June 1, 2004 / Vol. 29, No. 11 / OPTICS LETTERS 1203

Development of an amplitude-modulated Nd:YAGpulsed laser with modulation frequency

tunability up to 60 GHz by dual seed injection

Daniel C. Kao and Timothy J. Kane

Applied Research Laboratory, Pennsylvania State University, State College, Pennsylvania 16802

Linda J. Mullen

Naval Air Systems Command, Patuxent River, Patuxent River, Maryland 20653

Received November 26, 2003

A pulsed, modulation frequency tunable, frequency-doubled Nd:YAG laser has been devised for use in targetdetection through turbid media. A modulated pulse laser radar system offers many advantages in target de-tection, such as significant signal contrast enhancement, compared with conventional remote-sensing systems.By implementation of the dual-longitudinal-mode seed injection technique, the modulation frequency of thedesigned Q-switched laser can be tuned from 250 MHz up to 60 GHz in steps of 250 MHz while maintaining amodulation depth of at least 75%. This provides the ability to explore propagation and scattering propertiesfurther at previously unattainable high RF modulation frequencies. © 2004 Optical Society of America

OCIS codes: 140.3460, 140.3410.

Remote sensing through turbid media poses manychallenges for traditional laser radar sensing systems.However, by use of a RF-modulated laser pulse witha direct detection scheme, the signal-to-noise ratio(SNR) of light backscattered from a target can begreatly enhanced, thereby improving contrast andallowing for the detection of a target that could nototherwise be seen.1 – 3 One of the current drawbacksto these modulated pulse laser radar systems is theinability to explore the SNR improvement over a rangeof high modulation frequencies. Existing systems relyon static external acousto-optic and (or) electro-opticmodulators that operate at a single f ixed modulationfrequency, making them impractical for modulationfrequency tuning. These modulators can operate onlyat modulation frequencies up to a few gigahertz andhave low damage thresholds, reducing the range ofany lidar system. By use of a dual-longitudinal-modeinjection-seeding scheme,4 these restrictions can beaddressed. Through this technique, an instrumentwas developed that will allow us to examine the SNRimprovement of modulated pulse laser radar systemsthrough an entire continuous spectrum of modulationfrequencies in turbid media such as the atmosphere,the ocean, and human tissue.

Single-longitudinal-mode injection seeding has longbeen demonstrated as an effective approach to ensuresingle transverse- and longitudinal-mode operationof Q-switched Nd:YAG systems.5 By building onthis technique, and introducing an additional seedbeam, it is possible to achieve dual-longitudinal-modeinjection seeding that generates an amplitude-modulated laser pulse whose modulation frequencycan be tuned over a wide range. This is accomplishedby simultaneously locking two independent seed lasersto two different longitudinal cavity modes of the slavecavity, as well as matching the spatial eigenmodes of

0146-9592/04/111203-03$15.00/0

the seed and slave cavities. The two seeded modesare heterodyned at the detector, which results in amodulation frequency that is an integer multiple ofthe slave-cavity free spectral range. In the currentconfiguration the slave laser has a 250-MHz cavityfree spectral range, which represents the minimummodulation frequency as well as the modulation fre-quency step size. Although modulation frequenciesup to 33 GHz have been experimentally verified, thefrequency separation between the seed beams can beincreased to generate modulation frequencies up to60 GHz. The 60-GHz limit is imposed by the tuningrange of the seed lasers, and the 33-GHz frequency islimited by the detector resolution in the current setup.The design process and experimental verification willbe highlighted below.

