JP5202315B2 - Bandwidth control of multi-chamber gas discharge laser by discharge timing - Google Patents

Description

The present invention relates to multi-chamber line narrowing gas discharges, such as excimer laser systems.
CROSS-REFERENCE TO RELATED APPLICATIONS This application Attorney Docket No. Name of U.S. Patent referred to as "multi-chamber gas discharge laser bandwidth control according to the discharge timing" of December 29, 2005 filed an No. 2005-0042-01 Application US Patent Application No. 11 / 323,604 which claims priority and is filed Aug. 29, 2001, entitled "Ultra-narrow band two-chamber high repetition rate gas discharge laser system" US Patent Application No. 10 / 012,002, filed Nov. 30, 2001, which is a continuation-in-part of 09 / 943,343, and has an agent serial number of 2001-0090-06. Continuation of U.S. Patent Application No. 10 / 627,215 entitled "Ultra Narrow Band Two-Room High Repetition Rate Gas Discharge Laser System" filed July 24, 2003 US patent application Ser. No. 11/199, entitled “Ultra Narrow Band 2 Chamber High Repetition Rate Gas Discharge Laser System,” filed Aug. 9, 2005, having agent serial number 2001-0090-15. No. 691 is a continuation-in-part application, and the disclosures of these patents are incorporated herein by reference.

No. 6,359,922 entitled “Single Chamber Gas Discharge Laser with Line-Narrowed Seed Beam” granted on Mar. 19, 2002, and this application was granted on Apr. 9, 2002. No. 6,370,174 entitled “Implanted Seed F ₂ Lithography Laser” and No. 6,381 entitled “Ultra Narrow Band Implanted Seed F ₂ Lithography Laser” granted April 30, 2002. No. 257, No. 6,765,945 entitled “Injected Seed F ₂ Laser with Preinjection Filter” granted on July 20, 2004, and granted on April 15, 2003. and No. 6,549,551, entitled implantation seed laser "with precise timing control, and granted on April 29, 2003" implantation seed F ₂ laser having a center line wavelength control " And No. 6,556,600 of cormorants name, and No. 6,590,922, entitled granted on July 8, 2003 "injection seed F ₂ laser with a line selection and distinction", 2003 No. 6, 567, 450 entitled “Ultra-narrow-band two-chamber high repetition rate gas discharge laser system” granted on May 20, and “Ultra-narrow-band two-chamber” granted on September 23, 2003 No. 6,625,191 entitled “High Repetition Number Gas Discharge Laser System” and No. 6, entitled “Timing Control for Two-chamber Gas Discharge Laser System System” granted on March 8, 2005, No. 865, 210, No. 6, 801, 560 entitled “Line Selection F ₂ Two-Room Laser System” granted on October 5, 2004, and “2” granted on February 10, 2004. Chamber gas discharge laser No. 6,690,704 entitled “Control System for Stem” and No. 6,704, entitled “Lithography Laser with Beam Delivery and Beam Orientation Control” granted on March 9, 2004. No. 339, No. 6,798,812 entitled “Two-chamber F ₂ laser system with F ₂ pressure-based line selection” granted on September 28, 2004, and “Long Life Optics” No. 6,904,073 entitled “High Power Deep Ultraviolet Laser” and No. 6,912,052 entitled “Gas Discharge MOPA Laser Spectrum Analysis Module” granted to Rao et al. On June 28, 2005. "Automatic gas for gas discharge laser" filed on Jan. 31, 2003 with agent serial number 2003-0001-01, which is related to US Patent Application Serial No. 10 / 356,168 entitled “System” and “Gas Discharge MOPA Laser System Output” filed on Dec. 18, 2003 with Attorney Docket No. 2003-0053-01. No. 10 / 740,659 entitled "Method and apparatus for control" and "Control system for two-chamber gas discharge laser" filed Jul. 30, 2003, having agent serial number 2003-0025-01. No. 10 / 631,349, and No. 10 entitled “Method and Apparatus for Cooling Magnetic Circuit Elements” filed on June 25, 2003, having an Attorney Docket No. 2003-0051-01. No. 607, 407 and “Development of Gas Discharge MOPA Laser System” filed on Dec. 18, 2003 with agent serial number 2003-0053. No. 10 / 740,659, entitled "Method and apparatus for controlling the system" and "Line selection F2 two-chamber laser system" filed on March 18, 2004 with agent serial number 1999-0013-14 No. 10 / 804,281, and No. 10/854, entitled “Line Select F2 Two-Room Laser System”, filed May 25, 2004, having agent serial number 1999-0013-15. No. 614, and No. 10 / 922,692 entitled “Timing Control for a Two-chamber Gas Discharge Laser System” filed on Aug. 20, 2004, with agent serial number 2001-0091-08, The name “Multi-room Excimer or Molecular Fluorine Gas Discharge Laser Fluorine Injection Control” filed on Sep. 29, 2004, having agent serial number 2004-0083-01 No. 10/953, No. 100. All of the above disclosure is incorporated herein by reference.

Used in various applications for lithography such as, for example, excimer laser technology, eg KrF at 248 nm, ArF at 193 nm, or F _{2 at} 157 nm, and currently emerging technologies, eg used in integrated circuit manufacturing lithographic processing For high power laser light sources, there is a need for laser output power in excess of 60 watts, and perhaps up to 200 watts to maintain desirable photolithography processing capabilities. Traditionally, the increase in power to meet the needs of, for example, integrated circuit photolithography businesses, always increases with the energy per pulse remaining relatively constant and the higher the repetition rate, the higher the average output wattage. Obtained through the number of pulse repetitions. For example, the demand for spectral power in both KrF and ArF has recently increased to 100 W / pm or more.

  Thus, typical laser systems up to about a year ago were operating at 4 kHz to obtain maximum output power, but to reach 60 watts, a one-chamber narrowed gas discharge, such as an excimer gas discharge There was a need for higher repetition rates for laser systems. While much effort has been made to solve many engineering challenges even by increasing the pulse repetition rate by 50%, ie 6 kHz, the challenge of obtaining more than 60 watts is, for example, About a year ago by the introduction of the combined “MOPA” laser system / architecture, which is a multi-chamber configuration including a master oscillator (MO) and a power amplifier (PA) by the assignee “Cymer, Inc.” It has been resolved. A similar master oscillator seeded laser system with other amplifier configurations such as a power oscillator (PO) can also be used. However, for the sake of brevity, except when explicitly indicated otherwise, the term MOPA or the terms MO and PA are used to describe any such multi-chamber laser system, eg, a two-chamber laser system, eg, Any kind of example mentioned above, which optimizes the beam parameter quality and then adjusts for this amplification process, leaving the specific beam quality parameters for the amplification function and optimized largely in the main oscillator part. It should be construed separately to include an oscillator seed pulse generation portion followed by amplification of the seed by an amplification portion that receives the seed pulse.

The main oscillator forms a generally conventional single-chamber laser oscillator cavity, such as the one sold by the assignee of the present applicant, for example under the name 6XXX or 7XXX series laser systems. However, the main oscillator in the MOPA or MOPO configuration can be used in the same laser chamber, regardless of excess energy absorption through an oscillator laser system in an attempt to produce a high power output, for example about 5 to 10 mJ, with additional line narrowing optics. Special tweaks can be made for width optimization and center wavelength control, and / or other beam parameter quality optimization. Such losses are particularly so-called line narrowing packages that reduce the relatively wide bandwidth output of gas discharge lasers, such as excimer lasers, measured in nanometers, to an output bandwidth of about picometers or less ( In LNP). Further, the amplifying part is then passed through the laser gain medium, eg to generate lasing while being generated between the electrodes in eg PA while the output of the MO is passing through the gain medium or eg in a PO configuration. It can be optimized for amplification of this output of the MO. Thereby, for example, about 15 mJ / pulse or more at 4 kHz can be generated by the assignee's assignee's XLA200 series MOPA configuration laser system, thus, for example in the case of ArF, a 60 watt output laser power and higher pulse energy and In the case of output power, eg 80 KrF, 80 pulses are generated and the functional limit is more influenced by the optical lifetime under the energy density obtained at 248 nm (KrF) or 193 nm (ArF) or 157 nm (F ₂ ). The

In the MO, for example to the electrode in the MO after receiving for example a trigger signal from a photolithographic scanner, for example, with this function of optimizing the beam parameters, for example the bandwidth and the amplification part, for example in the PA the output in the MO To continue to meet customer requirements for multi-chambers, such as bandwidth stability, dose control and stability arising from MOPA architecture, with timing of electrical energy supply to PA electrodes in a controlled time after gas discharge There are many engineering challenges.
Applicants propose certain control functions for multi-chamber laser control systems to address the presented requirements, such as those described above.

US patent application Ser. No. 11 / 323,604 US patent application Ser. No. 09 / 943,343 US patent application Ser. No. 10 / 012,002 US patent application Ser. No. 10 / 627,215 US patent application Ser. No. 11 / 199,691 US Pat. No. 6,359,922 US Pat. No. 6,370,174 US Pat. No. 6,381,257 US Pat. No. 6,765,945 US Pat. No. 6,549,551 US Pat. No. 6,556,600 US Pat. No. 6,590,922 US Pat. No. 6,567,450 US Pat. No. 6,625,191 US Pat. No. 6,865,210 US Pat. No. 6,801,560 US Pat. No. 6,690,704 US Pat. No. 6,704,339 US Pat. No. 6,798,812 US Pat. No. 6,904,073 US Pat. No. 6,912,052 US patent application Ser. No. 10 / 356,168 US patent application Ser. No. 10 / 740,659 US patent application Ser. No. 10 / 631,349 US patent application Ser. No. 10 / 607,407 US patent application Ser. No. 10 / 740,659 US patent application Ser. No. 10 / 804,281 US patent application Ser. No. 10 / 854,614 US patent application Ser. No. 10 / 922,692 US patent application Ser. No. 10 / 953,100

