Method of generating a gigahertz pulse burst and laser device therefor

Abstract

The invention provides a method of generating a burst of gigahertz laser pulses, wherein: 1) a time delay T2 of a delay section with respect to an undelayed portion of an input pulse is greater than a time period T1 between the input pulse and a next input pulse; 2) the burst of output pulses has an increasing number of pulses; 3) a burst-internal pulse spacing within the formed burst is equal to T3 = T2 - T1 and corresponds to an ultrahigh pulse repetition rate higher than 100 MHz. In another embodiment: 1) T2 is longer than M*T1, where M = 2, 3, etc.; 2) an output series of bursts consists of bursts of pulses, where M adjacent bursts have the same number of pulses; 3) T3 is equal to T3 = T2 - M*T1. A laser device for implementing the method is provided.

Description

TECHNICAL FIELD

[0001] The present invention relates generally to pulsed lasers. In particular, the present invention relates to a method and a laser device for providing bursts of short and ultrashort optical pulses at high repetition rate. BACKGROUND

[0002] Pulsed laser beams are convenient tools for material processing, medical or scientific applications. In cutting, drilling or surface structuring of materials or other industrial operations, the processing quality depends on the parameters of the laser pulses. Short and ultrashort laser pulses are preferred in high precision microprocessing because it combines high processing quality and high speed. Standard laser sources of short and ultrashort pulses provide pulses at repetition rates in the range of tens to hundreds of megahertz. However, higher repetition rates are often desired because they increase the throughput of the manufacturing process and also provide better quality because new processing regimes are ensured. GHz ablation is a very promising tool for structuring of semiconductor devices. Increasing the repetition rate from MHz to GHz level in ablation of stainless steel, silicon and tungsten reduces the temperature of the surrounding irradiated area because less laser energy is lost due to thermal diffusion. Pulse sequences with pulse intervals in the <1 ns range activate additional mechanisms of femtosecond modification of quartz glass and sapphire. Many applications require active control of the pulses per site or multiple subsequent pulses within a defined time period. Therefore, there is a need for a method and a laser device for generating short or ultrashort laser pulses at a pulse repetition rate higher than 100 MHz, preferably higher than 1 GHz.

[0003] Typically, periodic sequences of ultrashort laser pulses are generated in mode-locked lasers. Mode-locked fiber lasers of sub-picosecond pulses require gain media with large gain bandwidth. Mode-locked fiber lasers based on rare-earth doped active fibers allow to generate femtosecond pulses. High output power is not directly obtained from fiber laser oscillators but is achieved by using master oscillator power amplifier (MOPA) schemes employing fiber or solid state amplifiers.

[0004] There are several methods to generate bursts of laser pulses at GHz repetition rate. One possibility to generate bursts of laser pulses at GHz repetition rate is to generate GHz laser pulses at the oscillator and then to form the desired sequence of pulses, eliminating the unnecessary ones. In this case, the pulse repetition rate cannot be higher than the repetition rate of the oscillator. Typical mode-locked fiber laser oscillators have pulse repetition rates in the tens of MHz. Fiber laser resonators are typically too long to achieve GHz repetition rates by fundamental mode locking. Instead, harmonic mode locking is used (multiple pulses circulating inside the cavity), but it does not allow the laser to operate reliably in general.

[0005] Documents (Nakazawa et al., IEEE J. Quant. Electron. 25, pp. 2045-2052 (1989) and Aslam et al., AIP Advances 2, 022168 (2012)) propose that the interaction between nonlinear and dispersive effects in a nonlinear system can lead to modulation instability. Even in the absence of any mode-locking element in the laser cavity, this leads to the spontaneous splitting of a continuous wave into a series of pulses. The authors of the current solution demonstrate a series of pulses with tens of GHz repetition rate. However, the pulse bursts obtained from the series are very unstable and therefore not suitable for industrial applications.

[0006] Document (Tang et al., Opt. Fiber Tech., 20, pp. 610-614 (2014)) proposes a method to generate a series of GHz femtosecond pulses in a fiber resonator by the so-called cavity-induced modulation instability, which is different from the traditional modulation instability. A highly stable series of pulses with up to tens of GHz pulse repetition rate is obtained. However, the key condition to obtain the operating state in a fiber laser is related to the power and the cavity phase detuning, the pulses are chirped, and there is a weak continuous background of laser radiation. At higher power levels, the traditional modulation instability effect turns on, which is more unstable and has another pulse repetition rate. The main drawback of this solution in industrial applications is that the pulse repetition rate depends on the cavity parameters and cannot be controlled independently of other output pulse characteristics.

[0007] Document US6901174B2 (published 2005-05-31) describes a device comprising a mode-locked fiber ring laser consisting of an optical fiber, an optical amplifier, a modulator for optical modulation. The ring laser further comprises a Fabry-Perot filter and a compound cavity structure consisting of two optical fiber arms of slightly different lengths (connected by a pair of 50 / 50 optical couplers). The method of laser pulse repetition rate multiplication is based on intra-cavity optical filtering and customizing the laser cavity length via the compound cavity structure. When a high-frequency electrical signal is applied to the modulator, the modulator, which is the basis for active mode-locking, provides a series of optical pulses with a repetition frequency of f m The filter can pass the applied modulation frequency f mof the cavity modes realized by the filter. The authors of the current solution demonstrate a nine-fold multiplication of the pulse repetition rate. Thus, the device produces laser pulses at a repetition rate of up to 10 GHz. This solution, like other similar solutions based on matching lengths of several arms of the resonator, is very sensitive to temperature changes, requiring an active stabilization system. Feedback control is more difficult to organize in an oscillator. If a loss occurs, it is difficult to determine what should be changed inside the optical resonant cavity. Another difficulty is that the difference in the lengths of the arms of the compound cavity structure is small (about 2 cm compared to the total length of the cavity of 6 m), and a small deviation in the length leads to a large change in the least common multiple of the compound cavity.

[0008] There are several ways to produce a higher pulse repetition rate than the oscillator produces. The literature (Okhrimchuk et al., Sci. Rep. 7, 16563 (2017)) demonstrates a method for efficient micro-machining of silica and sapphire by attenuating the amplitude of femtosecond pulses in sub-nanosecond bursts. A modification of the intra-burst pulse separation on the scale of tens of picoseconds is observed. It corresponds to a pulse repetition rate of 10-100 GHz. An empty Fabry-Perot cavity is located on the path of the primary pulse series produced in the seed source. To produce pulse bursts with various different inter-pulse intervals, the authors use two different types of Fabry-Perot cavities: 1) a free-space cavity - made using two planar semi-transparent dielectric mirrors; 2) a monolithic Fabry-Perot cavity - consisting of a single planar parallel fused quartz plate with a partially reflective coating. The length of the free-space cavity is adjusted to produce bursts with inter-pulse intervals in the range of 70 ps - 8 ns. Depending on the thickness of the plate, the monolithic cavity produces inter-pulse intervals in the range of 10-70 ps. Each pulse from the primary pulse series is accompanied by an attenuated pulse burst. To avoid pulse distortion, a non-chirped dielectric coating is used. The accuracy of the intra-burst pulse separation and spatial alignment depends on the setup accuracy of the Fabry-Perot cavity. The main disadvantage of this solution is the exponential decay of the pulse amplitude within the burst. The length of the burst (the number of pulses with significant energy) depends on the reflection coefficient of the coating; however, this solution is only suitable for specific applications. This solution is of little use in MOPA systems, as the amplifier can better amplify the first pulse.

