RU2250531C2 - Method for calibrating time sweep of chronographic image converter-recorder - Google Patents

Abstract

FIELD: investigation of optical processes in pico- and femto-second ranges.

SUBSTANCE: proposed method involves establishment of ambiguous correlation between calibration marks on screen produced by electron beam coming from photocathode area illuminated by semiconductor pulse-modulated laser on screen region confined by shadow mask and respective time intervals used for time calibration in processing image converter time chart. This correlation is effected either using mathematical formula (in case of simple waveforms of sweep voltage) or by pre-numbering (during adjustment of image converter-recorder) at discrete points of sweep voltage. All these procedures make it possible to change laser mode of operation, that is to reduce requirements to its radiation thereby simplifying implementation of proposed method and reducing calibration error.

EFFECT: enhanced precision of time sweep calibration.

1 cl, 2 dwg

Description

The invention relates to the field of electronic technology and can be used to study the optical processes of the pico and femtosecond range. The image intensifier contains an electron-optical converter - image intensifier.

When picosecond pulsed signals are studied using chronographic image intensifier recorders, the signal under investigation is deployed along the time axis using sawtooth-shaped scanning pulses, the duration of which can be up to Tr ~ 100 ps / screen.

In the literature cited in the list of terms are used: time scan, time axis, spatial axis, also used by the applicant. To calibrate the time base, two main methods are currently used. The first method [1] consists in applying harmonic time stamps to a special calibration deflecting system, the period of which should be less than the scan time by at least 5-10 times. In this case, the image of the harmonic signal (similar to the oscillographic signal) is expanded on the screen, the period marks of which are used for calibration. The invention of the applicant [1] increases by 3 times the accuracy of time measurements in the first method, however, it has all of the following disadvantages inherent in this method. First of all, it is the need to install a special calibration deflecting system in the image intensifier tube, the presence of which negatively affects the main characteristics of the image intensifier device, such as temporal and spatial resolution on the screen, distortion, and others. In addition, there are fundamental problems with calibrating the fastest scans. For example, to calibrate a sweep of duration on the screen Tp ~ 200 ps / screen, at least 5 time stamp periods with a frequency of Fm = 25 GHz are required.

There are currently no deflection systems for deflecting a signal of this frequency. In addition, a recording speed of Vz = 2 · π · Fm · A is required, where A is the amplitude of the timestamps (MB), V _З > 8 · 10 ¹⁰ m / s, which also exceeds the current level achieved, V _З ~ 5 · 10 ¹⁰ m / s. It is no coincidence that the fastest image intensifiers, such as P920 PME (Pilips, France), N5716 (Hamamatsu, Japan), PV 001 (VNIIOFI, Russia), lack a calibration system and carry out optical calibration of the time axis.

The known method [2] of optical calibration of the time base scan is that a portion of the photocathode of the image intensifier tube is illuminated by a laser optical radiation modulated at the desired frequency. At the same time, on a portion of the screen corresponding to the image of the illuminated portion of the photocathode, when scanning, a sequence of time stamps in the form of flashes is formed, the period of which is equal to the period of repetition of short pulses of laser radiation, as shown in Fig. 1c (upper diagram). The shape of the envelope of the laser radiation of the prototype Rlaz npoot (t) is conventionally shown to be the same as the shape of a pulse-modulated, continuous within the modulating pulse, laser radiation Rlaz (t) in the proposed method (here t is time). The disadvantage of the method [2] of optical calibration of the time axis is the need to use, together with the image intensifier, an expensive and sophisticated laser equipment that generates picosecond optical pulses with different (depending on the duration of the sweep) Toy period. In addition, there are restrictions on the registration of short brightness marks of duration t and corresponding to time stamps. In [2], the relation for the indicated durations is given: t and << τ << Toy << Tr (*), where τ is the time resolution of the image intensifier tube. So for Tp = 200 ps / screen (temporal resolution τ ~ 1 ps) at Toi ~ 20 ps, the duration of one flash is ti≤0.5 ps, the width of the element along the time axis at Vp ~ 1.5 · 10 ¹⁰ cm / s L≤0.075 mm and the luminous flux from such a thin element may not be sufficient for a reliable registration of the MB element. From the eopogram in the description [2] (Fig. 2, p. 208), it follows that when trying to improve the visibility of the calibration strokes, the condition (*) formulated in the description of the method is violated, and therefore, the width of the strokes becomes comparable with the period which significantly, up to 50%, increases the calibration error. Thus, the use of known calibration methods imposes various limitations in the calibration of fast sweeps, and therefore, the accuracy of the calibration of the time axis is reduced.