A method and apparatus for controlling the bandwidth of a multi-part laser system including a first line narrowing oscillator laser system part that supplies a line narrowing seed pulse to a second amplifier laser system part, the first laser system Disclosed is the selection of the differential firing time between the generation of the seed pulse in the part and the generation of the laser gain medium in the amplifier laser system part affects the bandwidth of the laser output light pulse from the multipart laser system. This includes adjusting the differential firing time as a function of the measurement bandwidth and bandwidth target, estimating the current operating point, and halogen gas injection as a function of the current operating point and the desired operating point. Adjusting. Estimating the current operating point can include the use of a function that can be easily calculated or estimated from available measurements of laser system operating parameters that are monotonic with respect to differential firing time over the expected operating range. . The desired operating point can be determined as a function of at least one of target bandwidth, laser system duty cycle, and laser system output pulse energy. Estimating the current operating point may include utilizing a difference between a current differential firing time and a reference differential firing time. The method and apparatus can further include adjusting the differential firing time as a function of bandwidth error. The method and apparatus can include selecting a reference differential timing to optimize laser system efficiency. The method and apparatus provide a derivative of a laser system output pulse energy with respect to a differential firing time at a constant voltage and a difference between a current differential firing time and a reference differential firing time, and a laser system at a current operating point. Estimating as a function of at least one of the output pulse energies may be included. The method and apparatus provides a difference between a current differential firing time and a reference differential firing time, a derivative of the laser system discharge voltage with respect to the differential firing time at a constant energy, a laser system output pulse with respect to the laser system discharge voltage. Estimating as a function of at least one of a derivative of energy and a laser system output pulse energy at a current operating point may be included. The method and apparatus estimate a current operating point as a function of at least one of a derivative of a laser system discharge voltage with respect to a differential firing time at a constant energy and a laser system output pulse energy at the current operating point. Steps may be included. The method and apparatus may include estimating a current operating point as (1 / E) ^* dV / dt, where E is the laser system output pulse energy and dV / dt is the current operating point. A derivative of the laser system discharge voltage with respect to the differential firing time at a constant energy at a point. The method and apparatus provides a current operating point as a function of at least one of a derivative of laser system output pulse energy with respect to a differential firing time at a constant voltage and a laser system output pulse energy at the current operating point. An estimation step can be included. The method and apparatus can include estimating a current operating point that includes utilizing the relationship (1 / E) ^* dE / dt, where E is the laser system output pulse energy; dE / dt is the derivative of the laser system output pulse energy with respect to the differential firing time at a constant voltage at the current operating point. dE / dt can be estimated by applying a dither signal to the differential firing time and calculating dE / dt using the dither, laser output pulse energy, and actual differential firing time. dE / dt can also be estimated by applying a dither signal to the differential firing time and taking the respective ratio of the dither laser output pulse energy and the correlation with the actual differential firing time. dE / dt can also be determined from the product of dE / dV and dV / dt. The method and apparatus estimate a derivative of a laser system discharge voltage with respect to a differential firing time at a constant energy by applying a dither signal to the differential firing time, and a scaled version of the dither signal to a voltage. Applying, adapting the scale to minimize energy error, and taking dV / dt as a scaling factor. Desirable operating points can include operating points that maximize laser system efficiency. The apparatus and method may include adjusting the halogen gas injection size as a function of the difference between the current operating point and the desired operating point. A method and apparatus for controlling the bandwidth of a multi-part laser system including a first line narrowing oscillator laser system part that supplies a line narrowing seed pulse to a second amplifier laser system part, the first laser system Disclose that the selection of the differential firing time between the generation of the seed pulse in the part and the generation of the laser gain medium in the amplifier laser system part results in the bandwidth of the laser output light pulse from the multipart laser system. Adjusts the target operating point as a function of the measurement bandwidth and the target bandwidth, estimates the current operating point, and advances the current operating point and target to advance the current operating point to the target operating point. Adjusting the differential firing time as a function of the operating point, and a function of the current and desired target operating points. It may include the steps of adjusting the halogen gas injection. Estimating the current operating point can include the use of a function that can be easily calculated or estimated from available measurements of laser system operating parameters that are monotonic with respect to differential firing time over the expected operating range. . Estimating the current operating point may include utilizing a difference between a current differential firing time and a reference differential firing time. The method and apparatus estimate a current operating point as a function of at least one of a derivative of a laser system discharge voltage with respect to a differential firing time at a constant energy and a laser system output pulse energy at the current operating point. Steps may be included. The method and apparatus estimate a derivative of a laser system discharge voltage with respect to a differential firing time at a constant energy by applying a dither signal to the differential firing time, and a scaled version of the dither signal to a voltage. Applying, adapting the scale to minimize energy error, and taking dV / dt as a scaling factor. The method and apparatus provides a current operating point as a function of at least one of a derivative of laser system output pulse energy with respect to a differential firing time at a constant voltage and a laser system output pulse energy at the current operating point. An estimation step can be included. The time derivative of the laser system operating energy can be estimated by applying a dither signal to the differential firing time and taking the respective ratio of the dither correlation with the laser output pulse energy and the actual differential firing time. . The time derivative of the laser system operating energy is estimated by applying a dither signal to the differential firing time and calculating dE / dt using the dither, laser output pulse energy, and actual differential firing time. You can also. The derivative of laser system operating energy with respect to time can also be determined from the product of dE / dV and dV / dt. Desirable operating points can include those that maximize laser system efficiency.

  An embodiment of the present invention as used with one or more of the assignee's assignee's laser systems, now named XLA-XXX (1XX, 2XX, 3XX, etc.) series laser systems. In accordance with the theory of operation of an energy timing controller that can be used with aspects, the present application, for example, together with data flow and appropriate algorithmic timing, as well as the algorithm design itself, eg, an energy timing controller design, and an energy timing controller, for example, is an overall multi-chamber laser. Describe system level requirements related to how to fit the system.

  Includes the following acronyms: Analog to Digital Converter (ADC), Launch Control Circuit (FCC), Launch Control Processor (FCP), Internet Service Routine (ISR), Master Oscillator (MO), eg Line Narrowing Laser oscillator in main oscillation power amplifier (MOPA) or main oscillation power generator (MOPO) system, (MSD), photodetector module (PDM), resonance charger (RC), signal processing group (SPG) And a timing and energy control module (TEM) may be utilized in this application.

  The purpose of the energy and timing controller is to ensure that the laser is 0.3% for 30 laser system output light pulses in a pulse burst from certain operating parameter requirements, eg between 1500 Hz and 1875 Hz pulse rate operation, and 1875 Hz. To help meet 0.3% energy / dose stability over a period of about 16 ms of pulse repetition rate operation from to 4000 Hz. Such dose stability requirements vary, for example, from the assignee of the prior art to a one-chamber laser system, such as an ELS-6XXX or 7XXX laser system, ie, a 5 mJ laser to a 10 mJ to 15 mJ pulse system. It represents the required accuracy in the overall output over some defined time window and / or number of pulse definitions as an extension, and the selected minimum number of pulses in the window required to meet any dose stability requirements. The window size that guarantees the dose specification can be constant, for example, 30 pulses from 1500 Hz to 1875 Hz, and then can be increased linearly, for example, up to 64 pulses at 4 kHz, for example, (A) implies that a trapezoidal dose window can be used and may require 99.7% compliance of the worst overall window (maximum and minimum) over 1000 bursts .

  At the same time, such a control system must meet certain timing requirements in order to operate properly, for example when receiving a trigger signal from an integrated circuit lithography scanner, for example when controlling an external trigger signal. That is, for example, “the time from the external trigger to“ Light Out ”” between about 100 μs and 150 μs, the specified short-term jitter is about ± 0.1 msec, for example, jitter fluctuation between lasers of about ± 0.1 msec, and about 1 msec, long-term jitter fluctuation, For example, for a one-month drift, about ± 10 nsec to about 50 nsec, for example, from “Light Out” to “Sync Out”, about −500 ns to about 100 ns, for example, short-term drift / jitter about ± 10 nsec, long-term drift, for example, chamber use It is about ± 10 nsec over the lifetime. In addition, for certain other timing requirements, for example, a launch control panel (FCP), for example, (1) an energy feedforward calculation window (a command to the resonant charger (RC) from the time of pulse reception to the launch control circuit) (Time to write to (FCC)), about 20 μs, (2) (CSC) Feedforward calculation window (time from pulse trigger reception to time to write timing and energy module (TEM)) to FCC) About 75 μs, (3) energy feedback calculation window (time from reception of energy data from TEM to time when RC voltage should be written to FCC), about 200 μs, and (4) CSC feedforward calculation window (received from TEM) T from when the timing data is read into the FCC Time trigger delay for M until to be written to the FCC), it is believed to require to satisfy about 180 [mu] s.

  An exemplary system architecture is shown in FIG. 1 illustrating an exemplary laser timing and energy control system 20 that can illustrate, for example, the relationship between the energy and timing controller 22 and other elements of the laser timing and energy control system 20. As shown in FIG. The energy and timing controller 22 can include software residing on a launch control processor (FCP) that the laser timing and energy control system 20 can be part of. The energy and timing controller 22 can communicate with the rest of the subsystems in the timing and energy control system 20 through a fire control circuit (FCC) 24. The FCC 24 can communicate with a timing and energy module (TEM) 26, a resonant charger (RC) 28, and a customer interface 32.

  The FCC 24 can, for example, send MO and PA delay commands from the energy and timing control system 20 to the TEM 26 over the communication line 44. The TEM 26, for example, further sends MO and PA commutator triggers to the pulse power system 30 over the communication line 36 and is charged, for example, through a semiconductor switching element (not shown) in the semiconductor pulse power module (SSPPM) 30. Can be started. Each trigger generates a gas discharge that is obtained for electrical energy supplied to a respective pair of electrodes through a laser gas medium between the electrodes in each of the respective MOs and PAs. This can be, for example, after a relatively identical time delay that electrical energy traverses the pulse compression and deformation elements (not shown) in the SSPPM 30 for each respective MO and PA.

Using the MO and PA delay commands on line 44, how long a reference trigger on TEM 26, eg, trigger 34 from customer interface 32, should trigger each trigger on line 36, eg, MO and PA-SSPPM 30, respectively. Can show. The pulsed power system 30 (eg, there can be one for each of MO and PA) can transmit the MO and PA-V _cp waveforms to TEM 26 on line 38. Such a waveform is used by the TEM 26 to, for example, when the voltage on the peak capacitor V _cp connected along the electrodes in each one of the MO and PA crosses zero, and thus one for each MO and PA. MO firing time and PA firing time indicating the respective discharge start times through the laser gas medium in one can be calculated. The TEM 26 can measure, for example, the respective zero crossings of the MO and PA-V _cp waveforms relative to the reference trigger and is collected from, for example, one or more “photodiode modules (PDM)” 50. For example, this information along with the energy data can be communicated to the FCC on line 42. The FCC receives the voltage command received from the energy and timing controller on line 52 on line 40 as “resonant chargers” 28 (two RCs 28, for example, one for MO and one for PA. That is, there can be one for each pulse power system module 30). The FCC can calculate the pulse interval (the time between successive triggers on line 36) and supply it to the energy and timing controller 22 on line 52.

The sequence of steps can be performed by an energy and timing controller 22, such as that shown in FIG. The energy and timing controller can, for example, determine whether certain events, eg, two events, (1) a trigger pulse arrives from the customer interface 32 through the FCC 24, or (2) the energy mode or target voltage / energy is changed. Can be activated in response to any occurrence of As shown in FIG. 2, for example, there is a firing sequence of the firing control circuit 24. Processing can begin with receipt of a trigger pulse on line 34, for example. The interrupt service routine shown in box 60 can process the trigger pulse 34 and retrieve the time from the last trigger pulse that can be stored in memory in the FCC, and with this information the resonant charger The resonant charger device manager 62 within 28 can be invoked. The device manager 62 can look up the voltage from the voltage lookup table using the period between triggers. The voltage lookup table can be generated by the feedforward portion of the algorithm. The voltage contribution from the energy servo, energy target, and energy dither can then be added to this value in the FCC so that, for example, the command voltage V _c is lined in the form of the voltage command of the resonant charger 28. it can be transmitted to the resonant charger 28 at 40, to charge the charging capacitor in a single SSPPM each SSPPM to achieve selective V across and electrode over the peak capacitor C _cp. If there are two RCs 28, two high voltage commands can be calculated and transmitted, one for MO and one for PA.

The FCC 24 allows the resonant charger 28 until after the trigger pulse 34 is received in order to allow the voltage command to depend, for example, on the timing of the arrival of the trigger pulse 34, eg, a factor related to the burst interval or pulse repetition rate. It may be necessary to postpone the transmission of the voltage signal to. Such spacing and / or ratio may affect, for example, the temperature within the SSPPM, and thus the selected peak capacitor voltage, and, for example, the charging voltage required to achieve the desired laser output energy. This entire sequence can be performed to ensure that, for example, the 20 μs deadline of the transfer of the voltage command to the resonant charger is met to control the charging of C ₀ at the interrupt level.

  After the voltage command is communicated, the communicated value can be used to apply the correction to the MO and PA trigger delays, which can then be communicated to the TEM 26, and then after the TEM 26 ISR60 can exit. After the laser fires, the TEM 26 can send new energy and timing data 42 to the FCC 24, enable the energy and timing task 80, and the energy and timing SPG 82 can be unblocked. Task 80 may invoke SPG 82, which may read this data from, for example, memory 84 and then communicate relevant shot data 92 to shot and decision SPG 90. The SPG 82 uses this energy and timing data on line 42 to calculate an error that can be used, for example, to adjust the internal state of the system to reduce errors during successive shots. it can. The new voltage lookup table 66, energy servo command, energy target, and energy dither can be calculated and sent to the resonant charger device manager 64, for example, with energy and timing SPG. Finally, new values that calculate the MO and PA delays as a function of voltage can be sent to the TEM device manager 70, eg, on line 101.

  The sequence of events related to energy / voltage target or energy mode change includes (1) receiving a change command from a laser control processor (LCP) by a launch control manager (not shown) in the FCC 24, (2) changing the energy and Sending to timing SPG 82 may be included. The SPG 82 can then call the RC device manager 64, for example on line 100, to change the appropriate values (energy target, energy mode, voltage command).

  The energy and timing controller 22 can perform its logic using a selection algorithm. The algorithm can employ the concept of stratification so that control can be applied to the system on a layer-by-layer basis, for example. Each continuous control layer can be designed to correct problems that were not fixed by the previous layer without reducing the effects of the previous layer.

  At the highest control level, for example, there can be two control layers, such as timing control and energy control. Since timing control can significantly and directly affect the efficiency of the laser, the timing control loop can be designated as the primary control loop and closed first. Thus, the energy loop can be designated as a secondary control loop and closed on the timing group. For example, various functions can be incorporated into the timing controller to prevent the effects of voltage changes in order to prevent feedback interaction between the energy loop and the timing group.