[0009] The empty Fabry-Perot cavity is a delay line, in which the reflected part of the pulse is combined with the non-delayed pulse, forming a pulse sequence. In a fiber laser, the delay line is an optical fiber.

[0010] Document US9209592B2 (published on 2015-12-08) provides a fiber arrangement for increasing the repetition rate of a laser. An input beam is equally power split between two arms of a 50 / 50 fiber coupler (fiber coupler, sometimes called fiber splitter, is a component used in fiber systems, intended to separate and combine radiation coupled to its input ports). An additional length of transmission fiber is inserted in one arm of the coupler. The optical path length difference between the two arms is designed to introduce a delay of half a period of the initial pulse separation. The outputs of the two arms are combined with a second 50 / 50 fiber coupler to produce a pulse train with a repetition rate twice that of the initial pulse train. This repetition rate multiplication arrangement can be inserted for example after the oscillator, after the amplifier or after the Raman shifter of a laser system. Document (Kerse et al., Opt. Commun., 366, pp. 404-409 (2016)) describes a similar solution. A fiber laser amplifier system is able to produce laser pulses up to 3.5 GHz with a number of cascaded arrangements of 50 / 50 fiber couplers with different lengths. After each 50 / 50 fiber coupler, part of the pulse passes through a fiber arm with an optical path length difference corresponding to half a period of the pulse separation at the input of this fiber coupler. Then, the partial pulses are recombined by the next fiber coupler and a pulse train with a doubled pulse repetition rate is produced. A signal from a 108 MHz fiber oscillator is converted into a 3.5 GHz pulse train by a fiber multiplier composed of six cascaded arrangements of 50 / 50 fiber couplers. The drawback of this solution is that the repetition rate multiplier is complex and not flexible. Each fiber coupler arrangement increases the pulse repetition rate by a factor of two. A gigahertz pulse repetition rate requires many identical fiber arrangements. The accuracy of the final pulse repetition rate is highly dependent on the accuracy of the optical path length difference of the fiber arms. Moreover, this solution is not suitable for ultrashort laser pulses because the dispersion of the fiber arrangements is not controlled.

[0011] Document DE102016124087B3 (published on 2017-09-28) proposes to produce laser pulses in burst operation in at least two different lengths of fiber arms. For a delay of 200 ps (corresponding to a 5 GHz pulse repetition rate), the length difference of the two fiber arms needs to be approximately 4 cm (assuming quartz glass as fiber material). The document suggests that dispersion management can be done by choosing fibers with different dispersion parameters. It leads to equal pulse durations of the burst of laser pulses. However, a burst of long pulses requires many arrangements with precisely selected fiber arm lengths and dispersion parameters.

[0012] Document FR3063395A1 (published on 2018-08-31) describes a method to form a doublet or four-pulse sequence in free space or in a fiber delay line. The fiber solution is similar to US9209592B2 but with additional elements for spectral or temporal shaping and / or time delay adjustment between pulses. The spectral shaping system can consist of spectral filters (dielectric bandpass filters or Bragg gratings with fixed pitch) while the temporal shaping can be performed, for example, using a Bragg grating stretcher with variable pitch. However, the pulse repetition rate can only be doubled. If four laser pulses are needed, two delay lines must be built. With this method, only a limited (predetermined) number of pulses can be generated within a burst. The laser source according to this solution is very suitable for LIBS measurements (temporal shaping is suitable for producing asymmetric secondary laser pulses in terms of duration and power), but not for other industrial applications.

[0013] There are other methods to increase the pulse repetition rate and / or to produce a pulse burst, where the delay line has a loop shape. With the help of a beam splitter or several beam splitters, the delay loop is incorporated into the optical path (usually downstream of the laser pulse generation system, i.e. the seed source). This way a part of the laser pulses goes directly to the output, while another part of the laser pulses goes into the delay loop and then to the beginning of the delay loop (the beam splitter). The delayed part of the pulses is again split into two parts, where one part goes to the output and the other part goes again into the loop, and so on.

[0014] The recent prior art is the method with a fiber loop. In a fiber laser system, the device desired for forming a laser pulse burst is of fiber. By this way, it can be spliced to the main fiber system, avoiding the spatial alignment problem.

[0015] Document WO2009042024A2 (published on 2009-04-02) describes a laser device comprising a pulsed fiber MOPA: it comprises a mode-locked master oscillator, a preamplifier, a fiber pulse repetition rate (PRR) multiplication device and a fiber power amplifier. The pulse train from the fiber preamplifier is injected into the fiber pulse repetition rate multiplication device; the pulse train from the PRR multiplier is injected into the power amplifier.

[0016] The fiber PRR multiplication device splits each input pulse into two pulses and delays one pulse with respect to the other. A part of one pulse and a part of the other pulse are delivered by the fiber pulse repetition rate multiplication device as an output pulse train whose PRR is equal to double the PRR of the input pulse train. The delay time within the PRR multiplication device is equal to half the inter-pulse period of the oscillator. Several fiber implementations of the PRR multiplication device are presented in this document. One of them includes a loop made of passive optical fiber and a semiconductor optical amplifier. The loop is made using a fiber coupler with two input and two output ports, and the first output port is connected to the first input port. The input pulse train is injected through the second input port, while the output pulse train is ejected through the second output port. The splitting ratio of the fiber coupler is preferably 50 / 50. The length of the fiber loop is chosen to provide a round-trip time within it that is one half of the pulse period of the input pulse train. The pulse portions that pass through the fiber loop are either amplified or blocked by a semiconductor optical amplifier, whose active and inactive states are controlled by a driving signal. When the semiconductor optical amplifier is in the active state, it amplifies the pulse portions that propagate through the fiber loop, making the pulse portions in its output train have equal amplitudes. When the semiconductor optical amplifier is not active, it blocks the pulse portions and prevents them from exiting the loop and interfering with the pulse portions of the subsequent input pulses. However, this method also has some drawbacks. The semiconductor optical amplifier used for amplitude control of the pulses per second also interferes with its spectrum and phase (due to time-dependent carrier distribution, birefringence, and dispersion). Therefore, the pulses per second can have different pulse durations and compressibility. The semiconductor optical amplifier exhibits strong superluminescence and multi-photon absorption; therefore, it can only be used for peak powers not exceeding 1 W. The proposed device is suitable for short pulses but not for ultrashort pulses. In comparison with the PRR of the master oscillator, the pulse MOPA forms an endless pulse train with double the pulse repetition rate. The highest pulse repetition rate produced by the MOPA can only reach 200 MHz; and there is no discussion of the option to form pulse bursts.

