JP3143054U - Charged particle beam device and printed circuit board - Google Patents

Abstract

An object of the present invention is to provide a charged particle beam apparatus having a higher resolution (focus) while minimizing costs and taking advantage of current autofocus.
A charged particle beam irradiation means for irradiating a sample surface with a charged particle beam, a sample stage on which the sample is placed, a moving means for moving the sample stage, and a deflection means for deflecting the charged particle beam. A control means for controlling the deflecting means, and a secondary signal detecting means for detecting a secondary signal generated from the sample by irradiating the charged particle beam to the sample. The amplifier circuit provided in the control means has an oscilloscope means having an envelope function at the output terminal, and measures the resolution when the noise width is minimum while measuring the noise width by operating the envelope function for a fixed period. It is characterized by.
[Selection] Figure 1

Description

  The present invention relates to a charged particle beam apparatus equipped with an amplifier circuit and a printed circuit board used therefor.

  In recent years, in the control circuit equipped with operational amplifier, D / A (digital-to-analog) converter, resistor, etc. in charged particle beam equipment such as scanning electron microscope (SEM), higher accuracy of output against input is increasingly required. Has been. For example, in a lens deflection control board of a scanning electron microscope, there is a significant demand for higher accuracy of an output signal with respect to an input signal. In order to realize this high-accuracy output, components with higher accuracy such as resistors and operational amplifiers are selected and a circuit board is designed. A combination resistor is described as a highly accurate resistor. Examples of standard AC resistors using commercially available resistors have been proposed (see, for example, Patent Document 1 and Non-Patent Document 1).

  When considering adopting new candidate electrical components, the check terminal output of the circuit board may be checked with an oscilloscope and noise may be checked using an envelope function. At this time, a technique for measuring a voltage by sandwiching a self-made probe between printed boards is also known.

JP 2006-302546 A Proceedings of the previous meeting of the IEEJ 2007 Atsushi Dozen and Yasuhiro Nakamura: “Evaluation of Temperature Coefficient and Time Variation of Standard AC Resistors”

  In the case of an operational amplifier, the normal waveform acquisition of an oscilloscope is considered, but it is difficult to clearly determine the minimum noise instant, and if abnormal noise occurs occasionally, there is a possibility of measurement with variations in resolution measurement. is there.

  In addition, various types of alignment and focusing are performed in chronological order to measure resolution in a general-purpose SEM. However, even if noise is mixed on the display, the resolution is improved with technical support by judging the size of circuit noise. Since (focus adjustment) cannot be performed, the resolution is degraded.

  In order to reduce the noise of the printed circuit board, a review of the printed pattern will be considered. However, there is an empirical aspect and there is a possibility of remaking without entering the target specification.

  An object of the present invention is to provide a charged particle beam apparatus having a higher resolution (focus) while minimizing costs and taking advantage of the current autofocus.

  Alternatively, to provide a charged particle beam device with improved detection sensitivity by adding a probe and a control board to the detection circuit and operating the detector when noise is low, with focus remaining unchanged when detection sensitivity is important With the goal.

The present invention includes charged particle beam irradiation means for irradiating a sample surface with a charged particle beam, a sample stage for placing the sample, a moving means for moving the sample stage, and a deflection means for deflecting the charged particle beam. A charging means comprising an amplifier circuit, a control means for controlling the deflection means, and a secondary signal detection means for detecting a secondary signal generated from the sample by irradiating the sample with the charged particle beam. In particle beam equipment,
The amplifier circuit provided in the control means has an oscilloscope means having an envelope function at an output terminal, and when the noise width is minimized while measuring the noise width by operating the envelope function for a certain period, or A charged particle beam apparatus is characterized in that a sign that the noise is minimized is learned in advance, a control signal is sent to a control circuit, and the highest resolution immediately after the resolution is measured separately from the manual resolution.

  Alternatively, the minimum noise moment can be detected in advance while monitoring the output of multiple lens terminals, and when multiple lenses are both at the minimum noise, a control signal is sent to the control circuit to measure the maximum resolution separately from the manual resolution. Features.

  Alternatively, an analog switch means for time-controlling the moment of minimum noise and at least one small GND area in the printed circuit board for adding a return layer for the return current of the GND layer by receiving the signal of the switch did.

  According to the present invention, the accuracy of focus (resolution) can be improved with a minimum configuration.

A charged particle beam apparatus according to an embodiment of the present invention includes a charged particle beam irradiation unit that irradiates a sample surface with a charged particle beam, a sample stage on which the sample is placed, a moving unit that moves the sample stage, A deflecting unit that deflects a charged particle beam, an amplifier circuit, a control unit that controls the deflecting unit, and a second signal that detects a secondary signal generated from the sample by irradiating the sample with the charged particle beam. A next signal detecting means,
The amplifier circuit provided in the control means has an oscilloscope means having an envelope function at an output terminal, and measures the resolution when the noise width is minimum while measuring the noise width by operating the envelope function for a certain period. Provided is a charged particle beam device characterized by the above.

  Similarly, the detector means provides a charged particle beam apparatus characterized in that the measurement sensitivity of the detector is increased by measuring when the noise generated by the amplifier is minimum.

  Embodiments of the present invention will be described below with reference to the drawings.

  FIG. 1 shows a schematic configuration of a charged particle beam apparatus 100 to which the present invention is applied. The charged particle beam apparatus 100 deflects the charged particle beam by charged particle beam irradiation means for irradiating a sample surface with a charged particle beam, a sample stage for placing the sample, a moving means for moving the sample stage, and the charged particle beam. A deflecting unit, a control unit that includes an amplifier circuit, and controls the deflecting unit; and a secondary signal detecting unit that detects a secondary signal generated from the sample by irradiating the sample with the charged particle beam. Prepare. Details will be described below.