The time sweep calibration presented in [3] is carried out by using a time interval of 21 ps between two short laser pulses. Such calibration is suitable mainly for determining the total duration of the sweep and requires the installation and adjustment of a complex optical system.

Also known is a calibration system [4] for the time axis of an X-ray image intensifier tube for a special case when the studied train of X-ray pulses is synchronized with an optical train formed from a short laser pulse. As the calibration value, the period of the image of the x-ray pulses visible on the screen of the train is used. From the description it follows that the method [4] can be used only in the case when the radiation source and the image intensifier tube are a single installation for one scan value and the type of the signal under study.

The closest technical solution to this proposal is the well-known [2] method for calibrating the time base of a chronographic image intensifier recorder, which consists in illuminating a portion of the photocathode with pulsed laser radiation. In the method [2], the pulse repetition period is such that at least 10 luminance marks are placed on the calibrated scan. As noted above, the first drawback of the method [2] is its complexity associated with the complexity of the used laser equipment. The second disadvantage of the method [2] is the unsatisfactory calibration accuracy, which is associated with high requirements for the pulses of time stamps. In the method - the prototype due to the fact that the number and width of the marks do not correspond to the optimal values, the calibration error reaches several tens of percent, as mentioned above.

The first technical result of the proposed method is to increase the accuracy of the calibration of the time base of chronographic image intensifier recorders. The second technical result of the proposed method is its simplicity associated with ease of implementation.

The technical result of the proposed method for calibrating the time base of chronographic image intensifier tubes, which consists in illuminating a portion of the photocathode with pulsed laser radiation, is achieved by applying a shadow transmission or blocking to the inner surface of the screen, for example, at the beginning of the spatial axis, at the beginning of the spatial axis a beam of electrons a mask that creates during scanning along the time axis an image of illuminated or darkened time stamps, the number of which in in the optimal case, at least N = 10, simultaneously with the registration of the studied signal, the corresponding portion of the photocathode is illuminated, from which the photoelectrons enter the screen segment indicated above, by radiation from a semiconductor laser operating in a single pulse-modulated semiconductor laser at a power level from 1 to 10 mW, when the duration of the modulating pulse in the microsecond range is the same for all scan durations (since the duration of this pulse does not matter and must be greater than the longest sweep), the connection of time intervals with the corresponding calibration marks on the screen is carried out according to the formulas, depending on the method of interpolating the form of the sweep voltage, applicable in the entire range of sweep durations:

Δt = ΔП / (S · A) - for the linear dependence U (t),

Δti = ΔП / (S · A · n · (i / N) ^(nl) ) - for the power law U (t),

Δt _k = ΔP · Δτ / (S · (U ((i + 1) · Δτ) -U (i · Δτ))) - in the general case,

where Δt, Δti, Δt _k are the time intervals corresponding to the calibration marks for each form of the dependence of the scan voltage on time,

ΔP is the geometric distance between the calibration marks evenly distributed on the screen along the time axis,

S is the sensitivity of the deflection system,

A is the slew rate of the deflecting sweep voltage,

i, k - current indices of the time interval between calibration marks,

Δτ is the time interval between voltage samples when measuring its shape,

U (i · Δτ) is the value of the sweep voltage at time i · Δτ when measuring its shape.

As for the size of the illuminated or darkened time stamps, for the proposed method, their size can be significantly less than the period between the marks. So for a typical value of the period between marks of 2-3 mm, applying a shadow mask with a hole size of 0.2-0.3 mm does not represent technological problems. For confident illumination of such an element, even a power of 1 mW is enough when the laser is operating at a wavelength close to the maximum spectral sensitivity of the photocathode.