The timing control loop has two purposes: adjusting the relative time between MO and PA launches by first adjusting the time between, eg, V _cp crossing, t _VcpPA -t _VcpMO , It can be designed with the adjustment of the time between the TEM reference trigger and the PAV _cp zero crossing in mind.

By keeping the time between MO launch and PA launch close to the optimal target, for example, the efficiency of the laser can be maximized. Reducing the jitter at this time can reduce energy fluctuations and thus improve dose control. Because V _cp zero crossings tend to be less sensitive to signal amplification variations, for example, V _cp zero crossings can be used instead of fast photodiode threshold crossings.

In general, the target keeps the difference between the MOV _cp intersection and the PAV _cp intersection within, for example, nanometer seconds of the specified target (typically, for example, 30 ns to 40 ns), for example, 30 ns ± 2 ns to 3 ns can do.

  The laser synchronization signal can be fired after a certain duration of the TEM 26 reference trigger, and the TEM 26 reference trigger can be fired after a certain duration of the customer trigger 34. Thus, adjusting the time from the reference trigger to the PA firing can help the laser system meet the trigger and “light” and “light out” to “sync out” requirements.

The timing controller has four layers: (1) a primary layer that corrects the effect of voltage, (2) a secondary layer that adjusts the voltage correction of the effect of temperature, and (3) corrects the variation between pulses 3. Next layer, and (4) dtMOPA (eg, t + differential MO and within PA firing time, eg, t _VcpPA −t _VcpMO , ie, difference in V _cp zero crossing) target is adjusted to optimally correct drift It can be arranged in the fourth layer.

The source of variation of the MO trigger or PA trigger 36 with respect to the V _cp delay is, for example, that charge is transferred from C ₀ to C _p , ie from charging capacitor to peak capacitor, through a saturable magnetic element and step-up transformer that reduce the pulse length. For example, the delay time of the MO commutator, the PA commutator and the compression head during the transfer. These delays may depend at least approximately on the applied voltage applied from RC 28 to charging capacitor C ₀ at the input to the respective MO-SSPPM and PA-SSPPM commutator circuitry and compression head circuitry. Furthermore, since these electrical circuits cannot be perfectly adapted, the difference between the two electrical circuits also varies with the voltage (differential 20 ns). The primary layer of timing control compensates for this effect by using, for example, measurement delay: curve fitting to charge voltage. It has been found that the following types of functions (where Δt is of SSPPM) provide a very good fit to the MO delay curve and the PA delay curve.

（式１）

(Formula 1)

It has been found that curve fitting with Δt _SSPPM = α / V + B + γ / V ² is actually not a bit good, and from “trigger out” to “light out” is about 6 at 300 volts. It varies from .4 μs to 5.1 μs at 1200 volts, and the PA tends to be about 50 ns faster at each point along the curve, but in practice it varies from about 66 ns at 800 volts to about 850 volts. Sometimes it peaks at about 68ns and then drops to about 42ns at 1200 volts. The primary layer of the timing controller can therefore be a simple algebraic correction for the effects of these bolts.

  FIG. 3 shows a schematic diagram of an embodiment of a primary timing controller 110 that may be in the FCC 24. The primary timing controller 110 takes over the voltage command on line 52 from the energy and timing controller 22 and passes the MO delay estimation block 112 and the PA delay estimation block 114, where the MO and PA delay estimation blocks 112, 114 are 1 can be performed, for example, to obtain MO and PA commutator delay estimates, which are then used by the TEM 26 as shown schematically in FIG. MO and PA delay command estimates 44 can be calculated.

  The secondary layer 120 of the timing controller 22 can correct the estimated delay values of the MO and PA with respect to the influence of temperature, for example. When the laser system is activated, the SSPPM of the MO and PA is raised. Thereby, for example, a commutator that typically delays the time from trigger to “light out” when the temperature rises and may reduce the delay and affect the time delay when the voltage still rises For example, the electrical characteristics can be changed. Thus, for example, the MO and PA delay estimates may be inaccurate when the magnetic circuit in the SSPPM commutator is heated.

However, the delay curves are all very similar in shape and in fact all become a single curve when the x- and y-axes are properly _offset , so the following form of equation over _{ΔT SSPPM} over the operating temperature range _: It can be used as a good estimate of the delay, where δ (T) is a temperature dependent operating point.

（式２）

(Formula 2)

For example, if the temperature dependent SSPPM voltage operating point δ is available in real time, for example, the delay estimate can track the temperature change. This tracking can be achieved by measuring the difference between the measured delay and the predicted delay and then adjusting the voltage trip point to reduce this error during every shot.
MoOpPoint [k] = MoOPoint [k-1] + MoOPointServGain ^* MoDelayError [k-1] (Equation 3)
And PaOpPoint [k] = PaOpPoint [k-1] + PaOpPointServGain ^* PaDelayError [k-1] (Formula 4)

  Here, MoOPPoint and PaOPPoint replace δ in Equation 2. As shown in FIG. 3, the step of adding the secondary layer 120 to the primary layer 110 may include an MO-Op servo 122 and a PA-OP servo 124. The MO-Op servo can determine the MO-OP point and the MO delay error [with respect to shot k-1] xMO-OP servo with respect to shot k-1 by performing the calculation according to Equation 3. Similarly, PA- The Op servo 124 can determine the PA-Op point [related to the shot k]. The MO delay error and the PA delay error can be determined by adding in the respective adders 130 and 132. The adders 130 and 132 respectively add the MO delay estimated value to the MO delay command, and The reference trigger for the signal and the reference trigger for the PA can be subtracted.

  A useful effect of adding the secondary layer control 120 to the primary layer 110 in the energy and timing controller 22 is that the same set of coefficients can be used for the MO and PA delay estimates. Applicants therefore believe that using the same coefficient for different commutators results in a MOPA timing error of less than about 5 ns for voltage changes as high as 50V.

The secondary layer can be used to cope with slow temperature fluctuations. Also, for example, it may be desirable to have an additional control layer to handle variations between pulses. In the tertiary control layer, two more servos can be added to this design: MOPAServo 142 and TltServo 144, and three adders 150, 152, and 154. The purpose of these servos 142, 144 is to reduce dtMOPA and “trigger to light” timing errors as quickly as possible. The MOPAServo 142 can serve to reduce the relative time between the MO and PA V _cp zero crossings and the target.
mopaError = dtMOPATarget (refTrigToPaVcp−refTrigToMoVcp) (Formula 5)
mopaAdjust [k] = mopaAdjust [k−1] + gain ^* mopaError [k−1] (Equation 6)

  Here, refTriggerPAVcp and refTriggerMOVVcp display the “light out” time from the reference trigger from each MO and PA, for example, using the respective MOVcp and PAvcp zero crossings.

  In order to reduce large errors as quickly as possible, the gain can be switched to a higher value for larger errors, for example, if abs (mopaError)> ETmopaServoThreshold, gain = ETMOPAServoGainHigh; Gain = ETMOPAssevoGainLow.

The TTL servo can serve to keep the time between the reference trigger and the PA discharge as close as possible to the target. For example, it is as follows.
ttlError = refTrigToLightTargetRefTrigToPaVcp (Equation 7)
ttlAdjust [k] = ttlAdjust [k−1] + ETTtlServoGainLow ^* ttlError [k−1] (Equation 8)

  Here, “refTriggerLight Target” is the time between PA “light out” and the reference trigger, and refTriMoVcp is “light out” from the reference trigger for each MO and PA using, for example, MOVcp zero crossing. It shows the time until.

  The optimal dtMOPA target can drift when conditions change, and the control system can respond to this by, for example, continuously estimating optimal timing values. The optimum timing value can be calculated in the optimum timing value unit 160 in the controller 22 in two stages, for example. Initially, the MOPA gradient estimator 162 can be used to estimate the sensitivity of the laser output energy to the MOPA timing difference. The estimator 160 can output a timing dither value that can be subtracted from the Mo delay command. The dither signal can be a sine wave, and the period, the interval between periods, i.e. the interval between dithers, holdoff, i.e. the number of shots in the burst before the dither starts, and the amplitude are configurable. Yes, that is, a fixed value can be stored in software (also can be updated from time to time). The estimator 162 can maintain two state variables, eg, inputCorrelation and outputCorrelation. When the dither period is complete, inputCorrelation and outputCorrelation can be updated in the following manner.

（式９）

(Formula 9)

（式１０）

(Formula 10)

  Here, g can be a settable gain, and mopa [ki], energy [ki], and dither [ki] are respectively the i-th shot MOPA difference timing (dtMOPa). I.e., t), shutter energy measurements, and dither. The value N can be a dither period.

The MOPA slope can be calculated as the ratio of the two correlations.
mapa slope = outputCorrelation / inputCorrelation expression 10a

The MOPA target servo 164 can serve to advance the mopa gradient to some selected value, ie, zero. The dtMOPA timing servo 142 can operate whenever the MOPA gradient estimator updates, for example, according to:
DtMOPATarget [k] = DtMOPATarget [k−1] + mopa slope [k] ^* ETMOPATargetServGainLow
(Formula 11)

MOPA time difference target, i.e., t ₀ (dtMOPATarget 166) may form the state variables in FCP that can be sustained over power cycles. For example, when the laser system becomes spare, the current value of the MOPA target can be written to a data registry (not shown?). When the laser is turned on again, this MOPA target may be available for reading from the registry.

  As schematically shown in block diagram form in FIG. 4, the energy controller 200 within the timing controller 22 keeps the single purpose, eg, laser output system output light beam energy as close as possible to the specified target value, ie, The energy level within each pulse of the laser system output light beam pulse provided by the laser system can be configured to maintain the pulse repetition rate of laser system operation. In most cases, this target can be implemented, for example, by minimizing moving average energy, ie dose. However, depending on the mode, it can be further emphasized, for example, that the moving standard deviation, ie sigma, is minimized.

  As illustrated in FIG. 4, the energy controller 200 includes, for example, four layers, (1) a primary layer that can respond to changes in the energy target, (2) a secondary layer that corrects drift and energy transfer, (3) It can be placed in a third layer that adjusts the variation between pulses, and (4) a fourth layer that estimates the derivative of energy with respect to voltage.

  The energy control system 200 has several energy control modes, eg, (2) output energy control: sigma, internal sigma ODC, internal dose, internal dose ODC, external sigma, external sigma ODC, external dose, external dose ODC, external Voltage or constant voltage (LEELEnable = true), (2) Dose control mode: internal dose, internal dose ODC, external dose, external dose ODC, (3) Sigma control mode: internal sigma, internal sigma ODC, external sigma, external sigma ODC, (4) Internal energy mode: internal dose, internal dose ODC, internal sigma, internal sigma ODC, external voltage or constant voltage (LELEEnable = true), (5) External energy mode: external dose, external dose ODC, external sigma External sigma ODC, (6) ODC mode: internal dose ODC, external Quantity ODC, internal sigma ODC, external sigma ODC, (7) non-ODC mode: internal dose, external dose, internal sigma, external sigma, (8) voltage control: constant high voltage or external voltage (LELEEnable = false), (9 ) MO control: The MO energy control mode can be supported. Such control modes include (1) for outputting a constant voltage, for example, a fixed voltage (can be set by a settable one), (2) for a constant internal dose, for example, shutter energy (laser system) For adjusting the voltage so that the laser output light pulse beam energy (shutter) at the time of light output is kept constant, for example, the control gain can be set to minimize the movement error of the energy error. (3) Constant internal sigma, for example, for adjusting the voltage so that the shutter energy is kept constant, for example, the control gain is set so that the moving standard deviation of the energy error is emphasized (4) a constant external dose, one that can adjust the voltage to keep the energy at the customer energy sensor constant, eg, control The gain can be set to minimize the moving average of the energy error, (5) constant internal sigma, the voltage can be adjusted to keep the energy at the customer energy sensor constant, For example, the control gain can be set to emphasize the moving standard deviation of the energy error, and (6) the external voltage, eg, the output voltage (eg, integrated circuit manufacturing photo, as required by the customer). For example, between shots can be supported for commanding signals received from lithography scanners.