  In this embodiment, electrons are used as charged particles, and a container for keeping the electron path in a vacuum is not shown in the figure. In FIG. 1, a voltage is applied between the cathode 3 and the first anode 4 by a high voltage control power source 19 controlled by the CPU 21, and a predetermined emission current is drawn from the cathode 3. Since an acceleration voltage is applied between the cathode 3 and the second anode 5 by a high voltage control power source 19 controlled by the CPU 21, the primary electron beam 6 emitted from the cathode 3 is accelerated and proceeds to the lens system in the subsequent stage. To do. The primary electron beam 6 is focused by the focusing lens 7 controlled by the focusing lens control power source 22, and unnecessary areas are removed by the aperture holes provided in the aperture plate 8. Here, the actuator 18 is provided to drive the diaphragm plate 8. Thereafter, the primary electron beam 6 is focused as a minute spot on the sample 10 placed on the sample table 15 by the objective lens 9 controlled by the objective lens control power source 16, and two-dimensionally on the sample by the deflection coil 11. Scanned. The scanning signal of the deflection coil 11 is controlled by the deflection coil 11 according to the observation magnification. A detector 12 for detecting secondary electrons 14 is arranged on the electron source side of the objective lens 9. The signal detected by the detector 12 is amplified by the amplifier 13, processed by the CPU 21 in synchronization with the scanning of the primary electron beam 6, and displayed on the image display device 20 as a sample image. Here, the deflection coil control power supply 17 sends a signal for operating the deflection coil 11. The oscilloscope 1 and the arithmetic unit 2 are provided for reading out the current of the objective lens 9, for example. However, there is also a commercially available oscilloscope 1 that can read the current without the arithmetic unit 2.

  Here, when the current value measured by the oscilloscope 1 is important information, the CPU 21 may be provided so that the information cannot be read from the outside (for example, via the LAN), and only the control feed signal may be sent to the CPU 21 ( Not shown).

  FIG. 2 is an example of a printed circuit board inside the objective lens power source. The operational amplifier 26 and the operational amplifier 26 are connected in a printed pattern, and an output terminal 28 (analog GND terminal 29 for the ground of the oscilloscope probe is provided in another place) for confirming the operation of the printed circuit board 24 at the time of manufacture. Has been placed. When the current is passed through the objective lens 9 while observing the resolution, the oscilloscope 1 is connected between the two terminals described above, and measurement is always performed with the envelope function (one cycle time setting is required). Here, the signal output from the circuit of the multi-stage D / A converter 25 including the D / A converter 25 and the operational amplifier 26 is stabilized by the operational amplifier 26 of the printed circuit board 24 in the next stage, amplified by the transistor 27, and supplied to the objective lens 9. Apply current. Here, the probe is regarded as a capacitor (equivalent circuit) and needs to have a circuit configuration in which current does not flow from the oscilloscope 1 to the ground. Alternatively, a clamp tester may be used for current detection so that no current flows through the oscilloscope. Alternatively, the envelope X function may be realized by previously creating software by connecting the GPIB interface and the CPU 21 to a digital multimeter having a large input impedance in voltage measurement.

  Here, if necessary, a frequency converter may be provided in front of the transformer or in front of the switching power source as a more accurate AC power source in order to accurately measure the minimum noise. (Not shown) Alternatively, if necessary, an automatic voltage regulator (AVR) may be provided in front of the transformer or in front of the dropper type power supply (not shown) as noise removal.

  Here, when one inverting amplifier circuit is configured only with the combinational resistor proposed in Patent Document 2 and the output noise is measured, the inventors have newly found that the minimum noise can be achieved in the following procedure. . (1) Turn off the dropper type power supply that is the power supply. (2) When the dropper power supply is turned on, the envelope X (X is a number) of the oscilloscope is started and noise is measured.

  When this procedure was performed, the measurement from the first to the third time (continuous for about 120 seconds) immediately after power-on of the dropper system became the minimum noise. Accordingly, an operational amplifier (a circuit for exchanging signals with the CPU 21) for temporarily turning off the power supply of the operational amplifier amplifier circuit and the control board and the lens board for several seconds may be added. In this case, a high-precision resistor having a resistance value change rate R2 ′ that is convex upward with respect to the ambient temperature and a resistance value change rate R2 ″ that is convex downward with respect to the ambient temperature is usually a feedback resistor (starting from the operational amplifier output). Are set in the order of R2 ″ and R2 ′ and in the order of input resistance R2 ″ (op-amp input pin) and R2 ′ (in this case, R2 ′ = 5 kΩ and R2 ″ = 5 kΩ are selected) ). However, even if the feedback resistors (starting from the operational amplifier output) are set in the order of R2 ′ and R2 ″ and the input resistors R2 ″ (operational amplifier input pin) and R2 ′ in this order, there is reproducibility (here, R2 ′). We chose a resistance value of = 5 kOhm, R2 ″ = 5 kOhm).