[0017] Recent state-of-the-art is described in Wei et al., Laser Phys., 26, 025104 (2016). The pulsed fiber sub-nanosecond MOPA comprises a cascade of two fiber loops forming a pulse burst. The fiber loops are composed of passive fibers of different lengths and are connected via fiber couplers with splitting ratios of 30 / 70 and 40 / 60 to the optical chain of a fiber mode-locked seed laser, a preamplifier and a power amplifier. The pulse train from the seed laser (fiber master oscillator) is injected into the second input port of the first fiber coupler. Each pulse of the pulse train is split into two unequal parts: a 30% part is transmitted as an output pulse through the second output port of the first fiber coupler; a 70% part is transmitted through the first output port into a passive fiber (loop), the first output port being connected to the first input port of the coupler. The 70% part is again split into 30% and 70% parts, which are transmitted with a certain time delay into the output and the loop, respectively. A set of delayed pulses is formed from each pulse of the master pulse train. Similarly, each pulse from the delayed pulse train is injected into the second input port of the second fiber coupler and is split into two unequal parts. A 40% part is transmitted as an output pulse through the second output port of the second fiber coupler; a 60% part is transmitted through the first output port into a passive fiber, the first output port being connected to the first input port of the coupler. The 60% part is again split into 40% and 60% parts, and so on. The time delay between the pulses is related to the length of the fiber loop. The first fiber loop is more than three times shorter than the length of the fiber oscillator. The second fiber loop is exactly twice shorter than the first fiber loop. The pulse part delayed by the first loop after one round trip coincides in time with the pulse part delayed by the second loop after two round trips. Thus, the cascade of the two fiber loops forms a burst of six pulses of different amplitudes; the third pulse has the highest amplitude. Further pulses are hardly observable due to the fast decay of the amplitudes of the pulse burst. The final pulse repetition rate is more than six times higher than the master pulse train. The solution with a second fiber loop length of 12.3 m provides a pulse repetition rate of 16 MHz (the time interval between adjacent sub-pulses is 60 ns). The method with two passive fiber loops and without additional elements has several drawbacks. The pulse burst is limited to a few pulses. The pulse amplitudes within the burst are adjusted only by the choice of the splitting ratios of the fiber couplers. No further control of the amplitudes is provided. The lengths of the fiber loops should be chosen carefully (precise cutting) to form an equidistant pulse burst. Broadband pulses propagating through the fiber as a dispersive medium experience pulse broadening. The shortest time interval between adjacent sub-pulses is the inverse of the optical length of the shorter fiber loop. Due to the two splices of the fiber loop, which require at least 10 cm (the minimum fiber lead length required for the fiber splicing process is 5 cm), the highest pulse repetition frequency that can be achieved is approximately 2 GHz.

[0018] Therefore, there is a need for a method and a laser device for generating bursts of short laser pulses which:

[0019] - can generate pulse repetition rates in the range of several hundred MHz to THz, preferably in the GHz range;

[0020] - can generate pulse repetition rates higher than the seed source;

[0021] - can generate intra-burst PRRs which are not multiples of the seed source PRR;

[0022] - can generate any number of pulses within a burst;

[0023] - can generate pulses with the same amplitude or amplitudes controlled in a desired way;

[0024] - can generate bursts of pulses with the same duration and / or compressibility;

[0025] - can generate bursts of pulses with the same time interval between pulses;

[0026] - are suitable for wideband radiation, corresponding to ultra-short transform-limited pulse durations below 1 ps;

[0027] - will be stable and cost-effective.

[0028] Our invention describes a new method for generating GHz-range bursts of short, preferably ultra-short laser pulses which meet the above requirements. SUMMARY

[0029] According to the proposed invention, the method for generating bursts of laser pulses comprises delaying a part of an input pulse with respect to another part of the input pulse, wherein the collection of the undelayed and delayed parts of many input pulses forms a burst of output pulses. It is characterized in that: 1) the time delay T2 of the delayed part with respect to the undelayed part of the input pulse is greater than the time interval T1 between the input pulse and the next input pulse. 2) the burst of output pulses has an increasing number of pulses; 3) the intra-burst pulse interval within the formed burst is equal to T3 = T2 - T1 and corresponds to a super-high pulse repetition rate higher than 100 MHz.

[0030] According to the proposed invention, another embodiment of the method for generating a burst of laser pulses is characterized by: 1) the time delay T2 of the delayed part of any input pulse with respect to the undelayed part of this input pulse is longer than M*T1, where T1 is the time period between adjacent input pulses and M = 2, 3, etc.; 2) the output train of bursts consists of bursts of pulses, where M adjacent bursts have the same number of pulses; 3) the intra-burst pulse spacing T3 of the formed bursts is equal to T3 = T2 - M*T1 and corresponds to an ultra-high pulse repetition rate higher than 100 MHz.

[0031] According to the proposed invention, a particular embodiment of the method comprises a delay line incorporated into an optical chain with an element having two input and two output ports. The first output port is connected with the first input port via said delay line. The second input port receives input pulses of a master pulse train, while the second output port provides an output train of bursts. The input pulses received at the input port and their delayed parts are split between the first and second output ports with a certain splitting ratio. The partial pulses are amplified during their propagation in said delay line and are blocked under the control of a driving signal D2.

[0032] Other particular embodiments of the method are provided. The master pulse train comprises narrowband radiation, the input pulses are transform-limited pulses with a duration between 10 ps and 10 ns. Alternatively, the master pulse train comprises wideband radiation with a bandwidth corresponding to a transform-limited pulse duration in the range below 10 ps, and the dispersion management is done during the pulse propagation in the delay line. The pulses of the burst output train have the same duration.

[0033] An embodiment of the burst synthesizer according to said method consists of an active optical fiber loop and an optical fiber coupler with two input and two output ports. The active optical fiber loop with an optical length L opt connects the first output port of the optical fiber coupler with its first input port and introduces said time delay T2. The active optical fiber loop contains a piece of undoped optical fiber, an isolator, a wavelength division multiplexer, a laser diode, a piece of doped optical fiber, an optical switch controlled by a driving signal D2. The piece of doped optical fiber is used for pulse amplification. The optical switch controlled by a driving signal D2 is used to block or attenuate the partial pulses propagating in the active optical fiber loop. In yet another partial embodiment of the burst synthesizer, the active optical fiber loop contains also a circulator and a side arm comprising a dispersion compensation element designed to completely compensate the dispersion of the active optical fiber loop at the end of a full round trip.

[0034] In other particular embodiments of the burst synthesizer, the active optical fiber loop contains at least one optical router and at least one additional piece of undoped optical fiber. The total optical length L optand the time delay T2 depend on the path travelled by the pulse within the active optical fiber loop. The active optical fiber loop, whose optical length depends on the path travelled by the pulse inside, comprises at least one dispersion compensating element designed to compensate the dispersion of the active optical fiber loop at the end of a complete round trip.

[0035] According to the proposed invention, the laser device comprises a seed source, the burst combiner of the invention and a power amplifier. A preferred embodiment of the laser device comprises a seed source, an optical switch controlled by a driving signal Dl, the burst combiner controlled by a driving signal D2, an optical switch controlled by a driving signal D3 and a power amplifier. In yet another embodiment of the laser device, the optical switch controlled by the driving signal Dl attenuates and / or blocks selected pulses of the main pulse train. In yet another embodiment of the laser device, the optical switch controlled by the driving signal D3 blocks selected bursts of the output train of bursts. BRIEF DESCRIPTION OF DRAWINGS

[0036] Figures 1A-1B : Structure and operation principle of the fiber master oscillator power amplifier (MOPA) (100) of the invention, comprising the following parts:

[0037] - a seed laser (1);

[0038] - a burst combiner (2);

[0039] - an amplifier (3).

[0040] Figure 1C : Structure and operation principle of the practical implementation of the MOPA (101) of the invention, comprising the following parts:

[0041] - a seed laser (1);

[0042] - a burst combiner (2);

[0043] - an amplifier (3);

[0044] - optical switches (7, 8).