The essence of the invention is to establish an unambiguous relationship between the calibration marks on the screen, created by an electron beam from a portion of the PC illuminated by a semiconductor pulse-modulated laser on the screen area bounded by the shadow mask and the corresponding time intervals that are used for temporary calibration during processing of the eopogram. This connection is carried out either according to the mathematical formula (with the simplest forms of scan voltage), or by preliminary (at the setup stage of the image intensifier recorder) digitization at discrete points of the scan voltage. The foregoing allows you to change the laser mode, i.e. reduce the requirements for its radiation, which leads to a simplification of the implementation of the method and reduce the calibration error.

Below is a description of the implementation of the proposed method.

Figure 1 presents the timing diagrams of the signals involved in the work: registered, pulsed laser radiation (for the prototype and the proposed method) and scan.

Figure 2 shows the screen of the image intensifier tube with a conditional image of the applied mask, the direction of the temporal and spatial axis, the illuminated portion of the photocathode and the main dimensions.

On figa presents a trigger signal Uzap (t).

On figb presents the electronic equivalent of the recorded optical process, converted into the density of the electron beam at the input of the deflecting system (OS), taking into account all the delays Esig (1). OS is not involved in the claims, therefore, in figure 2 is not presented.

On figv - envelope of pulsed laser radiation Raz (t) for the proposed method and Raz pror (t) for the prototype.

Figure 1 g - voltage sweep (single phase) Upaз (t).

In figure 1, the notation:

t ₀ is the origin of the time intervals,

tsis - the total delay of the investigated signal at the input to the OS,

t _0.5 sig - width at half maximum (WPC) of the studied signal,

tlaz is the delay in turning on the pulsed laser radiation,

Eye - the duration of a packet of pulse-modulated laser radiation,

tzpr - delay start scan,

tzlu - the delay of the linear (working) section of the scan,

Traz - the duration of the working section of the sweep.

On figa schematically presents the screen of the image intensifier tube with the calibration section. The mask is indicated by 1, the schematic image of the screen with time stamps and sizes - 2. The mask creates highlighted or darkened time stamps. In FIG. 2b schematically shows a fragment of an image intensifier recorder. Accepted notation: laser 3, a portion of the fiber line 4, a schematic representation of the photocathode and the illuminated portion 5, the illuminated portion of the photocathode 6. In FIG. 2, the notation is used: ΔPi is the distance between the brightness marks, N is the number of calibration intervals, h _fc and t _fc - the length and width of the photocathode, Нрп and Lpп - the length and width of the working field of the screen along the spatial and temporal axes.

Implementation of the proposed method is as follows.

The recorded optical signal triggers the corresponding trigger element, which generates a trigger signal (Figa). The recorded optical signal is also transmitted through the input optics system to the working section of the photocathode. The photocathode converts the optical signal into an electron beam of a certain density, which, having passed the accelerating and focusing image intensifier tube, reaches the input to the OS (Fig.1b). From the starting pulse with the required delays, a pulse-modulated laser 3 and a sweep generator (not shown) are launched, providing signals in accordance with FIGS. 1 c and 1 g. Laser 3, in turn, transmits radiation, for example, over a fiber section line 4 and illuminates the portion 6 of the photocathode allocated for calibration, providing a constant illumination during the duration of the scan of the corresponding portion of the screen, deployed along the time axis in the area with the shadow mask 1 (Figure 2). In the remaining area of the screen (or part of it), a temporary scan of the signal under study takes place, in the form of a brightness-modulated eopogram strip (7 in FIG. 2a). Thus, after the end of the signal registration, there is a frame shown in Fig. 2, on which there is a section with brightness marks (at least 10) and directly an eopogram of the signal under study, which can be calibrated in the time domain using brightness marks.

To clarify the essence of the invention, we consider an example by setting the sweep voltage as a function of time U (t).

Let U (t) be the sweep voltage, which is associated with a deviation on the screen as U (t) = P (t) / S (1), where S (mm / V) is the sensitivity of the deflecting system, P (t) is the deviation, M - electron optical zoom.