  FIG. 3 is an example of an experimental result showing the transition of the noise width. When the voltage at the two terminals (the output terminal 28 and the GND terminal 29) is measured with the oscilloscope 1, a waveform of (a) is usually obtained. (Constant voltage control is performed so as to obtain the waveform of (a).) On the other hand, the inventors of this configuration are performing a resolution measurement experiment with a constant period setting of the envelope X of the oscilloscope 1, and the noise width of DC constant voltage control. Transits to (b) normal, (c) minimum noise, and (d) normal (return to (b) at the moment of returning from (c) to (b)) at irregular intervals of several milliseconds to several seconds or less I found out. (Here, Y1 is the voltage of the minimum value of noise.) The main cause (according to experiments by the inventors) is a dropper type power supply (provided near the deflection coil substrate of the charged particle beam apparatus). It was found that the device characteristics of the chip resistor and the metal thin film resistor appeared in the output through the large-scale integrated circuit 38 of the reference power supply and through the inside of the operational amplifier 26. At this time, it was found through experiments that the noise width can be accurately grasped if Y is displayed on the screen of the oscilloscope 1. At this time, if a line graph of Y with respect to time is displayed on the image display device 20, for example, by replacing one large-scale integrated circuit 38, it becomes easy to determine whether the noise of the new circuit is good or bad (not shown).

  According to the experiment of the inventor, when the dropper power supply, the reference large-scale integrated circuit 38, and the operational amplifier 26 of the inverting amplifier circuit are connected in series as an elemental experiment (investigation of the influence of each resistance of the inverting amplifier circuit on the output) In addition, (b) to (d) were reproducible. Furthermore, experiments were conducted with all types of operational amplifier 26, chip resistance, metal thin film resistance, and combination resistance changed, and (b) to (d) were reproducible. Here, it is known that the envelope function of the oscilloscope 1 mainly includes a peak hold circuit. Therefore, the detection of constant noise and minimum noise can be configured with a simple circuit, and at the time of minimum noise or when there is a sign of minimum noise (at this time, it is considered that the highest resolution image is displayed synchronously on the display of the oscilloscope 1. ) To send a feedback signal to the CPU 21 to obtain the highest resolution image once (or record a resolution of several seconds continuous on video or the like). Here, the detection of the sign of the minimum noise will be described in detail with reference to another drawing. Furthermore, although it can be considered that detection of the sign of the minimum noise is possible if the noise threshold Y1 (minimum noise voltage) can be detected, the flowchart is appropriately omitted in this embodiment. However, realization is easy.

  Here, according to the experiment of the inventor, it is normally displayed by the coupled DC, but the minimum noise can be detected even by the coupled AC. Further, immediately after determining Y1 with the combined DC, the CPU 21 may switch to the combined AC (when the maximum and minimum period of noise is faster than the combined DC) and determine a new Y1. Here, since the operational amplifier 26 is installed in the amplifier 13 of the detector 12, it can be considered that the phenomena (b) to (d) exist. If the CPU 21 performs processing at the moment of the noise threshold Y1, the detection sensitivity can be improved.

  FIG. 4 shows another embodiment in which it takes time to capture one cycle as a preliminary experiment for capturing the waveform of the envelope in the state of FIG. Here, a generally known offset function is used to expand the voltage display to several hundred mV / div, and the trigger position is set as shown in the figure. When estimating the time of one cycle in the envelope X by the waveform acquisition single sequence, it may not be possible to read at the measurement limit level of the oscilloscope 1 depending on the voltage waveform of the output terminal. Thus, the inventors have found that the detection efficiency of the oscilloscope 1 is increased by applying one or more noises with periodicity of noise created by the user. (If the added noise after detection is unnecessary, it may be removed by a noise filter or the like.) As a periodic noise, the inventors found the high voltage pulsed power supply 23 through experiments. Here, if necessary, the periodic noise is preliminarily tested and the threshold value is determined and removed by calculation inside the CPU 21, or two horizontal cursors are displayed on the gate measurement screen of the oscilloscope 1 to determine the position of the trigger position. It may be outside the measurement range.

  A control process in which the CPU 21 performs the above-described process, that is, the process of obtaining the highest resolution image once, in detail will be described with reference to a flowchart of FIG. Initial setting is performed first. That is, in step S1, the envelope X and the noise threshold Y1 (Y1 shown in FIG. 3) are registered in the CPU 21. At the time of resolution measurement, the user adjusts the alignment at S2, and the user adjusts the focus at S3 and presses the photographing key. Thereafter, in S4, it is determined whether or not the noise threshold Y1 of the objective lens is reached (FIG. 3), and in S5, the highest final resolution is photographed when the noise is minimum. Here, as shown in FIG. 6, the minimum noise value (filled square) and the normal time (two-dot chain line) may be clearly recognized together with the resolution image.

  Furthermore, another embodiment in which the oscilloscope 1 is removed and the highest resolution is taken will be described. Here, the inventor has newly found out as means for finding the sign of the minimum noise at a time earlier than the determination of the noise threshold value Y1 described above. The detection of the sign of the minimum noise is not mounted on the normal oscilloscope 1. Therefore, the transition of the waveform is stored in the hard disk in advance with the oscilloscope 1 connected to the charged particle beam apparatus 100, and the tendency of the minimum noise of the output terminal (for example, alignment and then just focus) is read while reading the waveform information from the hard disk. The output is set to the output terminal of the operational amplifier 26 from the D / A converter 25 (for the objective lens) for focusing when the user is set to the specified voltage (DC fixed voltage) after a few seconds, and then the minimum after a few seconds The CPU 21 stores in advance as a parameter the learning whether the noise is output. After that, the highest resolution immediately after the parameter is taken separately from the moment when the user presses the switch for taking the resolution.

  Here, when a large voltage is measured on the oscilloscope screen depending on the circuit due to disturbance, two vertical cursors may be displayed on the gate measurement screen to make the position of the trigger position out of the measurement range.

  Here, when Y1 is used, since the current measurement is not performed, it may be a noise voltage that tends to increase the noise after the instantaneous voltage when the immediately preceding noise transitions from the maximum to the minimum.