[0045] Figures 2A-2B : The burst combiner (2) of the invention consists of an active optical fiber loop (2a) connected by a fiber coupler (70). The active optical fiber loop (2a) comprises the following parts:

[0046] - an isolator (75);

[0047] - a wavelength division multiplexer (76);

[0048] - a laser diode (80);

[0049] - a doped optical fiber (77);

[0050] - optical switch (78);

[0051] - undoped optical fiber (79).

[0052] Figure 3 - time frame showing the operation of the pulse synthesizer of the invention.

[0053] Figures 4A-4B : The burst synthesizer (2) of the invention for ultrashort pulses, contains elements for dispersion compensation. The list of parts is as follows:

[0054] - optical fiber coupler (70);

[0055] - circulator (83);

[0056] - a piece of undoped optical fiber (87);

[0057] - dispersion compensation element (81);

[0058] - optional wavelength division multiplexer (76);

[0059] - laser diode (80);

[0060] - doped optical fiber (77);

[0061] - optical switch (78);

[0062] - a piece of undoped optical fiber (79).

[0063] Figures 5A-5C : Embodiment of the burst synthesizer with selectable intra-burst pulse spacing. The improved active optical fiber loop (2b) comprises the elements of the active optical fiber loop (2a) and additional elements:

[0064] - multiple dispersion compensation elements (81a, 81b, 81c);

[0065] - multiple pieces of undoped optical fiber of different lengths (88a, 88b, 88c, 89a, 89b, 89c);

[0066] - optical router (90) with multiple output ports or several optical routers (90a, 90b) with two output ports.

[0067] Figure 6 : The burst synthesizer (2') of the invention, consists of a long active optical fiber loop (2c) connected by an optical fiber coupler (70). The active optical fiber loop (2c) comprises the elements of the active optical fiber loop (2a) or the improved active optical fiber loop (2c), but a greater delay is introduced. DETAILED DESCRIPTION

[0068] It is an object of the invention to produce bursts of short laser pulses with a selected pulse repetition rate, higher than the pulse repetition rate of the seed source, in particular in the range from 100 MHz to 1 THz. It is a further object of the invention to produce bursts of ultra-short laser pulses, in particular in the range of a few picoseconds or hundreds of femtoseconds. It is a particular object of the invention to obtain a burst of pulses with a controllable number of pulses, identical pulse duration and spacing, and a desired pulse burst amplitude envelope shape.

[0069] We propose a method to synthesize a burst of pulses with the help of an active fiber loop comprising fiber components for pulse delay, amplification and dispersion control. A burst termination element is also included within the active fiber loop. The preferred embodiment of the method is used in a pulse master oscillator power amplifier laser device. It is possible to produce bursts of GHz range pulse repetition rate with a desired length and amplitude envelope.

[0070] Figures 1A-1B A very simplified block diagram of the invention is shown. It is an object to have a method for forming a burst (or a series of bursts) from a master pulse train. The combination of a pulse synthesizer 2 between a seed laser 1 and an amplifier 3 constitutes the basis of a GHz range MOPA source 100 producing bursts of short or ultra-short laser pulses. The seed laser 1 can be a single unit (fiber or solid state pulse oscillator, or laser diode), or can consist of multiple units (oscillator followed by a preamplifier or oscillator followed by a preamplifier and additional components such as a pulse stretcher etc.). The term "short" in the following refers to a pulse duration below 10 ns. The term "ultra-short" in the following refers to a pulse duration below 10 ps. The term "pulse" is to be understood as a discrete beginning of electromagnetic radiation separated by inter-pulse periods when the light is absent or negligible. The output of the seed laser 1 is to be understood in this text as "seed pulses" or "seed pulse train" (train of primary pulses 4 in the following). A "pulse train" refers to a periodic occurrence of pulses; the spacing between adjacent pulses can be a long period of time compared to the time scale of the pulses. A "burst of pulses" refers to a series of an arbitrary number of consecutive pulses with a spacing compared to the time scale of the pulses, while a "pulse train" refers to a periodic or non-periodic series of bursts. Figure 1B

[0071] ​The mode-locked fiber oscillator is one of the main families 4 to obtain ultra-short optical pulses at high repetition rate (pulse interval T1) in the range of tens to hundreds of MHz. The burst synthesizer 2 of the invention forms a pulse train 5 in which the intra-burst pulse interval T3 is not necessarily a divisor of T1 and can be very small (ultra-high pulse repetition rate). The burst synthesizer 2 splits each input pulse into two parts and performs a delay on one of them. Moreover, additional control of the amplitude and of the dispersion is performed. The delay T2 introduced by the burst synthesizer 2 is longer than the pulse interval T1 within the main pulse train 4. Thus, the delayed part 41.2 of the first input pulse 41 is coupled out later than the non-delayed part 42.1 of the second input pulse 42. The first "burst" 51 is composed of a single pulse (the non-delayed part 41.1 of the input pulse 41); the second burst 52 is a doublet of 42.1 and 41.2 part-pulses. The subsequent bursts 53, 54, 55 with increasing number of pulses are formed in the same way. A detailed description of the burst formation will be given below. It is possible to provide constant amplitude pulses within the bursts and throughout the pulse burst train 51-55. In the simplest embodiment, the amplitude of the bursts 51-55 is equal to half the amplitude of the pulses 41-45 of the main train 4. The power amplifier 3 provides an amplified pulse train 6. The intra-burst pulse interval T3 is constant and equal to T3 = T2-T1. It is the same in all bursts, while the first pulse in each burst has a constant time period T1. The pulse amplitude in the amplified train 6 of the bursts depends on the amplification conditions.

[0072] Figure 1C A block diagram of a more complex MOPA 101 of the invention and its timing diagram are shown. This configuration is more practical, for example, when a constant number of pulses is needed in all bursts. Thus, other bursts containing a smaller number of pulses or a larger number of pulses should be eliminated from the output train. The MOPA laser device is supplemented with an optical switch 7 inserted between the seed source 1 and the pulse synthesizer 2. Moreover, another optical switch 8 is inserted between the pulse synthesizer 2 and the amplifier 3. The optical switches 7, 8 can operate as variable optical attenuators and / or as pulse selectors.

[0073] For example, only bursts containing four pulses are needed. In the given diagram, optical switch 7 has no effect on the pulse repetition rate; therefore, series 9 is the same as series 4 of the seed source. Drive signals D1 and D3 switch optical switches 7 and 8 to "ON" (maximum transmission), "OFF" (zero transmission), or intermediate (partial transmission for attenuation, as needed) states. Burst synthesizer 2 has its own internal optical switch, controlled by drive signal D2. Drive signal D2 manages the start and end of "burst evolution": when signal D2 is in the "ON" state, the number of pulses in each subsequent burst increases by one; when the state is "OFF," burst evolution stops; when D2 is "ON" again, the next burst has one more pulse. The next burst with four pulses is formed after four cycles (time interval T4 = B * T1, B—the number of pulses in the burst). The output series 10 of burst synthesizer 2 now has multiple subsets (subsequences) 10.1, 10.2, etc., with the number of pulses increasing; each subset continues until the desired number of pulses in the burst is obtained. If only those bursts containing four pulses are needed, the drive signal D3 only turns the optical switch 8 to the "ON" state when the desired bursts 21-24 are formed. Then, the series 11 of bursts 21-24 is sent to amplifier 3, where they are amplified; and the amplified burst series 12 is formed.