  The primary layer energy control layer 210 can be arranged to provide energy target agility, for example, an estimate of the inverse of the derivative of the output energy with respect to the voltage (dV / dE) at the signal line 204 in the multiplier 212. The energy target signal 202 can be simply scaled by the value to provide the energy feedforward forward signal 214. Thereby, for example, the control voltage level can be quickly adjusted in line 40 to accommodate changes in the energy target 202. For example, in the constant energy mode, the dV / dE estimate 204 signal and the energy target 202 signal can be the shutter / external energy estimate and target.

  As schematically shown in block diagram form in FIG. 2, the energy control primary layer 210 provides an energy reference value 202a that can be a general value of the energy target 202, eg, an intermediate value of some selected energy range. It can be performed with the energy target 202 subtracted. The result can then be multiplied by dV / dE, for example at amplifier 212, and the result added to default voltage 206 to obtain voltage point 40.

  The secondary energy control layer 220 can accommodate, for example, energy transfer and drift, and can be arranged, for example, as an adaptive feedforward system. The secondary layer 220 can serve to correct DC errors that are inherent in the primary control law where the energy: voltage curve has a non-zero y-axis intercept. The secondary layer 220 can, for example, change the voltage command for the first few pulses in a burst to correct, for example, a burst correlation error. The secondary layer 220 can be applied in a constant output energy mode. In the MO energy control, the feed forward can always be set to zero.

In the secondary layer, the energy target 202 can be subtracted from the measured energy signal 52 in the adder 222 to generate an energy error signal 224, for example, according to the following equation when a new measurement result is available: .
energyerror [k] = measuredenergy [k] energytarget [k] (Equation 12)

This energy error signal 224 scales the dV / dE signal 204 in a multiplier 226 and converts it to an equivalent voltageError signal 228, which is then the voltage applied by the adaptive feedforward circuit 230 using the voltageError signal 228. The waveform can be adjusted to provide an energy feedforward signal 232 to the adder 240, which receives the energyForward signal 214 from the multiplier 212 in the primary layer controller 210. The following formula can be used:
voltageError [k] = dvde [k] ^* energyError [k]
(Formula 13)
F [i, k] = F [i-1, k] + voltage error [k] ^* K _FF
(Formula 14)

Equation 14 can be used, for example, for energy feedforward updates to voltage for a burst based on, eg, errors in the previous burst, where F [i, k] is the kth shot of the i th burst. Occasionally applied energy feedforward voltage, _KFF is the adaptive gain. In special cases, for example, a shot that has just been fired may occur when the ETENergyFF InversionSize is exceeded, in which case the final value may simply continue until a new burst begins.

  A phenomenon that has been noticed to occur in laser systems is the effect of laser firing on efficiency. FIG. 7 shows the general response of a laser system of the type referred to above and firing, for example in voltage mode. The efficiency in this case increases as the laser continues to fire. For example, if the laser output light pulse beam stops firing for a sufficient period of time, the efficiency may return to a value similar to that at the beginning of the burst as shown in FIG. The shorter the stop time, the more likely the laser system will return to some value that is greater than the efficiency at the start of the burst described above.

  The implications of what is illustrated in FIG. 7 by the example is that, for example, the required DC offset for the start of a burst may depend on the length of the previous burst (and the number of repetitions) and the interval between bursts. is there. Although it occurs frequently, the burst length and the number of repetitions tend not to fluctuate much, but the interval between several bursts may be used at different times. For example, a value may occur when changing between dies on a wafer that is exposed to light from a laser system, for example, in a scanner used in integrated circuit manufacturing. Another value may occur between wafers, and a third value may occur between cassettes holding many wafers. Thus, it may be important to make the initial voltage applied during a burst at least approximately dependent on the interval between bursts and / or the duty cycle over a predetermined window size. Also, after the first pulse in a burst has been fired, the system can store the corresponding voltage and add to all subsequent voltages in the burst, for example. Thereby, the DC offset associated with the first pulse correction is effectively added to every pulse in the burst.

  The desired behavior may be that Equation 14 is applied at a fixed burst interval, for example. However, when the burst interval changes, for example, a different set of bins can be used.

（式１４ａ）

(Formula 14a)

  Here, i is a burst number, k is a shot number, and b is, for example, a selectable burst interval bin. Generally, for example, one bin used for burst intervals less than 0.35 seconds, another bin from 3.5 seconds to 4.5 seconds, and a third bin at intervals greater than 4.5 seconds .

  It is thought that the following update logic can be used.

（式１４ｂ）

(Formula 14b)

  Here, ev [i, k] is a voltage error. When b remains constant, equation 14b returns to equation 14 as desired.

Another effect that can be incorporated is the effect of pulse repetition rate on burst transition. FIG. 8 shows energy data averaged over several bursts over the pulse repetition rate range, 290 Hz to 550 Hz, 291 Hz to 1050 Hz, 292 Hz to 1550 Hz, 293 Hz to 2050 Hz, 294 Hz to 2550 Hz, 295 Hz to 3050 Hz, and 296 Hz to 3550 Hz. Show. As can be seen, the transition at the front of the pulse burst changes abruptly as the number of repetitions changes. This behavior is, for example, for different transform waveforms for different repetition rate ranges, eg for the first pulse in a burst.
energyFeedForward = firstPulseOutput = firstPulseWaveform
And subsequent pulses can be addressed by including:

（式１４ｃ）

(Formula 14c)

  In certain cases where the voltage hits the rail, the feed forward can continue to adapt. The state variables in the feedforward algorithm (firstpulsewaveform and waveform) can continue to grow without being constrained by a phenomenon known as servo windup. To limit this, logic can be implemented within the feedforward algorithm. The following logic can be applied after 14b.

（式１４ｄ）

(Formula 14d)

  Here, the operator sat (a, b, c) is a saturation operator.

（式１４ｅ）

(Formula 14e)

The tertiary energy control layer 250 can be used, for example, to account for variations between pulses, which can be in the form of, for example, energy servos 252. The energy servo 252 can be used to correct the voltage in units of pulses, for example. The energy servo can receive the voltageError signal 228 and can use this voltageError signal 228 to provide a correction signal 254 that is directly applied to the applied voltage output signal at the adder 240. 40, and for example, for part of the input to the voltageError signal 228, for example, the EnergyFeedForward signal 232 from the voltageError signal 228 (minus addition) in the input 262 adder 260 to the adaptive feedforward unit 230 By subtracting towards the input to the unit 230 for formation. This can be used, for example, to prevent the energy servo 252 from buffering with the behavior of the adaptive feedforward circuit 230. For example, the error signal 262 used in the adaptive feedforward circuit 230 is an energy servo correction signal. Thus, for example, the energy servo correction signal 254 is not observable by the adaptive feedforward circuit 230 when executing the algorithm. Next, the daptError signal 262 can be obtained by the following equation, for example.
adaptError = VoltageErrorReservationServer
(Formula 15)
The energy servo 252 can employ integral or integral square feedback.

（式１５ａ）

(Formula 15a)

Where K ₁ is the integral gain, K ₂ is the I-square gain, u [k] is the servo output, x [k] is the additional servo state, and e _v [k] is the voltage error. is there.

  For example, in a constant output energy mode, the energy servo 252 can be reset, for example, during the first burst of all bursts. This can be used to ensure that the DC offset at the end of one burst does not affect the voltage applied in the first pulse of the next burst. In MO energy control or if feedforward is disabled, the servo may not be reset during the first pulse of the burst. When the state is reset, both state variables can be set to zero. Similar to feedforward, the energy servo can protect against windup. For example, after applying Equation 15a, the following logic can be applied:

（式１５ｂ）

(Formula 15b)

  Several versions of the energy servo 252 can be installed, for example, three versions. One can be aimed at minimizing dose error (used in internal and external dose control modes) Another is for example energy sigma (in internal and external sigma control modes) The third can be intended for MO energy control, for example. FIG. 5 shows gain optimization for an energy servo for dose control, for example. The plot shows, as an example, the ratio of open and closed loop responses for a Gaussian white noise disturbance (in this case, a good approximation of energy error). Optimal reduction is achieved for all window sizes with a controller gain of 1. The represented pulse window sizes are 280 10 pulse windows, 282 20 pulse windows, 284 30 pulse windows, and 286 40 pulse windows.

  For energy sigma control, the frequency increase can be emphasized. FIG. 6 shows, as an example, the expected closed loop (MSD) value over the expected open loop MSD value for a wide bandwidth disturbance. The optimum gain can be at zero. FIG. 6 shows the ratio of open and closed loop sigma for Gaussian white noise disturbance. From the analysis results, it can be seen that the gain tends to degrade the sigma for any window size. In practice, some gain should be used by the servo to rule out drift, but in general, the energy sigma gain is much lower than the energy dose control.

  For example, as shown in FIG. 4, the fourth layer of energy control can be dV / dE estimation. The purpose of this layer is to determine, for example, the voltage requirement to be applied to the system, for example, to produce a desired energy change at the output (shutter at the output of the entire laser system) or an MO output energy sensor, for example. can do. For a single room controller, e.g. 7XXX, dE / dV is estimated instead. For multi-chamber laser systems, such as MOPA systems, the reverse can be used, for example, to eliminate some divisions. The dV / dE estimate 204 can then be reciprocally written, for example, as dE / dV in the shot record so that customer data is obtained in a manner that the customer is familiar with.

  The estimator can operate as follows. The dither signal 255 can be generated by the estimator 270, and the dither signal 255 can be added directly to the voltage command output, for example. The dither signal 255 can be a sine wave and the period, interval between periods, holdoff, and amplitude can be settable. The estimator 270 can maintain, for example, two state variables, inputCorrelation and outputCorrelation. When the dither cycle is complete, inputCorrelation and outputCorrelation can be updated as follows.

（式２６）

(Formula 26)

（式２７）

(Formula 27)

（式２８）

(Formula 28)

（式２９）

(Formula 29)

  Here, g can be, for example, a settable gain, and energy [ki], voltage [ki], and dither [ki] are i-th previous energy measurements, application Can be voltage, dither. The value N can be a dither period. For values between zero and one, inputCorrelation and outputCorrelation can be, for example, a scaling estimate of the true correlation with the respective energy / dither, the true correlation with the voltage / dither. Equations 28 and 29 can be used with dither signals, eg, shifted by τ units. In general, τ can be selected to be, for example, about ¼ of the dither period. Thus, equations 28 and 29 can be used to give the imaginary part of the correlation. For example, to emphasize that the estimation treats energy as an independent variable and voltage as a dependent variable, reference can be made to the input to energy and the output to voltage.

  The gain estimate can constitute, for example, the real part of the ratio of two complex correlations.

（式３０）

(Formula 30)

  Here, two gain estimators can be used in the control system. For example, one is for output energy and one is for MO energy. The output energy dV / dE estimator 250 can, for example, supply a dither signal to both gain estimators, for example.

According to aspects of embodiments of the present invention, for example, active spectral control on a multichamber laser can be achieved, for example, by adjusting the differential firing time between the two chambers. This can be done using bandwidth control rather than adjusting dtMOPATarget to maximize laser efficiency and determine the value of dtMOPATarget. This can lead to two fundamental problems with, for example, energy and timing control. The first problem may be caused by the F ₂ control algorithm. For example, in the MO-F ₂ injection controller used in the assignee's assignee's XLA-1XX, 3XX, and 3XX series multi-chamber MOPA and similar lasers, MOPA timing (adjusted for optimal efficiency) ) And bandwidth weighting can be used to determine, for example, the timing and size of the MO injection. If the bandwidth is controlled using MOPA timing, both of these values may be unavailable for F ₂ control. To determine the F ₂ injection size, it may require a different fundamental quantities of MOPA-Op point. Another issue would be, for example, timing dither issues that may be required for the “MOPA Op point”. Since lasers may be operating away from the peak of the energy: MOPA timing curve, energy is generally considered to be much more sensitive to differential timing and especially timing dither variations. It is done. Thus, according to aspects of embodiments of the present invention, for example, a method that provides information about the derivative of laser efficiency with respect to timing while simultaneously reducing the effect of timing dither on energy will be needed.

  The purpose of the MOPA-op point selection can be, for example, to provide a measure of “distance” away from the most efficient timing when the laser system is operating, for example. The selection for the MOPA-op point can be selected, for example, to satisfy some basic requirements. That is, the MOPA-op point can be selected to be substantially zero, for example, when the laser output light pulse beam is operating at a differential timing at which maximum efficiency is obtained. The MOPA-op point should be monotonic in differential timing, for example over the expected operating range. The MOPA-op point should be reasonably insensitive to changes in the energy target. The MOPA-op point can be estimated during laser system operation.