  FIG. 7 shows an example in which the data stored in the hard disk described above is read again and displayed on the image display device of the charged particle beam device 100 (the display on the oscilloscope 1 is kept as it is). According to the experiment of the inventor, the situation in which the two voltage widths of the two-dot chain line normally transition to the noise threshold Y1 with the constant voltage setting in the envelope X state is A1, A2, A3, A4, A5 (A1 is A5 is displayed in order of the last time at the start time of the minimum noise, and B1, B2, B3, B4, B5 (the saved one is read again, so after A1 to A5 are displayed sequentially, B1 is the minimum noise (B5 is displayed in order of the final time at the start time). Therefore, the angle θ1 may be defined in advance, and if it exceeds this angle, the CPU 21 may determine that the noise is a sign of minimum noise. Of course, θ1, θ2, and θ2 may be determined as parameters. Furthermore, at least two of θ1, θ2, θ3, and θ4 may be selected as parameters. If the accuracy of θ1, θ2, θ3, and θ4 is not improved, the inclination may be calculated using the least square method, and the angle formed by the two-dot chain line may be used as a parameter. At this time, if the calculation becomes complicated, a CPU 1 for calculation in the CPU 21 may be newly installed. Alternatively, high-speed computation may be performed on another computer via a network (not shown).

  Alternatively, depending on the circuit configuration, if it is difficult to determine the angle from the waveform using a filter, the area surrounding A1 to A5 and B1 to B5 is calculated to determine the normal and minimum noise size. The threshold value may be determined and determined.

  At this time, in this figure, A1, A2, A3, A4, A5 and B1, B2, B3, B4, B5 are set to display the image in the Y direction sequentially on the image display device 20 after displaying in the X direction. . For this reason, by determining the minimum noise with θ1 or θ2 consisting of A1, A2, A3, A4, and A5 or with θ1 and θ2, the signs of the minimum noise can be captured at the shortest, so the processing efficiency is good. In this method, there is no influence on the charged particle beam device 100 from the probe of the oscilloscope 1, and the resolution can be improved only by changing the software of the CPU 21. Alternatively, in a situation where the target resolution of the current charged particle beam apparatus 100 can be obtained with certainty, the oscilloscope 1 and the arithmetic unit 2 may be connected to capture the sign of the minimum noise and obtain the resolution in the same procedure. Alternatively, the transition of the noise width displayed on the image display device 20 (or the oscilloscope 1) (for example, as shown in FIG. 3) is picked up by a hybrid camera, image-processed, a feedback signal is sent to the CPU 21, and the same resolution is used in the same procedure An image may be obtained once.

  Here, the description has been given for the case where the minimum noise is a constant voltage, but the normal time and the time when the noise is minimum can also be distinguished in a transition waveform such as a scan waveform (meaning not a constant voltage) to the deflection coil 11. If the SEM image is displayed on the image display device 20 in a time with little noise, the image quality of the SEM image can be substantially improved.

  Although one output terminal is described here, two or more terminals (for example, output terminals of a stage control motor (eg, DC motor)) may be synchronized with the minimum noise to obtain the maximum function.

  FIG. 8 is a conceptual diagram in which the minimum point of noise optimized from a plurality of detection signals is detected to obtain the highest resolution. In the above description, one output terminal is described, but two or more output terminals may be sequentially synchronized with minimum noise. (A) is the display screen of the oscilloscope 1 which measured the output of the output terminal 28 of the objective lens 9 with the envelope X of the envelope function. Since the voltage of the objective lens is constant voltage control, it is two two-dot chain line bolts. Noise (black spots) due to disturbance is confirmed in a preliminary experiment, a threshold is determined, and data is deleted in the oscilloscope 1. Here, the resolution is measured with a delay of time seconds obtained by adding Δt0, Δt, and Δt1 from the time when the user presses the switch for taking the resolution. This will be explained in more detail below. The sine waves of (b), (c), and (d) return from the noise (normal) including the voltage set in (a) to the minimum noise (FIG. 3B), and from the minimum noise to the noise (normal). In order to make it easier to explain the transition from the results of the experiment of the creator, an approximate display was made. After the detection of the minimum noise point of the objective lens (b), the minimum noise point is detected after a preset Δt seconds of the GND voltage of the control board (c), and after the preset Δt 1 second of the GND voltage at another location on the control board. The minimum noise point is detected (d). Alternatively, instead of (c), the parameter of the minimum noise point of the alignment voltage may be entered to obtain the resolution. At this time, Δt and Δt1 may be determined by a preliminary experiment. Alternatively, the noise threshold Y1 may be detected in a preliminary experiment within Δt and within Δt1.

  FIG. 9 is an example of an initial setting screen when detecting minimum noise. (A) is newly displayed at the time of automatic setting when automatic selection of the menu of the initial setting is selected (Δt and Δt1 are parameters determined by preliminary experiments). A column 32 represents the current value, and a new parameter is set by the scroll button 31a. Thereafter, the registered value is confirmed with the menu bar 33. When the manual of the initial setting menu for the parameter at the time of detecting the minimum noise (within Δt, noise threshold Y1 detection within Δt1) is selected (b) is displayed. (The display configuration of (b) is not shown because it is the same as (a)) The column 32 displays the current value, and a new parameter is set with the scroll button 31a. Thereafter, the registered value is confirmed with the menu bar 33. At this time, displaying the parameters on the image display (resolution image) improves the operability of the apparatus, and has the advantage that the next resolution can be taken quickly.