[0074] The burst synthesizer 2 of this invention has the characteristic of forming a desired shape of the burst amplitude envelope. The pulse amplitude is controlled by amplification conditions, and to some extent by optical switching. This is to form a constant amplitude A at the output of the entire system. out Bursts 31-34 and bursts 21-24 with increased pulse amplitude should enter power amplifier 3. The first pulse 21.1 of burst 21 has the lowest amplitude A1 at the input of power amplifier 3; the last pulse 21.4 of burst 21 has the highest amplitude A4. The first pulse 21.1 enters the amplifier with undepleted gain medium and is amplified better than the subsequent pulses 21.2-21.4. Due to gain saturation (depletion of gain medium), the last pulse 21.4 is amplified the weakest.

[0075] The aforementioned burst amplitude shaping is achieved by adjusting the amplification conditions. The amplitude A1 of the first pulse 21.1 is equal to the amplitude A of the seed pulse 41. seed Half of the amplitude of the burst 21, while the other pulses 21.2-21.4 of the burst 21 have higher amplitudes because they are amplified during propagation in the delay line of the burst synthesizer 2. The pulses in Figures 1A-1C These are described as the starting points of electromagnetic radiation that are infinitely short. Of course, they have a certain duration, very short or ultra-short. The inter-pulse interval T3 corresponds to an ultra-high, preferably gigahertz, pulse repetition rate.

[0076] The initial moment t=0 is chosen as the moment when the first part 41.1 of the first pulse 41 crosses the reference plane. The optical switch of the burst synthesizer 2 can be opened (driving signal D2) at any time between moments t1 and t2, but must be closed before the second part 41.2 of the first pulse 41 reaches said optical switch. The optical switch of the burst synthesizer 2 can be closed at any time between moments t3 and t4, but must be opened before the four delayed pulses circulating inside the burst synthesizer 2 reach the switch.

[0077] The same way can be used to synthesize different numbers of pulses within a burst. Moreover, the burst selection - the optical switch 8 is located downstream of the burst synthesizer 2 - can be performed according to the needs of the custom application. The maximum repetition rate of identical pulses depends on the number of pulses within a pulse and the initial pulse repetition rate. The other hand is related to the duration of the burst. To avoid the generated burst to overlap with the next burst, the burst duration must be shorter than the inter-pulse interval of the input series entering the burst synthesizer 2. In case of a longer burst is needed, the repetition rate must be reduced by the optical switch 7 located upstream of the burst synthesizer 2. The series of pulses 9 will form the modified pulse repetition rate.

[0078] Figure 1C The method presented in the introduction allows amplitude control of individual pulses. The internal optical switch of the burst synthesizer driven by signal D2, and the optical switch 7 driven by signal D1 can have partial transmission, and thus can be used for amplitude control. It is easier to attenuate the amplitude of individual pulses using the optical switch 7 located upstream of the burst synthesizer 2, because each new input pulse has a pulse repetition rate in the MHz range, which is a relatively low rate to change the transmission value of a standard optical switch.

[0079] The burst synthesizer 2 according to the invention comprises Figure 2A ): a 2x2 optical fiber coupler 70 and an "active optical fiber loop" 2a. The optical fiber coupler 70 has two input ports 71, 72 and two output ports 73, 74. The loop is formed by connecting the first output port 73 with the first input port 71. The main pulse series 4 is injected into the second input port 72 of the optical fiber coupler 70, while the output radiation 5 is output coupled through the second output port 74.

[0080] Figure 2BA detailed scheme of the burst synthesizer 2 is shown, depicting all elements of the active fiber loop and explaining the basic operation principle. Each input pulse 41-43 is split into two parts: one part is output-coupled through the second output port 74, while the other part is transmitted to the first output port 73 of the coupler and thus to the active fiber loop 2a. Therefore, in the next cycle, all radiation entering the input ports 71 and 72 is split between the output ports 73 and 74. The output train 5 of bursts 51-53 is formed by the pulse parts coupled to the second output port 74. Other components of the active fiber loop include: an isolator 75, a wavelength division multiplexer (WDM) 76, a piece of doped fiber 77, an optical switch 78, and a piece of undoped fiber 79. The isolator 75 prevents backscattered light from propagating back to the fiber coupler 70. The WDM 76 is used to couple pump light from a laser diode 80 into the doped fiber 77. Either core or cladding pumping can be used. Radiation propagating from input 77.1 to output 77.2 is amplified in the doped fiber. Amplification is necessary to obtain bursts of pulses with equal or controllable amplitudes.

[0081] The delay T2 within the active fiber loop 2a is determined by the pulse period T1 of the master pulse train 4 and the desired intra-burst pulse spacing T3 of the output train 5 of bursts as follows: T2 = T1 + T3. The required total optical path length L of the active fiber loop 2a opt Calculated from T2, or determined experimentally by measuring the time delay. To obtain GHz bursts, the physical length L of the loop is about 4 m for a master oscillator repetition rate of 50 MHz. To achieve the required total length L, the piece of undoped fiber 79 is spliced into a loop. Elements 75 and 76 can be separate fiber components, or a special hybrid wavelength division multiplexer-isolator (WIDM) component. The optical switch 78 is a fiber-coupled acousto- or electro-optical modulator, an electronically controllable device that allows changing the amplitude and duration of the pulses. Adjustment of the optical path length of the fiber loop 2a can be performed by heating or cooling the undoped fiber 79 or a part thereof. This allows fine-tuning the delay T2 in case of GHz bursts with an external resonator, e.g. for SOPO pumping. Alternatively, mechanical stretching of the undoped fiber 79 can be used to adjust the optical path length of the fiber loop 2a, and thus the delay T2. If all components are polarization maintaining, the output laser radiation is linearly polarized.

[0082] In the first embodiment of the invention, the splitting ratio of the fiber coupler 70 is 50 / 50. This means that the pulses entering the input ports 71 and 72 of the coupler 70 are split into equal parts and transmitted to the output ports 73 and 74. The part-pulses transmitted to port 73 propagate through the fiber loop and enter port 71 again. They are split and transmitted again to ports 73 and 74.

[0083] The whole operation isFigure 3 The explanation is made with the help of the following time frames: t=0, t=T1, t=2*T1. The simplest case is analyzed: a burst 51-53 of synthetic equal-amplitude pulses; the splitting ratio of the fiber coupler 70 is 50 / 50. The pulses are observed near the reference planes A, B, C, D located on the input ports 71, 72 and on the output ports 73, 74, equidistant from the fiber coupler 70 (see Figure 2B ). In the reference plane A, the radiation at the end of the fiber loop is observed. In the reference plane B, the input radiation (the pulse train 4 of seed pulses 41, 42, 43) is observed. In the reference plane C, a part of the radiation is observed that is output-coupled through the first output port 73. In the reference plane D, a part of the radiation is observed that is output-coupled through the second output port 74. Now all the reference planes are superimposed (as if the fiber coupler 70 had no size and did not introduce any delay).

[0084] The first time frame (t=0) illustrates that the first input pulse 41 (reference plane B) is split into two equal parts. One part 41.2 is delivered into the active fiber loop (reference plane C). The other part 41.1 is delivered to the output (reference plane D); it is the first pulse of the output radiation 5. At this moment the fiber loop end (reference plane A) is free of radiation.