The first selection of the MOPA-op point may simply use the difference between the current differential firing time t _mopa and a reference value, eg, the value corresponding to the maximum efficiency, t _ref . The MOPA-op point can simply be as follows:

（式３０ｇ）

(Formula 30g)

It can be seen, for example, that it follows the first three requirements of the MOPA-op point described above, but does not necessarily satisfy the estimability requirement. t _ref can indicate that it fluctuates during operation, eg, due to duty cycle, F ₂ injection, or energy setpoint change. Thus, it may not be directly available for calculation, for example. However, for example, if certain assumptions are made on the relationship between energy, voltage, and timing, an estimate of this MOPA-op point can be derived. A suitable approximation for this relationship can be as follows:

（式３０ｈ）

(Formula 30h)

Where E is the total output energy from, for example, a multi-chamber laser system, i.e. PA or PO, V is the selected operating voltage for one or both chambers, e.g. a voltage, e.g. a MOPA or MOPO configuration, This voltage can simply be the same voltage for timing and other reasons, and is selected, for example, by a resonant charger, eg, by determining the voltage supplied to each respective magnetic pulse power system of MO and PA V ₀ is, for example, a voltage offset, E ′ is, for example, a derivative of energy with respect to voltage at maximum efficiency, and w is, for example, a peak in an energy: timing curve or a voltage: timing curve ( Or other extreme value). Taking the derivative of E with respect to t _mopa gives:

（式３０ｉ）

(Formula 30i)

  Combining these last three equations yields the MOPAOpPoint equation.

（式３０ｊ）

(Formula 30j)

For example, timing dither can be used to estimate the derivative of energy ＭＯE / ∂Vt _mopa with respect to MOPA timing, which can then be made available directly as a measurement result. Since it is believed that some low-pass filtering is required to estimate ∂E / ∂Vt _mopa , Equation 30j can be modified to obtain energy measurements with similar convergence times.

（式３０ｋ）

(Formula 30k)

Here, E _ave is the energy measurement value of the low-pass filtered version. Further, in some embodiments, it may not be possible to estimate the energy derivative with respect to MOPA timing; instead, the voltage derivative with respect to MOPA timing at constant energy is available and used. can do. In this case, the equation can be used with the following modifications.

（式３０ｌ）

(Formula 30l)

  The structure of equations 30k and 30l can illustrate, for example, how to form other MOPA-op point estimates. In both cases, the derivative with respect to the timing can be used to generate a measurement that is, for example, zero when the timing is near maximum efficiency, for example, and is monotonic, for example at the MOPA timing. This measurement can then be reduced, for example, by dividing the energy, for example by reducing the sensitivity to, for example, changes in energy. The constants w and dE / dV actually add functionality to the estimate. It is not an addition and therefore, if removed, these two new equations for the MOPA-op point occur as follows to reflect the execution of the desired MOPA-op point decision.

（式３０ｍ）

(Formula 30m)

（式３０ｎ）

(Formula 30n)

  Equation 30n is implemented in the assignee of the prior art assignee's XLA model laser control system.

According to aspects of embodiments of the present invention, there may be two basic algorithms that can be used to perform ASC. In the first algorithm, the differential timing can be advanced directly, for example using bandwidth control, and the MOPA-op point is used for gas injection control in other cases, for example. Can do. This can be done as follows: (1) Adjust dtMOPATarget as a function of bandwidth error, (2) Estimate MOPAOpPoint in the energy and timing control algorithm, and (3) MOPAOPoint to target value Adjust F ₂ injection to advance to (eg, zero). Another algorithm can control bandwidth using, for example, MOPAOpPoint. This can be performed as follows: (1) adjust the desired MOPAOpPoint as a function of bandwidth error, (2) estimate MOPAOPoint in the energy and timing control algorithm, and (3) bandwidth control From the point of view, adjust dTMOPATarget to advance the MOPAOpPoint toward the desired MOPAOpPoint, and (4) adjust the F ₂ injection to advance the MOPAOpPoint or the desired MOPAOpPoint to the target value. The former algorithm will be easier to implement.

  FIG. 23 shows a somewhat simplified structure of the timing controller as shown in FIG. 3, for example, with a modification that supports the above-described modifications to support ASC. To support ASC, the timing controller 100 1) accepts, for example, a MOPA target supplied by, for example, a bandwidth control algorithm when the MOPA timing mode is set to ASC, and 2) for example, the MOPA timing mode is laser. When set to maximize efficiency, for example, dV / dt estimates from energy and timing algorithms should be used, for example, to adjust the MOPA target. FIG. 14 illustrates, for example, how the algorithm of FIG. 3 can be modified to support ASC. The energy signal 156 is low pass filtered, eg, in a low pass filter, eg, 420 as shown in FIG. 11, and is generated from the energy algorithm, eg, MOPA, to generate eg, a MOPA-op point, eg, used by a gas control system. A slope estimator, for example from 164 as shown in FIG. 3, can be combined with the dV / dt estimate. When ASC is enabled, DtMOPATarget will come from the bandwidth control algorithm. When ASC is disabled, the behavior can return to maximum efficiency. The MOPA-op point at which this is done can be used by a MOPA target servo, eg, 164 shown in FIG. 3, to advance, for example, DtMOPATarget to, for example, the most effective value.

Using the energy controller 200, for example, two objectives can be achieved. The first is to estimate dV / dt _mopa , and the second is to reduce the sensitivity of the energy output to timing dither. This can be accomplished, for example, using an LMS adaptive filter, eg, 275, schematically shown by way of example in the block diagram in FIG. FIG. 24 shows a simplified version of the energy control system as shown in detail in FIG. 4 with a modification of the algorithm. An open loop energy error, for example, an approximation of a voltage error minus the action of the energy servo 252 can be used as an error signal to the filter 275, for example. For example, a timing dither signal from, for example, the MOPA gradient estimated value 162 can be used as a reference in the timing algorithm as shown in FIG. The filter 275 can generate a scaled version of the timing dither, and when the scaled version is added to the voltage command, the signal “timing dithercellation” can be used to counteract the effects of the timing dither signal. Is generated. As a secondary effect, dV / dt _mopa can be estimated using the LSM filter 275.

  The dither cancellation voltage can be calculated as follows:

（式４０）

(Formula 40)

Here, V _cancel the offset voltage, t _dither timing dither, dV / dt _MOPA is, for example, estimated value of the voltage derivative with respect to MOPA timing during constant energy. dV / dt _mopa can be updated using the LMS adaptive filter equation. For example, it is as follows.

（式４１）

(Formula 41)

  Where μ is the adaptive gain and ε is the open loop energy error.

The bandwidth and F ₂ control algorithm can be as follows. The bandwidth control algorithm can adjust the MOPA timing, for example, to advance the system to a target bandwidth value, for example. In a laser system as described above, the bandwidth increases with increasing MOPA timing, ie dtMOPA. Thus, the bandwidth control law can take the following form:

（式４２）

(Formula 42)

  Since it is often desirable to prevent the laser system from leaving the maximum efficiency too far away, the use of a MOPA-op point can be a way to limit this. For example, when the MOPA-op point is outside the specified range, the following logic can be used. When MOPAOpPoint is greater than the maximum desired value and when MOPAOPoint is less than the minimum desired value, respectively.

（式４３）

(Formula 43)

The F ₂ control algorithm may need to be used, for example, to adjust the MO injection to advance the MOPA-op point to the target value, for example. The local behavior of the MOPA-op point can be described by the following equation, for example.

（式４４）

(Formula 44)

Here, u is the MOPA-op point, and F ₂ is the amount of F ₂ in the chamber, eg, the MO chamber. The bandwidth control system is thought to adjust t _mopa to keep the bandwidth constant.

（式４５）

(Formula 45)

Using this equation for dt _mopa gives:

（式４６）

(Formula 46)

When the MO chamber is injected, for example, the maximum efficiency MOPA-op point timing may move down, and therefore the MOPA-op point is considered to increase with respect to, for example, F ₂ injection. By design, the MOPA-op point can always be increased with MOPA timing. Increasing the MOPA timing decreases the bandwidth, and increasing F ₂ increases the bandwidth. Combining this with the above formula gives the following:

（式４７）

(Formula 47)

The MOPA-op point can be monotonically increased for MOF ₂ injection, for example, when the timing is adjusted to maintain a constant bandwidth. Therefore, the following logic can be adopted for the F ₂ algorithm: if the MOPA-op point is too high, the F ₂ injection rate is reduced, and if the MOPA-op point is too low, the F ₂ injection is Increase speed. Table 1 illustrates the symbols and terms used in illustrating aspects of embodiments of the present invention.

According to aspects of embodiments of the present invention, the addition of other forms of existing controllers, such as those used in the assignee's assignee's XLA series lasers, including the concept of the enhancement plant 350, can be added to the present invention. For example, as shown in FIG. 13, it can be used according to the aspect of the embodiment. The enhancement plant 350 can include a laser 352, an energy controller 354, and a timing controller 366. The enhancement plant 350 has, for example, four inputs: (1) an energy target E ₀ , (2) a voltage dither signal V _d , (3) a timing dither signal t _d , and (4) a differential MOPA chamber firing time target t _0. Can have.

According to an aspect of an embodiment of the present invention, the voltage dither signal V _d can be used as an additional voltage input introduced by, for example, timing and energy algorithm modifications. This is not the same as a voltage dither signal that is used, for example, to estimate dE / dV and that is not explicitly included in, for example, the description of aspects of the presently described embodiment of the invention. The output of the enhancement plant 350 is (1) energy error E _e , (2) energy: voltage gradient estimate, ∂E / ∂V, (3) open loop voltage error V _e , (4) output energy E, (4 ) Measurement differential MOPA chamber firing time t. The voltage applied to the laser 352 can be, for example, the sum of the voltage dither input V _d and the command voltage V _{c from} the energy controller 358. The energy controller 3584 can generate estimates of E _e , ∂E / ∂V, and V _e , for example. Differential MOPA commutator firing command t _a signal, for example, may be the sum of the output of the timing dither signal t _d and the timing controller t _c. According to aspects of an embodiment of the present invention, the laser system 352, for example, take the firing voltage V _c signal and a differential MOPA commutator firing time t _a signal 384 as an input, for example, measuring energy E and MOPA chamber firing time t can be generated as an output.

According to aspects of embodiments of the present invention, for example, to adjust the time _ttl from trigger to light using a further aspect of timing control, for example, the command mode part of the commutator firing command for MO and PA Can be adjusted. This portion of the controller 350 can be completely separated from, for example, the energy and timing controllers 354, 356 described herein, and is omitted from the figure for clarity.

In accordance with an aspect of an embodiment of the present invention, the timing controller 356, for example, in response to an error between the desired t ₀ and the measured MOPA chamber firing differential time t, for example, the command MOPA commutator firing time t _c Can be adjusted. In its simplest form, the timing servo in timing controller 356 can be a separate integrator.

（式４８）

(Formula 48)

Here, g _t may be constant gain. For example, further enhancements to the timing control currently used in accordance with aspects of embodiments of the present invention in one or more of Applicant's assignee's XLA series laser systems as previously shown by way of example An example can be a delay based on the size of the error: the effect on the commutator by voltage and / or correction of gain scheduling. In accordance with an aspect of an embodiment of the present invention, the timing controller 356 may, for example, provide a desired differential MOPA commutator, for example, for the next pulse k based on an error relative to the preceding pulse k−1 scaled by some gain g _t . Based on the firing command t ₀ and the measured tMOPA chamber differential firing time, t ₀ can be calculated.