  As another example, it would be advantageous if the minimum noise moment could be controlled by the user in advance. For this reason, as shown in FIG. 10, a small second GND layer 35 (surrounded by an insulating layer 36) is newly set between the high-speed wiring 37 of the multilayer substrate and the first GND layer 34. . Ideally, it is only necessary to minimize the noise at the time of designing the printed circuit board such as a resistor, but it is necessary to minimize the noise of an important circuit for resolution and to improve the characteristics of the element typified by the resistance temperature coefficient of the resistor. There is an advantage that the variation is absorbed at the time of substrate assembly adjustment, and the performance is improved by optimizing selection and non-selection of at least one second GND layer 35.

  Here, it is known that when a current flows through the high-speed wiring 37, a reverse current is generated in the first GND layer 34 immediately below. Accordingly, the newly added second GND layer 35 is connected to the newly-added second GND layer 35 using the analog switch large-scale integrated circuit 38 (one is connected to the first GND layer 34, and one is off of the analog switch large-scale integrated circuit 38). In this state, the wiring is connected to the second GND layer 35. In order to minimize the noise, the switch is turned on and the signal line is connected to the ground). (For example, in the wiring layout of the Vcc layer and the GND layer of the memory module (72-pin SIMM), the larger the wiring width, the smaller the current density. Therefore, if the GND layer increases temporarily, the return current flow of the GND layer will increase. It is natural to think that the indirect influence on the signal layer can be reduced and the minimum value of the noise can be shifted.) Therefore, when the GND area is expanded, the EMC (electromagnetic compatibility) can be suppressed. The resolution can be obtained by shifting the moment of the minimum noise and synchronizing the moment of the minimum noise of the objective lens current and the alignment current, for example. Here, in order to reduce noise, the minute second GND layer 35 between the high-speed wiring 37 and the first GND layer 34 may be changed to a low transmission loss material or a low transmission loss multilayer material. Furthermore, all the insulating layers of the multilayer substrate may be changed to a low transmission loss material or a low transmission loss multilayer material.

  This procedure (synchronizing the moment of minimum noise to obtain resolution) will be described in more detail with reference to FIG. In S11, the user registers the envelope X and the noise threshold Y1 in the CPU 21. Next, in S12, the CPU 21 checks whether it is within the designated time of the day. Here, during the measurement, a display during the maximum resolution measurement may be displayed on the image display device 20 (not shown).

  Here, the level of noise is not a problem in a normal scanning electron microscope, but it is set to obtain the best resolution that exceeds that of a normal scanning electron microscope. A supplementary explanation will be given for the elemental experiment that provides the basis for this. (According to the experiments by the inventors, the AC 100-volt outlet (the best one was selected from a plurality of distribution panel candidates was chosen by experiment). Here, for the output of the inverting amplifier circuit, two op amps 26 that are optimally used for deflecting the electron beam and manufactured combined resistors (0. 005%, 10 kOhm, 0.6 w, temperature coefficient of 0.02 ppm / ° C. or less (−55 to 125 degrees) was used. Here, referring to the configuration of the standard AC resistor described above (background technology), the commercially available resistor is a similar metal thin film resistor (the resistor with nickel chrome and ceramic bonded with an adhesive is estimated to be the same) Selected. Immediately before, the same circuit was used to measure the noise (using the envelope X function of the oscilloscope 1) using the chip resistor, and the minimum noise of the operational amplifier output using the combined resistor was 40.0 millivolts or less (at a later date, digital voltimeter 80 Microvolt equivalent) and smaller than chip resistance. Alternatively, in another element experiment, the minimum noise of the operational amplifier output using the combination resistance for the reference power source constituted only by the combination resistance for the deflection circuit was 36.0 millivolts or less during the experiment, which was smaller than the chip resistance. Ideally, in a D / A converter circuit, it is estimated that the maximum resolution will be achieved if a combination of an operational amplifier and a deflection operational amplifier as a reference power supply is composed of only resistors, but it is estimated that it will be about 40 seconds (converted from 36.0 millivolts or less). The inventor estimated that.

  Next, according to experiments by the inventors, a quadro-type to single-type operational amplifier (Precision Operational Amplifier) and two combination resistors (10 kilohms, 0.6 W, 0.02 ppm / ° C. or less (−55 ° C. to 125 ° C.) In place of measurement of a digital voltimeter for 5 minutes (naturally, it is natural to think that mixing of harmonic noise from 100V voltage is reduced in this measurement), the actual measurement of the large-scale integrated circuit 38 for the reference voltage is performed. The value (40 microvolt fluctuation) and the measured value of the output of the operational amplifier 26 (40 microvolt fluctuation) coincided. Therefore, since the fluctuation width is small compared with (110 microvolt fluctuation) with two chip resistors of 10 kilohms, for example, it is natural to think that it is easy to find the moment of minimum noise using the function of the envelope X of the oscilloscope 1. is there. In addition, a quadro-type operational amplifier (deflection amplifier, precision operational amplifier) and two combination resistors (10 kilohms, 0.6 W, 0.02 ppm / ° C or less (-55 ° C to 125 ° C)) are used for 5 minutes. In the measurement of the digital voltimeter (it is natural to think that harmonic noise from 100V voltage is reduced in this measurement), the actual measurement value (40 microvolt fluctuation) of the large-scale integrated circuit 38 for the reference voltage and the operational amplifier The measured value of the output of 26 (40 microvolt fluctuation) matched. Therefore, since the fluctuation width is small compared with (110 microvolt fluctuation) with two chip resistors of 10 kilohms, for example, it is natural to think that it is easy to find the moment of minimum noise using the function of the envelope X of the oscilloscope 1. is there.