[0085] The second time frame (t=T1) shows the moment when the second input pulse 42 arrives at the reference plane B. The pulse 98 that propagates inside the fiber loop has not yet arrived at the reference plane A. The pulse 98 is obtained by amplification of the pulse part 41.2 (its amplitude is controlled by the doped fiber 77 and the optical switch 78; see Figure 2B ). Other features of the radiation can also be affected by the fiber components of the fiber loop, however, it is expected that their net effect is to compensate each other and only the amplitude is doubled. The radiation that is delivered again to the loop (reference plane C) consists of a part 42.2 of the pulse 42 and a part 98.2 of the pulse 98. The radiation that is delivered to the output (reference plane D) consists of a part 42.1 of the pulse 42 and a part 98.1 of the pulse 98. The pulse parts 98.2 and 98.1 are delayed with respect to the pulse parts 42.2 and 42.1. The pulses 42.1 and 98.1 form a pulse burst 52 of the output radiation 5. The amplitudes of 42.1 and 98.1 are equal.

[0086] The third time frame (t = 2*T1) illustrates that when the third input pulse 43 reaches the reference plane B, the radiation pulses 99 propagating inside the fiber have not yet reached the reference plane A. The radiation 99 at the end of the active fiber loop is formed by the pulses 42.2, 98.2 by doubling their amplitudes. The radiation delivered to the loop again (reference plane C) is formed by the part 43.2 of the pulse 42.2 and the part 99.2 of the radiation 99. The radiation delivered to the output (reference plane D) is formed by the part 43.1 of the pulse 43 and the part 99.1 of the radiation 99; it is the burst 53 of the output radiation 5. Other bursts of the output radiation 5 are formed in the same way.

[0087] Figure 2B The embodiment of the burst synthesizer 2 depicted does not have dispersion compensation elements. Therefore, it is suitable for narrowband radiation corresponding to transform-limited pulse widths greater than or equal to 10 ps. If short bursts are needed, the pulse synthesizer of Figure 2B The pulse synthesizer of the burst synthesizer 2 can be used for moderate bandwidth radiation. The necessity of compensating dispersion also depends on the length of the burst.

[0088] For example, to form a burst of 11 pulses, the part 41.2 of the first input pulse 41 circulates 10 times to and fro inside the active fiber loop; the part 42.2 of the second input pulse 42 circulates 9 times to and fro; and so on.

[0089] After 10 round trips inside the 4 m fiber loop, a transform-limited 5 ps (bandwidth ≈ 0.31 nm) near-Gaussian pulse can be broadened to 5.02 ps, after 30 round trips - up to 5.17 ps, and after 50 round trips - up to 5.47 ps. In some applications, the variation of the pulse width within a burst of 50 pulses is still allowed.

[0090] After 10 round trips inside the 4 m fiber loop, a transform-limited 1 ps (bandwidth ≈ 1.55 nm) pulse can be broadened to 2.40 ps, but after 30 round trips - to 6.68 ps. Without dispersion management components, the use of the burst synthesizer of Figure 2B The burst synthesizer of the burst synthesizer 2 can be used for moderate bandwidth radiation. The necessity of compensating dispersion also depends on the length of the burst.

[0091] Figure 4AAnother embodiment of the inventive pulse combiner 2 is shown. Preferably, the broadband radiation, as the dispersion compensating element 81 is included. Dispersion control is necessary for broadband radiation (ultra-short transform-limited pulses or short chirped pulses). The delayed pulses travel a longer path in the dispersive medium compared to the undelayed pulses. Due to dispersion, the various spectral components experience different phase shifts. As a result, the duration of the delayed pulses can change and the pulse shape can distort. Each subsequent pulse of the burst travels a longer path. The spectral phase differences accumulate. This means that the second pulse of the burst will be different from the first pulse, the third pulse will be different from the second and the first pulse, and so on. To compensate for the differences in the spectral phase and to obtain a burst of pulses with the same duration, an element 81 with dispersion opposite to that of the fiber within the active fiber loop (elements 71-79) is needed. If the seed source is a Ytterbium doped laser and radiates around 1 pm wavelength, most of the fibers have positive (normal) dispersion in this spectral range. The dispersion of element 81 must be negative (anomalous). For Tm or Ho doped laser sources that produce around 2 pm, the burst combiner 2 must include a dispersion compensating element 81 with positive (normal) dispersion. The most probable choice for the dispersion compensating element in a fiber system is a chirped fiber Bragg grating (CFBG), however, other types of dispersion compensating elements (e.g., a chirped volume Bragg grating (CVBG) or a photonic crystal fiber) can also be used. Since the radiation power inside the active fiber loop is relatively low and the nonlinear effects are negligible, only the second order dispersion (group velocity dispersion GVD) is observed. A CFBG with any selected chirp parameter for compensating the second order dispersion is commercially available. Thus, the phase shift experienced during the propagation within the active fiber loop is fully compensated. A CFBG with higher order dispersion compensation can also be used. If the nonlinear processes within the fiber loop (mainly the doped fiber 77 or subsequent elements) cannot be neglected, they should be used. The only case that does not require compensation is around the zero dispersion point.

[0092] Figure 4A The path inside the active fiber loop is as follows: the radiation from path 82 is turned by element 83 into path 84, propagates inside the dispersion compensating element 81, is turned to optical path 85 and path 86. If the dispersion compensating element 81 is a chirped fiber Bragg grating, it has a single fiber connector and acts as a mirror, i.e. reflects the light back into the same fiber (the reflected light has been dispersion managed). Thus, paths 84 and 85 are in the same fiber 87 and element 83 is a fiber circulator. It also turns the light from path 85 to path 86. The circulator 83 simultaneously acts as an isolator, preventing the light to propagate in the reverse direction. Figure 4AThe side arm (elements 83, 87, 81) of the active fiber loop 2a in Fig. 5 is located just after the fiber coupler 70. Thus, the dispersion control is performed before the pulse propagates in the dispersive loop; i.e. the dispersion management is done in advance (called pre-compensation). The duration of the pulse varies within the loop. In the ideal case of compensation, the pulse duration at the end of the loop is the same as the duration of the part of the pulse that is output-coupled through 74 without propagating through the fiber loop. It is also possible to insert the side arm (elements 83, 87, 81) at other locations of the active fiber loop 2a, but it is not recommended between the WDM 76 and the doped fiber 77. The dispersion of the chirped fiber Bragg grating 81 can be controlled by placing it in a heated mount. It allows fine tuning of the dispersion, and thus of the pulse duration.

[0093] There are several aspects to choose the right location for the side arm responsible for dispersion control. Let us analyze the case of Fig. 5 - the side arm is located before the amplification system (elements 76, 77, 80). If the initial pulses (pulses of series 4) and their parts coupled into the active fiber loop are not chirped, then the side arm of the loop chirps them. This means that the amplification in the doped fiber 77 is done on the stretched pulses, and nonlinear effects are not observable. If the initial pulses are chirped, then the side arm can reduce their chirp and reduce their duration. Due to the increased peak power, nonlinear effects can occur more likely. If the side arm is placed at other locations, then isolators are needed, or a WIDM is used instead of a WDM. Figure 4A

[0094] Figure 4B Another embodiment of a burst synthesizer 2 with dispersion compensation is shown. Here, the doped fiber section 77 is inserted in the side arm of the active fiber loop. Thus, the radiation propagates twice through the amplification medium, and the inversion is used more efficiently. The pump radiation from the laser diode 80 is coupled through the dispersion compensation element 81, so that no WDM is needed. Figures 4A-4B The pulse train synthesizer of Fig. 4 is suitable for pulses of < 1 ps. Other alternative configurations of the active fiber loop 2a of the burst synthesizer 2 can be obvious to the skilled person.