The system consists of two master oscillators(Lightwave Laser Model 126) and a Q-switchedNd:YAG slave cavity, as depicted in Fig. 1. The twoseed beams are spatially combined through BS1, a50% beam splitter, while power is routed to a beatfrequency detection scheme. An isolator is necessarynot only to prevent backref lections of the seed beamsbut also to prevent the rejected slave pulse fromdamaging the seeds. A focusing lens �F1� collimatesthe seed beams so that the spatial eigenmode ofthe seed laser matches that of the slave. A Glanlaser polarizing cube (P) is used to reject energyduring the pulse buildup interval and also servesas the medium through which the seed lasers areinjected into the slave cavity. The plano–concaveslave cavity formed by mirrors M3 �99% R� andM4 �70%R, 10-m radius of curvature �Rc�� allows forgreat stability as well as high spatial mode rejec-tion.5 Mirror M3 is mounted on a piezoelectric (PZT)stack to provide ultrafine cavity length adjustmentsnecessary to match a seed longitudinal mode to a

© 2004 Optical Society of America

1204 OPTICS LETTERS / Vol. 29, No. 11 / June 1, 2004

Fig. 1. Dual-longitudinal-mode modulated pulse systemschematic. D, detector; A, aperture; Q, Q-switch; P, po-larizer; BS2, beam splitter. Other abbreviations definedin text.

slave longitudinal mode. An aperture placed im-mediately adjacent to M3 effectively rejects higherspatial modes while preventing backref lections thatmay stimulate the lasing of additional longitudinalmodes. Half-wave plates (WPs) placed on either sideof the YAG rod prevent spatial-hole burning and arenecessary to ensure single-mode operation.6 WP2 isa half-wave plate that matches the polarization stateof the output laser pulse and the frequency-doublingKTP crystal. A focusing lens �F2� is used to increasethe energy density to optimize doubling eff iciency.The 532- and 1064-nm light is then separated by adichotic beam splitter and collimated.

The slave oscillator was first conf igured to producea single Q-switched pulse while minimizing the lasingthreshold current and maximizing the hold off of theQ-switch and polarizer. Cavity components wereadjusted until all prelasing and postlasing pulseswere eliminated. Once this was accomplished, seed 1was tuned to the middle of a continuous tuning modeaway from mode hops and was injected into the slavecavity. The cavity length was adjusted by the PZTstack until the slave laser exhibited characteristicsindicative of being injection seeded, such as improvedprofile smoothness and a decrease in Q-switch builduptime.5 Next, seed 1 was blocked and seed 2 wasinjected into the slave cavity without altering theslave cavity length. The temperature of seed 2 wasthen tuned until it was also seeding the slave cavity.To increase the modulation frequency, we tuned thetemperature of seed 2 away from seed 1 in 250-MHzsteps until the desired modulation frequency wasachieved. For modulation frequencies below 33 GHz,seed 1 was always maintained at the same tempera-ture and served as a reference point. To optimizethe modulation depth in the output laser pulse, weequalized the power of the two seed lasers and alignedthe two beams so that they were spatially collinear.The difference frequency between seeds 1 and 2 canbe frequency locked, but this was not necessary toachieve modulation, as these lasers were inherentlystable with minimal drift. However, longer-termmodulation experiments will require active locking ofthe two seed lasers.

A fully modulated pulse at 250 MHz capturedby a Newport D-15 high-speed photodetector andoscilloscope is illustrated in Fig. 2. This detectionmethod is reliable only for modulation frequencies upto a few gigahertz because of oscilloscope samplinglimitations. To verify the modulation properties athigher frequencies, we used a Hamamatsu streakcamera (Model C5680) with a maximum temporalresolution of 1.2 ps. With this streak camera, themodulation was verif ied from 250 MHz up to 33 GHz.The difference frequency of the cw seed beams wasmonitored with a high-speed photodector and an elec-trical spectrum analyzer. This provided the meansfor coarse modulation frequency adjustment as shownin Fig. 3 and also served to verify that the pulsedmodulation frequency matched the cw seed injectedbeat frequency. Once the temperature of seed 2 wasadjusted so that the beat note stabilized on a desired

Fig. 2. Full modulation 532-nm pulse at 250 MHz, cap-tured by a high-speed photodetector.

Fig. 3. (a) Beat note of two combined independent cw seedbeams as seen on an electrical spectrum analyzer. (b) Re-sulting modulation frequency in the slave laser pulse asdetected by a streak camera.