  In addition, a quadrature operational amplifier (objective lens) and a combination of two resistors (10 kilohms, 0.6 W, 0.02 ppm / ° C. or less (−55 ° C. to 125 ° C.)) can be measured with a digital voltometer for 12 minutes. (It is natural to think that harmonic noise from 100V voltage is reduced in this measurement). Actual measurement value of large-scale integrated circuit 38 for reference voltage (30 microvolt fluctuation) and actual output of operational amplifier 26. The values (30 microvolt variation) were in agreement. Therefore, since the fluctuation width is small compared with (110 microvolt fluctuation) with two chip resistors of 10 kilohms, for example, it is natural to think that it is easy to find the moment of minimum noise using the function of the envelope X of the oscilloscope 1. is there.

  Also, a single-type operational amplifier (Precision Operational Amplifier) and two resistors (relative value 0.01 ppm / ° C or less (-55 ° C to 125 ° C), R2 'characteristic, 5 kilohms, 0.005%, 0.3 W) The two resistors do not cut the lead and are connected with a jumper wire on one commercially available breadboard and measured with a digital voltimeter for 5 minutes in a continuous manner (in this measurement, mixing of harmonic noise from 100V voltage is considered to be less Is natural), the measured value of the output of the operational amplifier 26 was small (30 microvolt fluctuation). Accordingly, since the fluctuation range is small compared to two chip resistors of 10 kilohms, it is natural to think that it is easy to find the moment of the minimum noise by using the envelope X function of the oscilloscope 1, for example.

  Although the combination resistance is selected here, it is natural to think that it is better to divide the time setting whether it is within the specified time even for the chip resistance (for example, the printed board shown in FIG. 2 in which only the chip resistance is mounted). Here, when selecting the normal use, a command is sent from the keyboard to the CPU 21, and the best resolution measurement function exceeding the normal one is utilized (the user can return to the function immediately if necessary). Good. Subsequently, in S13, the user aligns the alignment. In S14, the user adjusts the focus and presses the shooting key. In S15, a noise threshold Y1 of the objective lens is found. In S16, a control signal for turning on the minute second GND layer 35 immediately below the alignment signal of the printed circuit board is sent. In S17, the CPU 21 automatically measures the resolution.

  At this time, a plurality of minute second GND layers 35 may be prepared in advance, and a large-scale integrated circuit 38 of the analog switch may be turned on and off for experimental evaluation while manufacturing the product, so that it can be designed at low cost. is there. Here, the current position of the first GND layer 34, the second GND layer 35, and the insulating layer 36 is estimated, and the installation position and the size that bring about an appropriate effect as the design specifications of the new second GND layer 35 are determined with respect to the printed circuit board. You may make some prototypes and make elemental experiments to estimate.

  According to the experiment of the inventor, when a variation in the GND layer (for example, measurement using the above-described screw probe) greatly affects the desired output terminal as an error as a preliminary experiment, it cannot be measured by the CPU 21 during that period. It may be set.

  Furthermore, a method for obtaining the highest resolution by defining the minimum noise created from the result of the element experiment conducted by the creator will be described. Here, a circuit configuration is made up of a dropper power supply, a large-scale integrated circuit 38 for generating a reference voltage, and an operational amplifier 26 for inverting amplification. In this configuration, first, the power of the dropper type power supply is turned on. This output becomes minimum noise at t5 when measured with the envelope X of the oscilloscope 1. Here, one large-scale integrated circuit 38 is included. (Delay time t6) The output (delay time t7) of the operational amplifier 26 becomes the minimum noise after the time obtained by adding t5, t6 and t7. Next, in the configuration in which the large scale integrated circuit 38 is added between the large scale integrated circuit 38 (delay time t8) and the switching power supply, the output of the op becomes the minimum noise after the time t5, t6, t7 and t8 are added. Thus, when a plurality of large-scale integrated circuits 38 are connected, the transmission delay may be estimated in advance and the minimum noise may be defined.

  At this time, the first stage large-scale integrated circuit (or in the case of a plurality of continuous D / A converter circuits may be regarded as the first stage D / A converter circuit except the reference power supply large-scale integrated circuit 38). The operational amplifier 26 is greatly affected by noise. Pay attention to the minimum noise of the large-scale integrated circuit 38 in the first stage (or the minimum noise of the dropper type power supply) in advance, and measure the minimum noise (measured with the envelope X, taking into account the delay time of each element) The output of the final stage operational amplifier 26 in which a plurality of large-scale integrated circuits are connected is regarded as the minimum noise (even if it is determined by the oscilloscope 1 because it is a measurement limit and cannot be determined), and the maximum resolution is measured. May be. Here, in the scan waveform for driving the deflection coil 11, the filter cut in the high frequency band can be estimated empirically. Therefore, when measuring the envelope X, the band limit of the oscilloscope 1 may be used to estimate the minimum noise value.

  FIG. 12 is an example in which the transition state between the objective lens and the minimum noise of alignment is displayed on the image display device in an easily understandable manner. In the charged particle beam apparatus 100, the normal resolution is manually determined after the user has aligned and focused. At this time, there may be some time between the final focusing and the pressing of the shooting key. Therefore, the objective lens and the minimum noise of alignment are displayed on the image display device 20, and the cursor 39 for focus guide at the time of manual determination is created. First, the cursor 39 repeatedly changes in the horizontal direction from left to right in one cycle (fixed interval) between the maximum noise and the minimum noise of the objective lens. At the same time, the cursor repeatedly changes in the horizontal direction from left to right in one cycle (constant interval) between the maximum noise and the minimum noise. At this time, the user visually observes the cursor of the objective lens and presses an imaging key for resolution at the minimum noise point. At this time, the deviation time L from the minimum noise point of alignment may be displayed with a cursor. At this time, when the period shown in FIGS. 3B to 3D of the objective lens and the alignment is greatly different, the same period as shown in the figure is displayed so that the user can easily recognize the minimum noise point. Also good.