[0095] Figures 5A-5C An embodiment of a burst synthesizer 2 with selectable intra-burst pulse spacing is shown. The active fiber loop 2b of the burst synthesizer has a selectable length. There are several ways to achieve a selectable length. The loop can have an additional side arm with optical length control, or the existing side arm can be supplemented with an optical router (an optical switch type that selectively directs an optical signal from one channel to another).

[0096] Figure 5B ​A part of the active optical fiber loop 2b with a selectable length is depicted. The entire modified active optical fiber loop 2b comprises the elements of the active optical fiber loop 2a and several additional elements. The loop side arm 87 connected to the main loop with the help of the circulator 83 ends with an undoped fiber segment 88a, 88b or 88c. An optical router 90 with multiple output ports directs the radiation to one of the alternative segments 88a-88c. The various lengths of the active optical fiber loop 2b are achieved by selecting the lengths of the multiple fiber segments. The total length of the loop 2b is equal to LI when the radiation propagates through the segments 87 and 88a. The total length of the loop 2b is equal to L2 when the radiation propagates through the segments 87 and 88b. The total length of the loop 2b is L3 when the radiation propagates through the segments 87 and 88c. The different lengths of the active optical fiber loop 2b ensure the possibility to select one of the alternative intra-burst pulse repetition rates. The selection is made by controlling the router 90 with the driving signal D4. For ultrashort pulses it is important to compensate the dispersion acquired during the propagation in the optical fiber. Therefore, different lengths of the active optical fiber loop 2b require different dispersion compensation elements 81a, 81b, 81c. For short pulses the dispersion management is not that important (especially if LI, L2 and L3 are not significantly different); the same dispersion compensation elements 81a, 81b, 81c or a single element can be used. Various types of optical routers with multiple output ports are suitable.

[0097] Figure 5C A more practical implementation of the active optical fiber loop 2b with a selectable length is depicted. Several 1x2 optical routers (90a, 90b) are used. The alternative lengths of the active optical fiber loop 2b are achieved by selecting the lengths of the multiple fiber segments: LI - when the radiation propagates through the segments 87 and 88a; L2 - when the radiation propagates through the segments 87, 89a and 89b; L2 - when the radiation propagates through the segments 87, 89a and 89c. In the same way, a burst synthesizer with many selectable intra-burst pulse repetition rates can be implemented.

[0098] Figure 6 An alternative method of forming a burst of pulses is described. A burst synthesizer 2' with a long active optical fiber loop 2c in order to ensure that the pulses are delayed more than two periods Tl can form a series of bursts, where multiple subsequent bursts contain the same number of pulses. If the length of the active optical fiber loop 2c satisfies the rule: T2 = 2*Tl + T3, then two subsequent bursts have the same number of pulses (see Fig. 6). The active optical fiber loop 2c is connected to the main loop with the help of the circulator 83. The loop side arm 87 ends with an undoped fiber segment 88a, 88b or 88c. An optical router 90 with multiple output ports directs the radiation to one of the alternative segments 88a-88c. The selection of the alternative intra-burst pulse repetition rate is made by controlling the router 90 with the driving signal D4. The total length of the loop 2b is equal to LI when the radiation propagates through the segments 87 and 88a. The total length of the loop 2b is equal to L2 when the radiation propagates through the segments 87 and 88b. The total length of the loop 2b is L3 when the radiation propagates through the segments 87 and 88c. The different lengths of the active optical fiber loop 2b ensure the possibility to select one of the alternative intra-burst pulse repetition rates. The selection is made by controlling the router 90 with the driving signal D4. For ultrashort pulses it is important to compensate the dispersion acquired during the propagation in the optical fiber. Therefore, different lengths of the active optical fiber loop 2b require different dispersion compensation elements 81a, 81b, 81c. For short pulses the dispersion management is not that important (especially if LI, L2 and L3 are not significantly different); the same dispersion compensation elements 81a, 81b, 81c or a single element can be used. Various types of optical routers with multiple output ports are suitable. Figure 6). The first burst 91, 92 of the first series E contains one pulse; the second burst 93, 94 of the second series F contains two pulses, etc. Two identical bursts occur, spaced by Tl equal to the inter-pulse spacing of the master series. This can be an advantage in some applications requiring identical bursts repeated at small time intervals. However, the time of the multiple bursts forming many pulses is increased. However, if a certain application requires several bursts and can tolerate a longer time interval between the next series of several identical bursts, the burst synthesizer 2' is very suitable. The time period T5 required to obtain the first burst 93 containing two pulses is: T5 = 2*Tl. The time period to obtain the first burst containing three pulses is equal to 3*Tl. The rule to synthesize M subsequent bursts repeating at period Tl is to create a long active fiber loop 2c which satisfies the rule: T2 = M*Tl + T3. The time period required to obtain the first burst containing N pulses is equal to N*Tl.

[0099] A significant feature and advantage of the burst synthesizer and MOPA of the present invention is that a relatively slow amplitude modulator can be used. It is possible to use the optical switch 7 to control the amplitude of a single pulse in a GHz burst with very high precision, as it has to be faster than the full round trip time of the fiber loop. For example, for a seed source with a repetition frequency of 50 MHz, the opening and closing time of the optical switch has to be less than 20 ns. This is a moderate requirement on the fiber modulator. A slowly varying amplitude envelope of the burst can be successfully implemented with an internal optical switch 78. Yet another advantage of the present invention is that the number of pulses within a burst can be selected and changed by driving the optical switch only. There is no limitation on the pulse repetition rate. The length of the active fiber loop is in the range of a few meters, while the inter-pulse spacing within a burst is in the range of gigahertz or even terahertz. The active fiber loop contains all the necessary components for pulse amplitude control and dispersion compensation. Pulse bursts shorter than 1 ps can be generated with the same pulse duration and repetition rate. Variants of the method of the present invention (given below) allow to synthesize long bursts which are particularly attractive for certain applications. The only limitation on the burst length is the precision of the dispersion compensation.

[0100] According to a preferred embodiment of the present invention, the pulses of each burst do not overlap. This means that the minimum inter-pulse spacing within a burst T3 is equal to the pulse duration. For example, 10 ps duration pulses can be packed without overlapping at a 100 GHz repetition rate; 1 ps pulses allow for 1 THz. If chirped pulses with a duration of 150 ps are used, it is possible to create 6 GHz bursts.

[0101] Today, the applications of burst-mode pulsed lasers extend beyond various micromachining and materials processing operations. Similar advantages are found in radiative interactions with biological objects: the amount of information collected per unit time can be increased without exceeding the damage threshold of the biological object. Bursts of short pulses are used to generate photoelectron packets by irradiating the photocathode of an electron accelerator. For such applications, precise spacing between light pulses is crucial, and this is relatively easy to achieve using the method of this invention. High repetition rates open new possibilities for nonlinear frequency conversion in gases, liquids, and crystals. The design and complexity of synchronously pumped OPOs or coherent pulse stacking systems are highly dependent on the repetition rate accuracy of the pump radiation.