In accordance with aspects of an embodiment of the present invention, the energy controller 345 shown schematically in FIG. 8A can employ several levels, for example, three levels of control. In a first level, according to aspects of embodiments of the present invention, for example, energy target E ₀ , signal 358 can be, for example, energy: voltage gradient, as shown schematically and in block diagram form in FIG. 8A. The E ₀ energy scaling amplifier 480 can be scaled by the reciprocal of the estimate ∂E / ∂V, ie, dV / dE. This scaling in amplifier 480 can produce, for example, one component 481 of command voltage V _c 3624. According to the aspect of the embodiment of the present invention, this can be used to reduce the sensitivity of the energy error E _e to, for example, a large change in the energy target E ₀ . In the second level of control, for example, between the energy target E ₀ , the input signal 358 (eg, can be delayed by one sample in the delay unit 486) and the actual measured energy E, the input signal 390 The difference can be filtered in an E ₀ -E signal amplifier 482 to generate a correction signal 483 that can tend to advance the energy error to zero, eg, from the function dV / dE. For example, in the current example of an XLA series controller as described above, this control law can be, for example, integral-integral square feedback. The energy error E _e can be scaled by, for example, ∂E / ∂V before being filtered, for example to keep the loop gain constant. The third level of control can be, for example, feedforward control (not shown) between bursts. The controller 354 can, for example, calculate the inverse waveform dV / dE, which can be calculated, for example, for the first 20 or so shots of a burst to correct for energy transfer. Can be applied. This waveform dV / dE can be adapted on an interburst basis.

In order to obtain a stable adaptation, the feedback algorithm can use, for example, the open loop voltage error estimate V _e , signal 404, along with the effect of the removed energy servo. This is, for example, the signal V _e 404 shown in FIG. A useful characteristic of this V _e signal 404 is the voltage dither, the closed loop transfer function from V _d to V _e being, for example, approximately 1 when the error in the estimate of the energy: voltage gradient ∂E / ∂V is small. Is that you can. ∂E / ∂V adds a four-pulse voltage signal to the laser voltage V _c and, for example, converts this signal to the laser input voltage V to reach an estimate of energy: voltage sensitivity ∂E / ∂V signal 402. It can be estimated by correlating with _c −V _d and the output energy E. This estimator can now be omitted from the description of the energy controller, for example in FIG. 13, for the sake of clarity. The energy controller 354 can calculate the voltage command V _c , signal 404, which can be aimed, for example, to minimize the error between the measured energy E and the desired energy target E _0. .

In the present assignee's assignee's current laser series, XLA-1XX laser system, for example, the MOPA timing target t ₀ (desired MO and PA and Differential discharge timing) can be adjusted. FIG. 16 shows an example of an energy: MOPA timing curve for example at a constant voltage, eg V _c . The energy E can be maximized, for example when t = t _opt , and then can be lowered when leaving the optimum t _opt on either side of the curve from t with varying timing as shown. There is sex. For example, in a typical energy: differential MOPA chamber firing time curve, the voltage can be kept constant, for example when changing the MOPA timing. The maximum energy output can be achieved with the optimal differential chamber firing time (eg, 30 ns) shown by way of example. FIG. 17 shows, by way of example, how the present MOPA target control described above can be performed, for example, in the assignee's assignee's XLA-1XX series multi-chamber laser. The timing dither signal t _d can be generated as an input to the enhancement plant 350.

This dither signal t _d may have a constant amplified sine wave at the Nyquist frequency, for example. In other words, the dither signal t _d may fluctuate between a positive value and a negative value at every other shot, for example. The slope of the energy: timing curve, for example, correlates the timing dither signal with, for example, the differential MOPA chamber firing time t measurement 392 and the measured output energy E signal 390 and takes the ratio ∂E / dt to form the output signal 436. Can be estimated. This operation is shown schematically in FIG. The timing target servo 414 can step, for example, the MOPA timing target t ₀ value 372 “upward” until the estimated MOPA slope ∂E / dt signal 436 is zero according to Equation 36, for example, very simple. A control law can be implemented.

（式４９）

(Formula 49)

FIG. 17 illustratively shows the operation of this controller. For example, in the operation of the present MOPA timing controller, for example, in the above-mentioned MOPA laser system of the assignee of the present applicant, as described above, the timing dither t _d signal 374 is, for example, before being sent to the laser controller 352, for example, It can be applied to the command timing t _c 376 at the current operating point. The dither signal t _d can be applied, for example, at the current operating point, and the controller 352 can serve to move the operating point, for example, to the peak portion of the energy: timing curve. A useful characteristic of this controller 352 is that, for example, when the timing target t ₀ converges (eg, the laser system is operating at the top of the curve), the sensitivity of the output energy E to the timing dither t _d is zero. It can be said that it can be. Therefore, the use of the timing dither t _d can be used without ignoring the stability of the output energy E.

According to aspects of embodiments of the present invention, the system of the present invention can be operated, for example, away from the peak of the MOPA timing curve, for example, to adjust the laser system output, eg, bandwidth. FIG. 10 illustrates aspects of an embodiment of the present invention. For example, instead of operating at the top of the energy: timing curve, the laser can be operated away from the peak. According to aspects of embodiments of the present invention, Applicants have noted that increasing the distance away from the peak can have a strong impact on the spectrum of light emerging from the laser system. Thus, according to aspects of embodiments of the present invention, the bandwidth output of the laser system can be controlled, for example, by adjusting the distance that the laser operates away from the peak t _offset .

FIG. 14 exemplarily and schematically shows a controller designed to do this. The desired offset t _offset from the peak can be calculated by the bandwidth control servo 412. The simple law of this control servo is as follows.

（式５０）

(Formula 50)

Where g is a gain selected to obtain a timely reduction in bandwidth error. The rest of the control servo can be designed, for example, to advance the operating point of the laser above the dE / ddtMOPA operating curve to this desired offset. The MOPA gradient estimator 244 may continue to operate, for example, similar to the applicant's assignee's current XLA control design. In addition, a low pass filter 420 can be applied to the measured energy E signal 390 to obtain, for example, average energy E _ave and output signal 434 measurements. The time constant of this filter can be adapted, for example, to the time constant of filters 462, 472 used to estimate the MOPA slope.

  The offset from the peak can be estimated by making some assumptions on the shape of the dtMOPA timing curve, taking into account the slope and energy value of the curve. A model that is not irrational is to assume a law of the form

（式５１）

(Formula 51)

Where E _p is the peak energy that can be expected to operate at the peak of the MOPA timing curve during launch at the current voltage, and σ is a known parameter that determines the width of the curve. The ratio of gradient to energy can be expressed as:

（式５２）

(Formula 52)

  Accordingly, the estimated offset value from the optimum timing that can be executed by the “optimum MOPA timing estimator” 424 can be expressed as follows, for example.

（式５３）

(Formula 53)

  Using the estimated value Δt of the distance between the current timing operating point and the optimal efficiency timing operating point, the timing target servo 414 of FIG. 14 can take the following form.

（式５４）

(Formula 54)

  This controller can serve to advance the laser to the desired operating point.

One notable drawback of the previous illustrative version of the modified controller 408 is that the effect of timing dither t _d is easily observable in the laser output energy E at this point when the controller 408 converges on the operating point. That's what it means. This shortcoming can be addressed, for example, as shown in an illustrative schematic diagram of another version of the modification controller 408 ′, for example, as shown in FIG. In order to eliminate the effect of timing dither t _d on output energy E, timing dither t _d signal 374 can be used as a reference for adaptive feedforward controller 452. This embodiment of the controller 452 implements an LMS algorithm that uses the estimated open loop voltage error V _e as an error signal that adjusts the voltage dither signal V _d to advance a portion of the energy error correlated in the timing dither to zero. It is thought that it can be implemented.

  To offset the effect of the timing dither signal in the output energy in this manner, a slightly different method of estimating the slope of the energy: timing curve may be required. Recall that the timing dither can be a sine wave at the Nyquist frequency.

（式５５）

(Formula 55)

  The voltage dither required to offset this is given by the following formula:

（式５６）

(Formula 56)

  By which the estimate of the slope of the MOPA timing curve can be given by the ratio of the amplitudes of the two dither signals scaled by the energy: voltage slope estimate.

（式５７）

(Formula 57)

Referring now to FIG. 13, an enhancement plant 350 is shown illustratively and in a block diagram, and the enhancement plant 350 can include a laser system controller 352, an energy controller 354, and a timing controller 356. The laser system controller 352 can have inputs V _d and t _d modified as described below to form input signals 382 and 384, and the measured energy signal E and differential MO and PA as outputs. A firing time signal t can be provided. The energy controller 354 can be provided with an E ₀ input signal 358 and an E signal and an output voltage command V _c signal 362 that can be combined with V _d in the adder 370 to form the input signal 382. Can do. The energy controller 354 can also provide an E _c signal 400, a ∂E / ∂V signal 402, and a V _e signal 404 as outputs.

Timing controller 356 can provide as output t _c that has t ₀ and t as inputs and can be combined with t _d to form input signal 384.

Referring now to FIG. 14, a modified controller 408 according to an aspect of an embodiment of the present invention is shown illustratively and in schematic block diagram form, where the controller 408 includes an enhancement plant 350 and a t ₀ generation circuit. The t ₀ generation circuit can receive a timing dither generator 410 that generates a timing dither signal t _d, and an error between a bandwidth and a desired bandwidth, Δλ _err as an input, t _offset , that is, a desired dtMOPA differential A band that generates time and represents, for example, each point on the graph of dtMOPA change to obtain the desired Δλ, for example, a band that zeroes the Δλ _err difference between the measured bandwidth and the desired bandwidth according to the value of the lookup table And a width control servo 412. Bandwidth control servo 412, also referred to as timing offset servo 412, can supply t _offset signal 432 to timing target servo 414, which receives Δt signal 438 and receives t ₀ according to, for example, Equation 50. A signal 372 can be generated. The timing target servo 414 can receive the Δt signal 438 from the optimal MOPA timing estimator 424, and the optimal MOPA timing estimator 424 receives the ∂E / dt signal 436 from the MOPA slope estimator 422 and the E _ave output signal. 434 can be received from the low pass filter 420, which can low pass filter the measured energy E signal 390 exiting the laser system on the enhancement plant 350. The MOPA gradient estimator 422 receives the t signal 392 from the enhancement plant 350 and the E _e signal 400 from the enhancement plant 350 together with the timing dither signal t _d from the timing dither generator 410 also shown in FIG. 57, the E / dt signal 456 can be calculated. Optimal MOPA timing estimator, the optimal Δt between the MO and the PA so that a predetermined energy is obtained, i.e., the filtering E _ave signal 438 to calculate, for example, the t ₀ 0 based on the desired bandwidth changes Calculate the desired value and supply it to the timing target servo 414 to achieve the selected bandwidth, ie, the t _offset signal 432.

Referring now to FIG. 15, for example, a modified control system according to an aspect of an embodiment of the present invention that can include the elements shown in FIG. 14 with an adaptive feedforward circuit 452 that calculates a voltage dither V _d signal according to Equation 56, for example. It is shown. The adaptive feedforward circuit 452 can provide the ∂V / ∂t output signal 440 calculated according to the timing dither signal t _d to the MOPA slope estimator 422, which then performs adaptive feedforward. The ∂E / dt signal 436 can be calculated by multiplying the ∂V / ∂t output signal 440 from the circuit 452 by the ∂E / ∂V output signal 402 from the enhancement plant 350.

Referring now to FIG. 8A, an input E ₀ signal 358 and an adder to obtain, for example, the value E _e = E−E ₀ where the E ₀ signal 358 is delayed for one pulse to a z ⁻¹ delay 486. An energy controller 354 that can include an input measurement laser system output energy E signal 390 summed within 460 is illustratively shown in schematic block diagram form. The desired laser output energy E ₀ signal 358 can be provided to the energy controller E ₀ dV / dE scaling amplifier 480 to achieve the voltage command V _c component output signal 481. Another command voltage V _c component signal 483 can be provided from the dV / dE scaling amplifier 482 from the energy controller E (k) -E0 (k−1) which provides the output signal 483 to the energy servo 488. Components 481 and 482 can be added in adder 490 to form command voltage V _c signal 362. Further, the output 483a of the energy servo 488 is delayed by, for example, a z ⁻¹ delay 492 and subtracted from the input to the energy servo 488 in the adder 494 to obtain the open loop voltage error V _e signal 404 of the energy controller 354. can do. E _e = E (k) −E _e (k−1) signal 484 may form E _e output 400 from energy controller 354.

Referring now to FIG. 18, a MOPA gradient estimator 422 according to an aspect of an embodiment of the present invention is illustrated in an exemplary and schematic block diagram format, wherein the MOPA gradient estimator 422 includes E _e signals 390 and t _An amplifier 460 can be included that receives the input of the _d signal 392 and provides the input to the low pass filter 462, which can provide the energy portion 464 of the ∂ E / dt signal 436 to the divider 466. . Similarly, amplifier 470 multiplies input td signal 374 and t signal 392 to form an input to low pass filter 472 and provides time portion 474 of ∂E / dt signal 436 to divider 466 for output ∂. An E / dt signal 436 can be formed.