  Here, the objective lens and the alignment may be controlled to a minimum noise for about several seconds by a user's requested operation (input from the keyboard). Here, it can be considered that there is periodicity from the newly controlled constant voltage of the large scale integrated circuit 38 to the minimum noise. Further, it is known that the current always flows in the opposing GND layer while the current flows through the signal line. As a means for making the minimum noise continuous, an ON control signal is sent to the at least one or more second GND layers 35, that is, the objective lens or the second GND layer 35 immediately below the alignment at least once in a timely manner. This control procedure may be determined in advance by a user through preliminary experiments. Using this method, the voltage of the objective lens is first minimized. Next, the noise of the alignment voltage is reduced. This information may be displayed for several seconds in FIG. 10 with the objective lens and the alignment cursor 39 vertically aligned with the minimum noise point at the minimum noise point, and may wait for the user's request for resolution measurement. Further, the voltage ExB may be controlled to a minimum noise for about several seconds.

  FIG. 13 shows an initial setting screen when the user selects the second GND layer 35 as means for making the minimum noise continuous. When the user selects GND1 (selects the second GND layer 35) immediately below the signal line of the alignment control circuit with the image display device 20, the menu bar 33 highlights from white to diagonal lines as shown in the figure. At this time, when the user presses the shooting key according to the above-described procedure (the objective lens and alignment are controlled to the minimum noise by the user's requested operation for about a few seconds), the shooting key is pressed only twice (here, only the alignment voltage). After that, GND is operated at t1, and GND1 is operated at t2 at a constant interval t3 (however, t1, t2, and t3 may be set to be variable)) The second GND layer 35 is connected in a timely manner to align with the objective lens. Take the highest resolution with the minimum noise time. Further, one or more GND menu bars 33 for the voltage ExB may be displayed.

  Alternatively, focusing on the vicinity of the edge of the resolution image and taking the final resolution, the cursor 39 is moved and displayed near the edge (in the image display) for easy understanding, and the shooting key is pressed when the minimum noise is present. Also good.

  Here, when the load fluctuation is large outside the charged particle beam device 100, it is estimated that the one-cycle length of the cursor 39 is largely switched according to the frequency fluctuation on the image display device 20, and the user is informed of the situation of the noise. The CPU 21 may determine the large and small time zones and notify them on the image display device 20. Alternatively, if the noise Y (Y shown in FIG. 3) also changes greatly (in the time zone where the noise is large or small), the CPU 21 determines the situation as the time when the noise is large or small, and the image display device 20 You may let me know above. At this time, if the operation is performed at a time when the noise is small, the probability of taking the highest resolution increases.

  Further, the information regarding the time when the noise is large or small is accumulated in the HDD every day and stored in the database, and displayed on the image display device 20, so that it becomes effective information for time determination at the time of continuous operation of resolution measurement.

  As another method for obtaining a stable and reproducible resolution, an output terminal 28 is provided between the large scale integrated circuit 38 of the reference power source of the objective lens control board and the D / A converter 25 shown in FIG. An example of measurement with the envelope X (see FIG. 4 for the trigger position) will be described. According to experiments by the inventors, in a coupling circuit composed of a switching power supply and a dropper power supply, a voltage that is 1.9 times or more rarely displayed in the normal voltage shown in FIG. Therefore, since it is considered that the resolution deteriorates at this voltage value, the user may determine the measurement time before and after in advance, and the CPU 21 may judge and control processing so as not to measure the resolution.

  Of course, the present invention can be easily realized in a general-purpose commercially available multilayer substrate. Specifically, at least one or more second GND layers 35 are arranged inside the electrical insulating layer between the high-speed wiring 37 and the first GND layer 34 for controlling the noise minimum time of the wiring layer described above. Jumper line means is created so that connection can be selected from the wiring 37 to the second GND layer 35, from the second GND layer 35 to the first GND layer 34, and to the high-speed wiring 37 and the first GND layer 34.

  Here, in order to reduce noise, the minute second GND layer 35 between the high-speed wiring 37 and the first GND layer 34 may be changed to a low transmission loss material or a low transmission loss multilayer material. Furthermore, all the insulating layers of the multilayer substrate may be changed to a low transmission loss material or a low transmission loss multilayer material.

  According to the present embodiment, it is possible to provide a charged particle beam apparatus equipped with a high resolution technology only by adding an oscilloscope means and a simple circuit to the scanning electron microscope. Alternatively, it is possible to provide a charged particle beam apparatus equipped with a high resolution technology simply by obtaining a plurality of data in advance using an oscilloscope means and a simple circuit.

Sectional drawing which shows the main structures of a charged particle beam apparatus. The block diagram which shows an example of the printed circuit board inside an objective lens. The graph which shows an example of the experimental result which shows the transition of the noise width of an Example. The graph which shows another Example when taking in of an envelope function takes time. The flowchart which shows the procedure in the case of taking the highest resolution at the time of resolution measurement. The screen figure which shows an example which displayed the content which shows the transition of the noise width of an Example on a screen. The graph which shows the concept which saves the noise width of an Example once, and reads and finds the minimum noise. The graph which shows the concept which detects the minimum point of the noise optimized from several detection signals, and takes the highest resolution. The screen figure which shows an example of the initial setting screen at the time of minimum noise detection. Sectional drawing which shows one cross section of a board | substrate. The flowchart which shows the procedure in the case of detecting the minimum noise moment from a plurality of detection signals, taking the highest resolution by shifting the minimum noise moment of the other detection signals with reference to one detection signal. The screen figure which shows an example which displayed on the screen the transition state of the objective lens and the minimum noise of alignment. The screen figure which shows an example which displayed on the screen the initial setting of the objective lens and the minimum noise of alignment.