June 1, 2004 / Vol. 29, No. 11 / OPTICS LETTERS 1205

Fig. 4. Additional modulation frequencies of (a) 10 GHz,(b) 20 GHz, and (c) 33 GHz achieved through dual-longitudinal-mode seed injection. The loss of uniformityis a function of increasing modulation frequency as a resultof streak camera sensitivity at high streak speeds.

frequency, the streak camera provided the feedbacknecessary for f ine tuning. Since the laser produces1.1-mJ pulses at 532 nm, greater than the damagethreshold of the streak camera, neutral-density f ilterswere necessary to reduce incident power levels. Byuse of this detection scheme, modulation frequenciesup to 33 GHz were experimentally verif ied. A modu-lation depth of at least 75% was always maintainedand was greatly affected by alignment, seed powerbalance, and seed injection polarization. The ESAcould not resolve seed beat frequencies above 33 GHz.Therefore, although the streak camera was not limitedto 33 GHz, the results could not be independentlyverified by the electrical spectrum analyzer.

An issue that arose with the streak camera was thetrade-off between resolution and sample length. In-creasing the resolution not only decreased the samplelength but also increased signal nonuniformities, sincethe streak camera became photon starved at highstreak speeds. High streak speeds were necessary toobtain the level of resolution necessary to resolve highmodulation pulses. To compensate for photon-starvedsignal nonuniformities at lower resolutions, the inci-dent laser power was increased until a more distinctsignal could be profiled by removing neutral-densityfilters. However, once the maximum incident powerlevels were reached on the streak camera, the nonuni-formities could no longer be avoided. Additionally, atmodulation frequencies above 20 GHz, the decreasein sample length prevented the detection of unwantedmodulation frequencies generated by the presence ofunwanted longitudinal modes. At higher frequencies,

the short sample length prevents the detection oflow frequency modes. Other possible contributors tothe nonuniformities at high modulation frequenciesinclude shot noise and streak camera amplifier (micro-channel plate) noise. For example, at 33-GHz, themaximum allowable input power was incident uponthe streak camera, and the gain of the microchannelplate was pushed to its upper limit. Only then couldthe profiles in Fig. 4 be captured. Lastly, becauseof the limited aperture size of 150 mm, the streakcamera could not provide an entire spatial profile ofthe pulse, only a selected portion. In the future, theentire spatial profile will be measured to examinethe possibility of simultaneous self-lasing of the slavecavity and its effects on modulation depth.

In summary, a system now exists that is capable ofproducing a verified modulated pulse from 250 MHzto 33 GHz with a modulation depth of at least 75%.Although the system can generate modulation fre-quencies up to 60 GHz, the upper frequencies remainunverified because of detection and instrumentationlimitations. Since the system does not rely on staticexternal modulators, and the modulation process isgenerated within the slave cavity itself, greater powerlevels and tunability are achieved in comparison withprevious modulated lidar systems. In addition togranting us the ability to investigate SNR and con-trast improvements through a continuous spectrumof high modulation frequencies, the laser systemwill also be used to verify theoretical models thatpredict modulated pulse system performance.7 Wealso intend on utilizing this laser system to explorethe improvements it can provide in atmospheric,underwater, and biomedical remote sensing.

This paper is based on work supported by the NavalAir Systems Command, Patuxent River, Maryland,under contract N00421-01-C-0223. T. J. Kane’se-mail address is tjk7@psu.edu.

References

1. V. M. Contarino, P. R. Herczfeld, and L. J. Mullen,“Modulator LIDAR system,” U.S. patent 5,822,047(October 13, 1998).

2. R. Pellen, R. Olivard, Y. Guern, J. Cairou, and J.Lotrian, J. Phys. D 34, 1122 (2001).

3. L. J. Mullen, V. M. Contarino, and P. R. Herczfeld, IEEETrans. Microwave Theory Tech. 44, 2703 (1996).

4. T. D. Raymond and A. V. Smith, IEEE J. Quantum Elec-tron. 31, 1734 (1995).

5. W. Koechner, Solid State Laser Engineering, 5th ed.(Springer-Verlag, New York, 1999), pp. 195–288.

6. A. E. Siegman and V. Evtuhov, Appl. Opt. 4, 142 (1965).7. L. J. Mullen E. Zege, I. Kotsev, and A. S. Prikhach, Proc.

SPIE 4488, 25 (2002).