designation of axes and other alphanumeric and other characters. A sawtooth generator can be used for display of curves can also be provided.

Another embodiment of a fast storage oscilloscope, that uses the analog / digital converter according to the invention is shown in FIG. In this embodiment, the entire The content of the image converter 6 is stored in a read / write memory 39. With this system it is possible to use the imager write several curves before transferring its contents to memory. This is done by the fact that the curves only then clocked out after several of them have been written to the imager. The full picture of all curves including Grids and labels can then e.g. on a cathode ray tube 37, which operates with a raster scan.

Blank page

Claims (11)

Hewlett-Packard Company Int. Ref .: Case 1503 February 12, 1982 PATENT CLAIMS | /. Device for digitizing analog curves, characterized by a cathode ray tube. (Z-) with a fiber optic Front panel (10) on the screen for supplying an impinging image to a semiconductor image converter (6) accordingly an applied analog input signal; a semiconductor image converter (6) on the fiber optic Faceplate of the cathode ray tube that receives the image. and holds and at least one of the impinging images 10 · speaking image signal generated; and . an evaluation circuit (18, 19) with the semiconductor image converter is connected and emits control signals to the image converter and that .processed in the image signal in such a way that at least one digital output signal is generated. 2. Device according to claim 1, characterized in that ·, ' .that the image converter signal contains image intensity information and that the evaluation circuit (18, 19) has a sensor (19) for the acquisition of the intensity information and the generation of an intensity output signal. 3. Device according to claim 2, characterized in that that the sensor (19) is an analog / digital converter and that the intensity output signal is a digital electrical one Signal is. 4. Device according to one of claims 1 to 3, .g e k e η η records by a glass fiber optic plate (8) which is arranged on the semiconductor image converter (6) in this way, that the impinging image from the cathode ray tube (2) is directed to the imager. Hewlett-Packard Company 3 20 555 2
Int. Ref .: Case 1503 - 2 - 5. Device according to Artspruch 1 or 4, characterized by a memory device (39) which is connected to the evaluation circuit (18, 19) and at least one digital output signal and, if necessary, the intensity ! 3 signal stores; a processor device (31) connected to the memory device and the stored digital output signal and if necessary converts the intensity signal in a suitable manner and generates at least one display signal; and a display device (37) which communicates with the processor device is connected and receives the display signal and reproduces it visually. 6. Apparatus according to claim 5, characterized in that that the display device (37) is a television receiver. . 7. Apparatus according to claim 5, characterized in that that the display device (37) is a cathode ray tube system is. . · 8. Method for digitizing analog curves, characterized in that that an analog curve on a fiber optic plate is shown as a tracer, 'on which plate a semiconductor image converter with a multitude of photosensitive elements is arranged in such a way, that the light trail is the light ■ 25 sensitive elements activated and that information the multitude of photosensitive elements in selected Sequence are read out, the activated light-sensitive elements are determined and a digital output signal according to which the analog curve is generated. Hewlett-Packard Company Int. Ref .: Case 1503 - 3 - 9. The method according to claim 8, characterized in that ■ that the digital output signal corresponding to the analog curve is displayed. 10. The method according to claim 9, characterized in that that the digital output signal is processed before the display. 11. The method according to claim 8, characterized in that that the light intensity at each activated photosensitive Element is detected and an output signal corresponding to the intensity is generated.