The purpose of the timing control is then to drive the system to the desired offset to achieve the operating point on the operating curve, and thus the desired bandwidth. To do this, the control system 408 can determine when the peak operation is obtained and then calculate, for example, the current value of t _offset . However, since the laser is operating away from the peak, without detailed knowledge of the shape of the energy: timing curve shown in FIG. 16 as an example, it is not possible to determine at what timing peak efficiency actually occurs. It will be difficult. This may be further complicated by the fact that the shape of the curve shown in FIG. 16 may vary depending on, for example, laser operating conditions (eg, duty cycle, pulse repetition rate, output energy level, etc.). However, the offset timing information can actually be used, for example, in a feedback loop around the bandwidth, so it is stable by using a value that is monotonic but more easily measurable in the offset timing, for example. An algorithm can be achieved. One such value can be, for example, the following quantity:

（式５８）

(Formula 58)

This amount u, referred to by the Applicant as the timing operating point, can include the ratio of the derivative of voltage to the average energy relative to the timing at fixed output energy. u can be shown to be monotonic, eg, in timing offset, over a wide range, eg, around the energy: timing peak of FIGS. 16, 17, and 10, for example. Further, for example, u can be shown to be zero when the offset timing t _offset is zero. Therefore, the following form:

（式５９）

(Formula 59)

Bandwidth controller 408 shown as an example in FIG. 11 for executing the control law be made to have the desired ', for example, a measurement operating point u timing controller 408 is to proceed toward the u _0' Where u ₀ is the desired timing operating point. FIG. 11 shows a controller 408 ′ designed to do this. The desired operating point u ₀ can be calculated by bandwidth controller 412 according to equation 59. The rest of the controller 408 can be designed to advance operating parameters, eg, E, to this desired operating point.

The problem with operating away from the peak of the energy: timing curve is that the output energy is much more sensitive to the timing dither signal used to estimate the slope of the energy: timing curve. . This can have a negative impact on laser energy stability. In order to eliminate the effects of timing dither in the output energy, the timing dither signal can be used as a reference for the adaptive feedforward filter 452 of FIGS. The implementation of this adaptive feedforward filter 452 within the controllers 408, 408 ′ of FIGS. 15 and 11 may be implemented, for example, by moving the voltage dither signal V _An algorithm, such as the LMS algorithm, can be performed using the estimated open loop voltage error V _e as the error signal to adjust _d . In addition to generating the desired voltage dither signal, the side effects of this embodiment of the operation of the adaptive feedforward filter 452 also use the adaptive feedforward filter 452 to achieve the constant energy time timing of FIG. The voltage derivative with respect to ∂V / ∂t | _E signal 500 can be estimated. This section and the average output energy E _ave in the operating point calculation block 504 in FIG. 11, taking the ratio of the signal 434 can be obtained the desired measurement result of the timing operating point u in accordance with equation 44.

Next, using the measured operating point and the desired operating point, for example, in the timing target servo 414, the desired differential firing time (t ₀ ) of the MO and PA chambers can be calculated. The servo 414 can implement a simple feedback law of the form

（式６０）

(Formula 60)

The conclusion of the timing control described above using the controller 408 or 408 ′ is that the timing value at which maximum efficiency is obtained is no longer available as a measurement result of other laser subsystems. A system that uses this value can be, for example, an F ₂ infusion control system. Some versions of this system may require using, for example, one or more of the laser chambers, eg, the MO chamber of a two-chamber MOPA laser system, using derivation of the measured optimum timing from the reference optimum timing value. The F ₂ level to be determined can be determined. FIG. 12 shows an example of such an architecture 520. The F ₂ controller 522 can take an optimal firing time t _offset signal 524 as input and use this value to adjust the injection size signal 526. The injection size signal 526 can be used to affect the level of fluorine injected into the laser system chamber, and thus the amount of fluorine contained in the chamber that affects the optimum firing time differential. FIG. 19 shows how the F ₂ control can be modified to accommodate the timing control algorithm as described above.

For example, if a desired operating point for a given desired bandwidth is calculated, for example, the bandwidth controller calculates a desired timing operating point u ₀ . The F ₂ controller 522 can take this target u ₀ as an input. The F ₂ controller 522 can then execute an algorithm that can advance the target operating point u ₀ to a desired value. For example, the operating point may be driven to zero or some other value other than zero, or to some other value that maximizes the laser system operating efficiency, for example, so that the laser is always at the selected point, for example maximum It is guaranteed to work with efficiency or at least in some desired range. Timing control 408, 408 ′ and bandwidth control 412 can both quickly bring the system to the desired bandwidth, and then F ₂ control 522 allows the laser system to operate at a selected point, for example, at maximum efficiency. The gas concentration can be adjusted slowly.

Referring now to FIG. 20, as an example, details of aspects of the embodiment of the present invention shown in FIG. 19 are shown. FIG. 20 shows a system 530 as an example, and the F ₂ injection control system 522 determines and controls, for example, the amount of F ₂ gas injected into the laser 353 during replenishment, while boosting. The plant 350 receives u ₀ and E, tVcpMO and tVcpPA, and activates the operating voltage V and a trigger signal MOTRIG that closes a semiconductor switch that determines the time of gas discharge between electrodes in the MO chamber, for example, a PA chamber. A trigger signal PATRIG that determines the closing of the semiconductor switch that determines the time of gas discharge between the electrodes in the inside and the actual BW are also generated.

The energy timing and control system, according to aspects of embodiments of the invention, can serve to advance u to equal u ₀ the shorter the time period, the longer the system is depleted of fluorine. In doing so, it is possible to move u ₀ over the gas lifetime, for example, to change the operating point to maintain the desired BW, or to move the BW target in the range of acceptable bandwidth values in gas lifetime units.

  A first line narrowing oscillator laser system portion that supplies a line narrowing seed pulse to a second gain generator laser system portion, wherein seed pulse generation in the first laser system portion and laser in the second laser system A method and apparatus for controlling the bandwidth in a multi-part laser system, where the choice of differential firing time during gain medium generation affects the bandwidth of the laser output light pulses from the multi-part laser system, Disclosed in the present application in accordance with aspects of an embodiment of the invention, which includes adjusting a differential firing time as a function of a measurement bandwidth and a bandwidth target, estimating a current operating point, and a current operation It will be appreciated by those skilled in the art that adjusting the halogen gas injection as a function of the point and the desired operating point can be included. For example, the differential firing time between the MO and PA in a multi-part laser system is two times within the tolerances of at least an available power source and control equipment that would be understood by those skilled in the art, for example within a few nanoseconds. A switch in the power supply for the MO, for example, to control the timing of the discharge between the electrodes of the MO laser part of the laser system and then the PA (or PO) part of the laser system with a selected timing difference between the discharges, or For example, it can be adjusted by the use of the energy and the timing signal generated by the timing controller leading to the closing of each of the other switches in the power supply for the power supply for PA. The measurement bandwidth can be measured, for example, according to known techniques for such measurements as shown in the various assigned patents of the assignee of the applicant referred to above and the current patent application pending. The bandwidth target is, for example, in response to the required bandwidth from equipment utilizing laser output light pulses generated by a laser, such as an integrated circuit photolithography apparatus as known in the art, for example of a laser system A selected bandwidth selected by the controller can be included. This difference may form a bandwidth error. The step of estimating the current operating point involves the use of a function that can be easily calculated or estimated from available measurements of the laser system operating parameters and that is monotonic with respect to the differential firing time over the expected operating range. For example, the operating point can be, for example, the operating point of a variable operating parameter, for example a constant energy or a change in the differential firing time as a function of the operating voltage at a constant operating voltage, eg at a constant operating voltage The operating point may be on a curve, for example, and may be determined empirically for a given laser system and / or individual laser device and may be updated periodically, eg according to measured laser system operating parameters it can. Halogen gas injection can be controlled as is known in the art, and the current operating point can be moved to the desired operating point by such injection, according to aspects of embodiments of the present invention. .

The desired operating point can be determined as a function of at least one of target bandwidth, laser system duty cycle, and laser system output pulse energy. Estimating the current operating point includes utilizing the difference between the current differential firing time and the reference differential firing time. The method and apparatus can further include adjusting the differential firing time as a function of bandwidth error. The method and apparatus can include, for example, selecting a reference differential timing to optimize laser system efficiency at the extremes of the operating voltage: differential timing curve or the operating energy: differential timing curve. The method and apparatus provides a difference between the current differential firing time and the reference differential firing time, for example, a derivative of the laser system output pulse energy with respect to the differential firing time at a constant voltage, and for example at the current operating point. Estimating as a function of at least one of the following laser system output pulse energies. The method and apparatus, for example, provides a derivative of the laser system discharge voltage with respect to the differential firing time at a constant energy, a derivative of the laser system output pulse energy with respect to the laser system discharge voltage, and a laser system output at the current operating point. Estimating the difference between the current differential firing time and the reference differential firing time as a function of at least one of the pulse energies. The method and apparatus estimate a current operating point as a function of at least one of a derivative of a laser system discharge voltage with respect to a differential firing time at a constant energy and a laser system output pulse energy at the current operating point. Steps may be included. The method and apparatus can include estimating a current operating point as (1 / E) ^* dV / dt, where E is the laser system output pulse energy and dV / dt is, for example, current For example, the differential of the laser system discharge voltage with respect to the differential firing time at constant energy at the operating point.

The method and apparatus can be used, for example, as a function of at least one of a derivative of the laser system output pulse energy with respect to a differential firing time at a constant voltage and a laser system output pulse energy, eg, at the current operating point. Estimating the operating point can be included. The method and apparatus can include estimating a current operating point that includes utilizing the relationship (1 / E) ^* dE / dt, where E is the laser system output pulse energy; dE / dt is the derivative of the laser system output pulse energy with respect to the differential firing time at a constant voltage at the current operating point. dE / dt can be estimated by applying a dither signal to the differential firing time and calculating dE / dt using the dither, laser output pulse energy, and actual differential firing time. dE / dt can be estimated by applying the dither signal to the differential firing time and taking the respective ratios of the dither correlation with the laser output pulse energy and the actual differential firing time. dE / dt can also be determined from the product of dE / dV and dV / dt.

  The method and apparatus estimate a derivative of a laser system discharge voltage with respect to a differential firing time at a constant energy by applying a dither signal to the differential firing time, and a scaled version of the dither signal to a voltage. Applying, adapting the scale to minimize energy error, and taking dV / dt as a scaling factor. Desirable operating points according to such embodiments may include operating points that maximize laser system efficiency. The method and apparatus may include adjusting the halogen gas injection size as a function of the difference between the current operating point and the desired operating point.

A first line narrowing oscillator laser system portion that supplies a line narrowing seed pulse to a second gain generator laser system portion, the seed pulse generation in the first laser system portion and the laser gain medium in the amplifier laser system Disclosed is a method and apparatus for controlling bandwidth in a multi-part laser system, wherein selection of a differential firing time during generation of the power affects the bandwidth of laser output light pulses from the multi-part laser system. According to an aspect of an embodiment of the present invention, adjusting the target operating point as a function of the measurement bandwidth and the target bandwidth, estimating the current operating point, and setting the current operating point to the target operating Adjusting the differential firing time as a function of the current and target operating points to advance to the point, and the current target operating point and the desired It may include the steps of adjusting the halogen gas injection as a function of a point. For example, the target operating point can be set such that the control system corresponds to, for example, bandwidth control, while the desired operating point is, for example, to advance the system operating point where the system also maintains bandwidth control. It will be appreciated by those skilled in the art that, for example, the system can set the operating point at which F ₂ injection control can be utilized.