The sensitivity of the OS in the general case is a frequency-dependent quantity. However, in the presence of a broadband OS, the upper cutoff frequency fgp for the beam deflection of which is significantly higher than the active spectral width of the scanning signal fa, the sensitivity S can be considered with sufficient accuracy a constant value called static sensitivity. If frp is commensurate with fa, then the frequency properties of the OS can be taken into account when calibrating the scan voltage. Then, assuming that calibration is performed at discrete points in the presence of N time intervals uniformly located on the time axis with a step Δt, the length of the corresponding interval will be (at Δt → 0): ΔP _i (t) = Δt · S · (dU (t) / dt) (2), where dU (t) / dt is the derivative of the scan signal in the ith time interval. We also take the duration of the sweep on the whole screen Tpaz = N · Δt and at t = Tpaz, U (t) = Umax.

Consider the simplest case of a linear sweep, which in practice often turns out to be admissible: U (t) = A · t, where A (V / s) is the rate of rise of the sweep voltage. From formula (2) we obtain the desired relationship of the geometric distance between the calibration marks uniformly distributed over the screen for this case and corresponding also to the uniform time intervals Δt: ΔP = Δt · S · A (3), where Δt = Traz / N.

In this case, if there are several values of the scan durations Tpj, for each of them the calibration will consist in adjusting for each scan the necessary value of the slew rate Aj (out of 3):

Aj = ΔP / (S · Δtj) (4).

In the presence of a linear and stable sweep with slew rates determined by formula (4), time intervals corresponding to time stamps for each sweep are defined in the simplest way as:

where j is the number of the corresponding scan.

For a more complex case, when the sweep voltage (for all values of Tpaзj) depends on time as:

where n is the exponent of the power function (it can be fractional), A ₁ is the calibration coefficient, A ₁ = Umax / Tp ⁿ (Umax is the value of the scan voltage at t = Tp) and A ₁ = A / Tp ^(n-1) , where A is the lanerized slew rate A = Umax / Tp, (V / s). Then, from the formula (2), substituting the derivative of the function (6), we obtain the length between the marks corresponding to the i time interval for the j-th sweep value:

where ti = i · Δtj, a Δtj is determined by the formula (5), i = 1 ... N is the number of the corresponding time interval, and

(V / s) - slew rate for each sweep, similar to the linear case, which should be provided during tuning. Thus, it can be seen that in this case, for each sweep at uniform time intervals Δtj, the length of the segments between the calibration points is uneven and is determined by formula (7). If it is necessary, as in the first case, to obtain a uniform geometric scale, then from formula (7) we obtain (omitting the index j for each scan) the value of the next time interval corresponding to a fixed length of the interval between calibration marks ΔP:

Thus, formulas (5) and (9) will determine in the two considered variants of the time dependence U (t), the relationship of time intervals with the corresponding calibration marks on the screen, while the slew rates Aj are determined for each scan value, respectively, by formulas (4) and (8).

In the General case, if for each scan value the waveform of the scan is different and is not described as described above, the proposed calibration method works as follows.

For each sweep value, when setting up the sweep unit, a highly accurate measurement of the shape of the sweep working section is made at M time intervals uniformly located on the sweep working section. This measurement can be performed once in the manufacture of a scan unit, provided that the subsequent stability of the shapes and parameters of the scan signals is ensured. This measurement can be carried out, for example, using a high-precision analog-to-digital or digital oscilloscope with an error of ~ 1-2%. To account for distortions in the shape of the sweep voltage, due to the presence of a frequency dependence of the sensitivity of the OS, consisting of the "cold" (frequency response of the OS as a four-terminal) and "hot" (effects of the frequency dependence of the deviation) components, when calibrated, the scan voltage is fed to the input of an analog-to-digital oscilloscope through "cold" deflecting system of the image intensifier tube. In this case, the shape of the deflecting voltage will be adjusted taking into account the dynamic characteristics of the “cold” OS. The “hot” component can be taken into account when using the expression for the frequency response taking into account the span effect (this expression is given, for example, in [5]), for convolution of the impulse response (IC), determined by the influence of the span effect, with the scan voltage:

U _kor (t) = U (t-τ) g (τ), where

g (τ) = Ψ (| A (ω) | · ^{j jφ (ω)} ),

g (t) - their OS due to the span effect,

U (t) is the sweep voltage during measurement,

U _kor (t) - scan voltage taking into account the span effect ,,

A (ω) and φ (ω) - frequency response and phase response of the OS, determined by the span effect,

ψ is the inverse Fourier transform

* - a convolution symbol.