[0102] Other Embodiments

[0103] A long burst longer than T1 can be formed as follows: 1) The first part of the burst is formed in the same manner as described above; 2) When a set of pulses circulating within the loop fills the entire length of the loop, the optical switch ( Figure 1C 7) Block the next input pulse. Pulses circulating within the loop are output during the next loop and combined with those output in the previous loop, thus forming a continuous pulse burst. Long bursts can be formed by keeping the input pulse blocked. Since amplification is performed in the active fiber loop, the pulse amplitude is not reduced. Because dispersion management is performed in the active fiber loop, the pulse duration can be preserved.

[0104] For example, if the burst pulse interval is 0.5 ns (2 GHz), then 41 pulses fill a 4 m long fiber loop. An 80-pulse burst can be formed within a 39*T1 time period: a 40-pulse burst grows incrementally over 39 cycles, and then is output before the 40 pulses circulating inside the loop are combined in the next cycle, even though no input pulse is added to the active fiber loop. If, after the next T1 cycle, another input pulse is blocked by optical switch 7, the burst is lengthened by another 40 pulses from the loop.

[0105] The time interval required for a long burst to form is equal to the time required to fill the entire loop length with pulses. The only limitation is the closing time of the optical switch 78 in the active fiber loop. Switch 78 is used to terminate the burst. Because the pulses within the loop are tightly packed in time (GHz pulse repetition rate), the closing time of switch 78 will form the tail of the burst. For example, if we generate a 120-pulse burst with a duration of 1 ns within the burst and use an acousto-optic modulator with a closing time of 6 ns, we will obtain 114 pulses of the same amplitude and 6 pulses with diminished amplitude.

[0106] Another method of forming long bursts is: 1) to form a train 9 with large pulse separation - larger than the duration of the desired long burst - with the help of optical switch 7; 2) to use a long active fiber loop which introduces a time delay slightly longer than said large pulse separation. The idea is to form the whole long burst in the way described in Figures 1A-3 the middle: the first pulse of the burst is part of the input pulse, the other pulses of the burst are from the loop; they are part of the input pulse of the previous cycle. The advantage of this method of synthesizing long bursts is to control the amplitude of each pulse and to obtain a rectangular burst envelope.

[0107] The method and the laser device of the invention can also be used to synthesize long rectangular pulses from several transform-limited pulses of 10 ps - 10 ns. If the length of the active fiber loop is such that the delayed part 41.2 of the input pulse 41 (see for example Figures 1B-1C ) overlaps with the non-delayed part 42.1 of the next input pulse 42, the result is a prolonged pulse. If T3 is smaller than the pulse duration, M-shaped pulses or rectangular pulses can be obtained.

[0108] The application of the method of the invention in a wideband laser source has additional aspects. The wideband radiation of the seed source can be delivered to the burst synthesizer of the invention as a train of transform-limited pulses or as a train of chirped pulses. A variant of the MOPA of the invention includes a pulse stretcher and a compressor. The pulse stretcher inserted between the seed source 1 and the burst synthesizer 2 (in the MOPA 100 of Figure 1B ) or between the seed source 1 and the optical switch 7 (in the MOPA 101 of Figure 1C ) chirps and thus temporally stretches the pulses. If the time period T3 is smaller than the duration of the chirped pulses, the pulses overlap in the fiber coupler of the burst synthesizer 2. The radiation spectrum is modified due to the interference of the delayed part of the input pulse and the next input pulse. The interference of many input pulses results in some spectral modulation of the radiation at the output of the burst synthesizer 2 of the invention. A pulse compressor added to the MOPA after the power amplifier 3 compresses the radiation and produces ultrashort pulses. The output pulse envelope is also modulated.

[0109] The active fiber loops 2a, 2b or 2c of the burst synthesizer 2 have many other alternative configurations which are not shown in the given figures without departing from the main idea of the invention. In principle, the same idea can be implemented in a free-space equivalent of the loop but with the need to control the diffraction in order to avoid the spreading of the beam while circulating many times back and forth inside the loop. Moreover, many alternative configurations of the MOPA are not shown in the given figures without departing from the main idea of the invention. The whole scope of the invention is protected by the appended claims.

Claims

1. A method for generating a burst of laser pulses from a master pulse sequence, the method comprising delaying a portion of an input pulse with respect to another portion of the input pulse, wherein, A collection of undelayed and delayed portions of a plurality of input pulses forms a burst of output pulses, characterized in that the time delay T2 of the delayed portion of the input pulse with respect to the undelayed portion is longer than the time interval T1 between the input pulse and the next input pulse; the burst of output pulses has an increasing number of pulses; the pulse interval T3 within the burst of the output pulse is T2-T1, and corresponds to an ultra-high pulse repetition rate of over 100MHz; a delay line for the delayed portion of the input pulses is coupled to an optical fiber coupler having a first input port (71), a second input port (72), a first output port (73), and a second output port (74). The optical chain of (70) is combined as follows: the first output port (73) is connected to the first input port (71) via the delay line, the second input port (72) receives the input pulses of the main pulse series, and the second output port (74) transmits the burst output series (5); the input pulses received at the first input port (71) and the second input port (72) and their delayed portions are separated between the first output port (73) and the second output port (74) at a certain splitting ratio; some pulses are amplified during propagation in the delay line; the pulse portions propagating in the delay line are blocked under the control of the drive signal D2.

2. The method according to claim 1, characterized in that, The main pulse series includes narrowband radiation, and the input pulse is a transformation-limited pulse with a duration in the range of 10 ps to 10 ns.

3. The method according to claim 1 or 2, characterized in that, The main pulse series includes broadband radiation with a bandwidth corresponding to the pulse duration of the transformation limit in the range of less than 10 ps, ​​and dispersion management is performed during pulse propagation in the delay line; the pulses of the burst output series (5) have the same duration.

4. The method according to claim 3, characterized in that, The pulses in the main pulse series are subject to transformation restrictions.

5. A laser apparatus for implementing the method for generating laser pulse bursts from a master pulse series as described in claim 1, characterized in that, The laser device includes: a seed source (1), a burst synthesizer (2), and a power amplifier (3). The burst synthesizer (2) consists of an active fiber optic loop and a fiber coupler (70) having a first input port (71), a second input port (72), a first output port (73), and a second output port (74). The active fiber optic loop connects the first output port (73) of the fiber coupler (70) to its first input port (71). The active fiber optic loop includes: an isolator (75); a wavelength division multiplexer (76); a laser diode (80); a doped fiber (77); an optical switch (78) controlled by a drive signal D2; and an undoped fiber (79). The doped fiber (77) is used for pulse amplification, while the optical switch (78) is used to block or attenuate the pulse portion propagating in the active fiber optic loop.

6. The laser device according to claim 5, characterized in that, The active fiber loop further includes a circulator (83) and a side arm (87), the side arm (87) including a dispersion compensation element (81); wherein the dispersion compensation element (81) is designed to compensate for the dispersion of the active fiber loop at the end of a complete round trip.

7. The laser device according to claim 5, characterized in that, The active fiber loop further includes: a circulator (83) and a side arm (87), including at least one optical router controlled by drive signals D4 and D5, and at least one additional undoped fiber segment; The at least one optical router is designed to direct pulses to a selected additional segment of undoped optical fiber; The total optical length Lopt and time delay T2 of the active fiber loop depend on the path along which the pulse propagates within the active fiber loop.

8. The laser device according to claim 5, characterized in that, The active fiber loop further includes at least one dispersion compensation element; wherein the at least one dispersion compensation element is designed to compensate for the dispersion of the active fiber loop at the end of a complete round trip.

Images