Explanation of symbols

1 Oscilloscope 2 Arithmetic Unit 3 Cathode 4 First Anode 5 Second Anode 6 Primary Electron Beam 7 Focusing Lens 8 Aperture Plate 9 Objective Lens 10 Sample 11 Deflection Coil 12 Detector 13 Amplifier 14 Secondary Electron 15 Sample Stand 16 Objective Lens Control Power Supply 17 Deflection coil control power supply 18 Actuator 19 High voltage control power supply 20 Image display device 21 CPU
22 Focusing lens control power supply 23 High voltage pulse power supply 24 Printed circuit board 25 D / A converter 26 Operational amplifier 27 Transistor 28 Output terminal 29 GND terminal 30 Fixing screw 31 Scroll button 32 Column 33 Menu bar 34 First GND layer 35 Second GND layer 36 Insulating layer 37 High-speed wiring 38 Large scale integrated circuit 39 Cursor 100 Charged particle beam device

Claims (16)

A charged particle beam irradiating means for irradiating the surface of the sample with a charged particle beam; a sample stage for placing the sample; a moving means for moving the sample stage; a deflecting means for deflecting the charged particle beam; and an amplification circuit. A charged particle beam apparatus comprising: control means for controlling the deflection means; and secondary signal detection means for detecting a secondary signal generated from the sample by irradiating the sample with the charged particle beam. ,
An oscilloscope means having an envelope function with a constant period at at least one output terminal of the amplifier circuit provided in the control means, and the noise is minimized or the noise is minimized. A charged particle beam apparatus characterized by comprising means for measuring a new resolution when the noise of the minimum time is minimum from the resolution, apart from the resolution of manual focus.   2. The charged particle beam apparatus according to claim 1, further comprising clamp tester means for detecting a current or a voltage at the output terminal or the GND layer of the printed circuit board.   2. The charged particle beam apparatus according to claim 1, further comprising a frequency converter means or an automatic voltage regulator means before the power supply means for operating the amplifier circuit or the transformer means before the power supply means.   2. The noise generating means for displaying on the oscilloscope a voltage for use as a reference at a preset time from outside the charged particle beam device, if necessary while operating the envelope function, and the noise A charged particle gland apparatus having a control means for controlling the generating means, and having means for removing the periodic noise and judging transition of the noise width.   2. The resolution according to claim 1, wherein when the noise is minimized, the envelope function is operated at a constant period to measure the noise width in advance, and the resolution is separated from the manual focus resolution when the noise width is minimized. A charged particle beam apparatus comprising means for measuring a new resolution when the noise of the minimum time is minimum.   2. The charged particle beam apparatus according to claim 1, comprising means for displaying a state from a normal output of at least one terminal of the noise to a minimum noise.   In Claim 1, the measurement time is determined in advance in a preliminary experiment immediately after the time when the noise is minimum, and then the measurement is started again when the noise at the other output terminal is minimum. With the user's pre-selection, start measurement again when the noise of another terminal is minimum. After measuring in advance when the noise of the preselected amount is minimum, it has means for measuring a new resolution. A charged particle beam apparatus that performs the above processing.   2. The method according to claim 1, wherein a measurement time is determined immediately after the time when the noise is minimum, and a time when the noise at the other output terminal is minimum is measured. Measures when the noise of another terminal is the minimum by user's pre-selection. After measuring in advance when the noise of the preselected amount is minimum, it has means for measuring a new resolution. A charged particle beam apparatus that performs the above processing.   2. The micro-volume GND layer according to claim 1, wherein a micro-volume GND layer is provided between a GND layer opposed to the high-speed wiring of the output of the amplifier circuit if necessary, and the micro-volume GND layer is electrically connected to the GND layer. The analog switch means is disposed near the high-speed wiring, and the terminal of the analog switch has a small volume GND layer and the other terminal has a sufficient return current to the GND layer. A charged particle beam apparatus comprising a printed circuit board connected to flow.   10. The charged particle beam apparatus according to claim 9, wherein a new insulating material is set between the minute volume GND layer and the GND layer for low noise.   2. If necessary, the minimum noise time of the first stage circuit is measured in advance using an envelope function having a constant period in the amplifier circuit if necessary, and the output of the large-scale integrated circuit in the last stage of the amplifier circuit is the minimum noise time period. A charged particle beam apparatus characterized by having a means for measuring the resolution by defining the minimum.   12. The minimum noise time according to claim 11, further comprising means for calculating the signal delay time between the first-stage circuit and the last-stage circuit by previously measuring and adding together with a constant period envelope function if necessary. Charged particle beam equipment.   7. The charged particle beam apparatus according to claim 6, further comprising means for displaying a cursor that indicates a moment of minimum minimum noise of the terminal and serves as a pointer for taking a resolution.   2. The means for displaying an operation start time and an operation time of a micro GND unless the micro GND layer of at least one terminal of the noise is selected, and a means for determining the operation start time and the operation time according to claim 1. A charged particle beam apparatus comprising:   A printed circuit board having a wiring layer, an electrical insulating layer, and a GND layer, wherein a microvolume GND layer is disposed between the wiring layer and the GND layer, and the microvolume GND layer is electrically connected to the GND layer. The analog switch means is disposed near the wiring layer, and the terminal of the analog switch is connected to the minute volume GND layer, and the other terminal is connected to the GND layer so that the return current sufficiently flows through the volume of the minute GND layer. A printed circuit board characterized by that.   16. The printed circuit board according to claim 15, wherein a new insulating material is set between the minute volume GND layer and the GND layer to reduce noise.

Images