  Estimating the current operating point according to such an embodiment, as described above, is monotonic for differential firing time over the expected operating range and is easily calculated from available measurements of laser system operating parameters. The use of functions that are possible or estimable can be included. Estimating the current operating point can include utilizing the difference between the current differential firing time and the reference differential firing time, as described above. The method and apparatus, as described above, is current as a function of at least one of the derivative of the laser system discharge voltage with respect to the differential firing time at a constant energy and the laser system output pulse energy at the current operating point. Estimating the operating point can be included. The method and apparatus includes estimating a laser system discharge voltage derivative with respect to a differential firing time at a constant energy by applying a dither signal to the differential firing time, as described above, and scaling the dither signal. Applying a scaled version to the voltage, adapting the scale to minimize energy error, and taking dV / dt as a scaling factor. The method and apparatus, as described above, are currently present as a function of at least one of the derivative of the laser system output pulse energy with respect to the differential firing time at a constant voltage and the laser system output pulse energy at the current operating point. Estimating the operating point of the. The time derivative of the laser system operating energy is obtained by applying a dither signal to the differential firing time and taking the respective ratio of the dither correlation with the laser output pulse energy and the actual differential firing time, as described above. Can be estimated. The derivative of the laser system operating energy with respect to time, as described above, applies a dither signal to the differential firing time and calculates the dE / dt using the dither, laser output pulse energy, and actual differential firing time. Can be estimated. The derivative of the laser system operating energy with respect to time can also be determined from the product of dE / dV and dV / dt, as described above. Desirable operating points can include operating points that maximize laser system efficiency, as described above.

  Aspects of the embodiments of the invention disclosed above are intended to be preferred embodiments only and are not intended to limit the disclosure of the invention in any way, in particular only certain preferred embodiments. It will be understood by those skilled in the art that the present invention is not intended to be limiting. Many changes and modifications may be made to the disclosed aspects of embodiments of the disclosed invention as will be understood and appreciated by those skilled in the art. The claims are intended to encompass, within its scope and meaning, not only the disclosed aspects of the embodiments of the present invention, but also equivalents and other modifications and changes that would be apparent to a person skilled in the art. In addition to the changes and modifications to the disclosed and claimed aspects of the embodiments of the invention described above, it is contemplated that others can be implemented.

  Specific aspects of the “MOPA gas discharge laser bandwidth control by discharge timing” embodiment described and illustrated in this patent application in the details required to satisfy “35 USC §112” are: Any of the above-mentioned objects of the above-mentioned embodiment aspects, and the problems to be solved by or for any other reason of the above-mentioned embodiments, or for that purpose, can be completely achieved, It should be understood by those skilled in the art that the aspects described herein of the above-described embodiments of the present invention illustrate and represent the broad subject matter of the present invention. The scope of the presently described and claimed aspects of the embodiments encompasses all other embodiments that are presently believed or will be apparent to those of ordinary skill in the art based on the teachings herein. Is. The scope of “MOPA gas discharge laser bandwidth control by discharge timing” of the present invention is solely and completely limited only by the claims, and nothing exceeds the detailed description of the claims. References to elements in such claims in the singular are intended to mean that such elements are "one and only one" unless explicitly stated in the interpretation. Not intended and meaningless, and intended and meant to mean "one or more". All structural and functional equivalents of any of the above-described aspects of embodiments known to those skilled in the art or later known are expressly incorporated herein by reference and are It is intended to be covered by a range. Any term used in the specification and / or claims of this application and expressly given meaning to this specification and / or claims of this application shall have any dictionary meaning for such terms. Or shall have its meaning regardless of other commonly used meanings. Each or all of the apparatus or methods described herein as any aspect of an embodiment is sought to be solved by the aspects of the embodiment disclosed in this application, as it is encompassed by the claims. It is not intended or necessary to deal with the problem. Any element, component, or method step in the disclosure of the present invention will be disclosed to the general public regardless of whether the element, component, or method step is explicitly described in detail in the claims. It is not intended to be dedicated. Any claim element in a claim is either explicitly recited using the phrase “means for” or the element is “action” in the case of a method claim. However, unless it is listed as “stage”, it shall not be construed in accordance with the provisions of paragraph 6 of “35 USC § 112”.

1 is a schematic block diagram of an exemplary system architecture useful for multichamber laser system timing and energy control. FIG. FIG. 5 is a diagram showing a modified version of a part of the controller of FIG. 4. FIG. 3 is a schematic block diagram illustrating an example of a timing controller. FIG. 2 is a schematic block diagram illustrating an example of an energy controller. FIG. 6 illustrates gain optimization for energy servos according to aspects of an embodiment of the present invention. It is a figure which similarly shows gain optimization. It is a figure which shows the energy characteristic of laser system discharge | release in a constant voltage mode. It is a figure which shows laser system operating energy data. FIG. 5 is a diagram showing a modification of a part of the controller shown in FIG. 4. FIG. 4 is a diagram showing a modification of a part of the controller shown in FIG. 3. An example where the controller advances the operation of the laser system to the point of the operating curve where the desired bandwidth output is obtained, for example by advancing the offset from the optimal operating point (eg at the maximum energy output) to the desired point on the operating curve It is a figure similar to FIG. 17 which shows a typical control system. FIG. 3 is a schematic block diagram illustrating a multi-room timing and output energy controller according to aspects of an embodiment of the present invention. FIG. 6 is a schematic block diagram illustrating a combination of short-term bandwidth control and long-term bandwidth control according to aspects of an embodiment of the present invention. 1 is a schematic block diagram illustratively showing a so-called augmentation plant of a multi-chamber laser control system according to an aspect of an embodiment of the present invention. 1 is a schematic block diagram illustrating a multi-chamber laser control system illustratively. FIG. FIG. 3 is a schematic block diagram illustrating a multi-room timing and output energy controller according to aspects of an embodiment of the present invention. A multi-chamber that alters an exemplary relationship between parameters of a multi-chamber laser control system, eg another laser system operating parameter, eg dtMOPA value, and another operating parameter, eg a bandwidth for deformation of the same parameter, eg dtMOPA value FIG. 3 is a diagram illustrating an example of an operating point curve for output energy resulting from operation of a laser system. FIG. 17 illustrates an example of operating a laser system according to a similar operating point curve, eg, for energy versus dtMOPA, as shown in FIG. 16 by a controller that advances the operating point on the curve to one operating parameter, eg, output energy maximum. It is a schematic block diagram which shows the example of a dtMOPA curve gradient estimation circuit. FIG. 6 is a schematic block diagram illustrating an illustrative example of bandwidth and laser gas control according to aspects of an embodiment of the present invention. FIG. 6 is a schematic block diagram illustrating an illustrative example of bandwidth and laser gas control according to aspects of an embodiment of the present invention. It is a figure which shows the plot of the data judged empirically regarding a duty cycle as an example. It is a figure which shows the order of a process. FIG. 4 illustrates a modification of the controller of FIG. 3 in the form of a schematic block diagram in accordance with aspects of an embodiment of the present invention. FIG. 4 illustrates in schematic block diagram form an adaptive controller portion according to an aspect of an embodiment of the present invention.

Explanation of symbols

112 MO delay estimator 114 PA delay estimator 122 MO operating point servo 124 PA operating point servo 142 MOPA servo

Claims (12)

Method for controlling bandwidth in a multi-part laser system including a first line-narrowed oscillator laser system part having a first gain generator, supplying a line-narrowing seed pulse to the second gain generator laser system part A selection of a differential firing time between generation of the seed pulse in the first laser system portion and generation of a laser gain medium in the second laser system portion is from the multi-part laser system To provide the bandwidth of the laser output light pulse;
Adjusting the target operating point as a function of the measurement bandwidth and the target bandwidth;
Estimating a current operating point by estimating a difference between a current differential firing time and a reference differential firing time, wherein the estimation is facilitated from available measurements of laser system operating parameters; Including the use of a function that is calculable or estimable and is monotonic with respect to the differential firing time over the expected operating range,
Adjusting a differential firing time as a function of the current operating point and the target operating point to advance the current operating point to a target operating point;
Including
Where the function is dE / dt estimated by applying a dither signal to the differential firing time and taking the respective ratio of the dither correlation between the laser output pulse energy and the actual differential firing time, Where E is the energy of the laser system output and t is time.
A method characterized by that. Adjusting the injection of the halogen laser gas into the chamber of the laser system as a function of the current operating point and the target operating point;
The method of claim 1 further comprising: A target operating point is determined as a function of at least one of target bandwidth, laser system duty cycle, and laser system output pulse energy;
The method according to claim 1. Method for controlling bandwidth in a multi-part laser system including a first line-narrowed oscillator laser system part having a first gain generator, supplying a line-narrowing seed pulse to the second gain generator laser system part A selection of a differential firing time between generation of the seed pulse in the first laser system portion and generation of a laser gain medium in the second laser system portion is from the multi-part laser system To provide the bandwidth of the laser output light pulse;
Adjusting the target operating point as a function of the measurement bandwidth and the target bandwidth;
Estimating a current operating point by estimating a difference between a current differential firing time and a reference differential firing time, wherein the estimation includes available measurements of laser system operating parameters Including the use of a function that can be easily calculated or estimated from and that is monotonic with respect to the differential firing time over the expected operating range;
Adjusting a differential firing time as a function of the current operating point and the target operating point to advance the current operating point to a target operating point;
Including
Here, in estimating the difference between the current differential firing time and the reference differential firing time, the derivative of the laser system output pulse energy with respect to the differential firing time at a constant voltage and the current operating point Using a function of at least one of the laser system output pulse energies, wherein the voltage is a voltage applied to an electrode of the laser system;
A method characterized by that. Method for controlling bandwidth in a multi-part laser system including a first line-narrowed oscillator laser system part having a first gain generator, supplying a line-narrowing seed pulse to the second gain generator laser system part A selection of a differential firing time between generation of the seed pulse in the first laser system portion and generation of a laser gain medium in the second laser system portion is from the multi-part laser system To provide the bandwidth of the laser output light pulse;
Adjusting the target operating point as a function of the measurement bandwidth and the target bandwidth;
Estimating a current operating point by estimating a difference between a current differential firing time and a reference differential firing time, wherein the estimation includes available measurements of laser system operating parameters Including the use of a function that can be easily calculated or estimated from and that is monotonic with respect to the differential firing time over the expected operating range;
Adjusting a differential firing time as a function of the current operating point and the target operating point to advance the current operating point to a target operating point;
Including
Here, in the step of estimating the difference between the current differential departure time and the reference differential emission time, the derivative of the laser system discharge voltage with respect to the differential emission time at a constant energy, the laser system with respect to the laser system discharge voltage Using a derivative of output pulse energy and at least one function of the laser system output pulse energy at the current operating point;
The energy is the pulse energy of the laser system output,
A method characterized by that . Estimating a derivative of the laser system discharge voltage with respect to the differential firing time at a constant energy by applying a dither signal to the differential firing time;
Applying a scaled version of the dither signal to a voltage;
Adapting the scale to minimize energy error;
Taking dV / dt as the scaling factor;
Further including
6. The method of claim 5, wherein dV / dt is a derivative of a laser system discharge voltage with respect to a differential firing time at a constant energy at a current operating point. Estimating the current operating point as a function of at least one of a derivative of a discharge voltage of the laser system with respect to a differential firing time at a constant energy and the laser system output pulse energy at the current operating point;
The method of claim 1 further comprising: Estimating the current operating point as a function of at least one of a derivative of a discharge voltage of the laser system with respect to a differential firing time at a constant energy and the laser system output pulse energy at the current operating point; , Where E is the laser system output pulse energy and dV / dt is the derivative of the laser system discharge voltage with respect to the differential firing time at a constant energy at the current operating point (1 / E) ^* Estimating as dV / dt,
The method of claim 7 further comprising: Estimating a derivative of the laser system discharge voltage with respect to the differential firing time at a constant energy by applying a dither signal to the differential firing time;
Applying a scaled version of the dither signal to a voltage;
Adapting the scale to minimize energy error;
Taking dV / dt as the scaling factor;
8. The method of claim 7, wherein dV / dt is a derivative of the discharge voltage of the laser system with respect to the differential firing time at a constant energy at the current operating point. Estimating the current operating point as a function of at least one of a derivative of the laser system output pulse energy with respect to a differential firing time at a constant voltage and the laser system output pulse energy at the current operating point;
The method of claim 1 further comprising: Estimating the current operating point as a function of at least one of a derivative of the laser system output pulse energy with respect to a differential firing time at a constant voltage and the laser system output pulse energy at the current operating point; ,
The step of estimating the current operating point includes: E is the laser system output pulse energy, and dE / dt is the laser system output pulse energy with respect to a differential firing time at a constant voltage at the current operating point. Using the relationship (1 / E) ^* dE / dt when it is a derivative,
The method of claim 10 further comprising: Adjusting the target operating point as a function of the error between the measurement bandwidth and the target bandwidth;
The method of claim 1 further comprising:

Images