After taking measurements and correcting the form U (t) for each sweep value Tj, with a step Δτ = Tj / M, there will be (M + 1) values of sweep voltage samples: U (iΔτ), i = 1 ... (M + 1). Assuming a uniform geometric scale of calibration marks in increments of: ΔP = Lrp / M, where Lpp is the length of the working field along the time axis, we obtain from the formula (2) for each scan value:

where k = 1 ... N is the number of the received time interval; i = 1 ... (M + 1) is the sample number during preliminary measurement of the shape of the working section of the sweep, and N = M.

Thus, formula (10) allows, by analogy with formulas (5) and (9), to carry out in the general case, for each scan value (the scan number index for simplification in formula (10) is omitted), a highly accurate connection of time intervals with the corresponding calibration marks on the screen. This allows us to apply the proposed method for calibrating the time axis in the study of picosecond pulsed signals using chronographic image intensifier registers, having a geometrically uniform high-precision scale for all scan values, with at least 10 points along the time axis, which increases the accuracy of the time axis calibration for all scan values .

In practice, this can be implemented by recording in the software of the image intensifier, for each scan value, the values of the specified time intervals, for the subsequent calculation of the required time values on the eopogram of the signal under study. These values can, if necessary, be verified during the life of the image intensifier.

From the description it follows that the implementation of the described method of illumination does not require a complex and expensive laser installation, which is used in the prototype, which confirms the achievement of the first technical result.

Since there are at least 10 clear and stable time stamps for all sweep values, the time calibration error by analogy with oscillography will not exceed several percent [6]. In the prototype method, due to the fact that the number and width of the marks do not correspond to the optimal values, this error reaches several tens of percent, which follows, for example, from the formulas for the calibration error given in [1].

LITERATURE

[1]. RF patent No. 1817613 dated 05/07/1997, “Time-analyzing device for registering fast processes”.

[2]. O. Brekhov et al. “Calibration of time scans of the Agat photochronograph. J. PTE No. 6, 1981

[3]. A.M. Prokhorov et al. “GaAs switch for controlling an electron-optical camera”. J. PTE No. 5, 1987

[4]. US Patent No. 4714825 of 12/22/1987, “System for calibrating the time axis of an X-Ray Streak Tube."

[5]. B.A. Shkunov, G.I.Semennik. “Broadband oscilloscope tubes and their application.” Energy, 1976

[6]. A.I. Naidenov, B.A. Novopolsky. “Electron beam oscilloscopes”, Moscow: Energatomizdat, 1983, p. 205.

Claims (1)

A method for calibrating the time base of chronographic image intensifier tubes, which consists in illuminating a portion of the photocathode of the image intensifier with pulsed laser radiation, characterized in that, in the manufacture of the electron-optical converter (EIT), a shadow mask transmitting or blocking the electron beam is applied onto the inner surface of the screen and creates a mask during scanning along the time axis image of illuminated or darkened time stamps, the number of which is at least N = 10, simultaneously with the registration of the signal under investigation yes, from which the photoelectrons enter the screen segment indicated above, are illuminated by a laser, the backlighting is performed in a single pulse-modulated semiconductor laser mode at a power level of 1 to 10 mW, with a modulating pulse duration in the microsecond range that is the same for all scan durations, the time relationship intervals with the corresponding calibration marks on the screen is carried out according to the formulas depending on the method of interpolation of the form of the voltage sweep, applicable throughout the range azone scan durations: Δt = ΔP / (S · A), Δt _i = ΔП / (S · A · n · (i / N) ^(n-1) ), Δt _k = ΔP · Δτ / (S · (U ((i + 1) · Δτ) -U (i · Δτ))), where Δt, Δt _i , Δt _k are the time intervals corresponding to the calibration marks for each form of the dependence of the scan voltage on time; ΔP is the geometric distance between the calibration marks evenly distributed on the screen along the time axis; S is the sensitivity of the deflecting system; A is the slew rate of the deflecting sweep voltage; i, k - current indices of the time interval between calibration marks; Δτ is the time interval between voltage samples when measuring its shape: U (i · Δτ) is the value of the sweep voltage at time i · Δτ when measuring its shape.

Images