Del Mar Photonics - Femtosecond autocorrelators brochure - Femtosecond autocorrelator Reef RT manual

The autocorrelation technique is the most common method used to determine laser pulse width characteristics on a femtosecond time scale.
The basic optical configuration of the autocorrelator is similar to that of an interferometer (Figure.1).

 

Figure 1.

An incoming pulse train is split into two beams of equal intensity. An adjustable optical delay is inserted into one of the arms. The two beams are then recombined within a nonlinear material (semiconductor) for two photon absorption (TPA). The incident pulses directly generate a nonlinear TPA photocurrent in the semiconductor, and the detection of this photocurrent as a function of interferometer optical delay between the interacting pulses yields the pulse autocorrelation function. The TPA process is polarization-independent and non-phasematched, simplifying alignment.
The two beams propagate in a collinear fashion (interferometric configuration). This configuration results in an autocorrelation signal that is on top of a constant background.
This background is produced by TPA photocurrent resulting from the portions of the scan during which the pulses are not overlapped.
The Reef RT autocorrelator has been specifically designed to measure the width of pulses from femtosecond lasers. For the measurement of laser pulses the only other item you need is an oscilloscope. Although not necessary, a storage oscilloscope is convenient when operating in the interferometric mode since it allows you to calibrate the display directly using the interference fringes that make up the pulse envelope.
 

KEY COMPONENTS
The Reef RT autocorrelator consists of an opto-mechanical assembly and electronics:
?Beam Splitter
?Scanning block with measuring converter
?Variable Delay Line
?Focusing Mirror
?TPA-Detector with amplifier
?Delay Sensor
?Control Unit

SPECIFICATIONS
Pulse width - 10fs?to 6ps
Input pulse repetition rate - 10kHz to CW modelocked
Scan nonlinearity - <1%
Sensitivity (P-av P-peak) - <10(mW)2
Input polarization linear - horizontal (vertical specify with order)
Wavelength range: -700 - 1000nm
(1000 ?1400nm Optional)
Scan rate variable - 0,1 - 20Hz
Detection - diode with TPA
Electrical power choices - 100-110V(220V), 50-60 Hz
Dimension: - optical unit ?170mmx134mmx105mm
- control unit ?80mmx190mmx250mm

Autocorrelator
 
Let's See What We Have Here

 Model   Product Name+   Buy Now 
 Reef-RT   Reef-RT  Buy Now 

 

WARNING! LASER SAFETY
The protective housing of this product should always in place during normal operation.
Removal of the protective housing may expose the user to unnecessary radiation, and should
be done only in accordance with specific instructions given in this manual. Be very careful
executing any step of alignment. Avoid any exposure to the direct and reflected laser beams.
Direct and reflected laser radiation can cause serious eye damage. Remember, that IR
radiation is invisible or looks like the red radiation from a low intensity He-Ne laser.
However, it is dangerous even at lowest intensity. Even diffuse reflections are hazardous.
Check all reflections and provide enclosures for beam paths whenever possible.
SAFTY CHECK LIST
?Wear protective goggles whenever possible.
?Please follow all safety instructions of your laser and use their
recommendations in your work.
?Keep all beams below eye level always. Never look in the plane of
propagation of the beams.
?When possible, maintain a high ambient light level in the laser operation area.
?Provide enclosures for beam paths whenever possible.

INSTALLATIONS AND ALIGNMENT
5.1 Reaching good alignment is easy if you use two external mirrors to direct the
laser beam into the autocorrelator. Input power should be <20mW!
5.2 Install optical unit of the autocorrelator on the optical table horizontally at
appropriate level. Use bottom screws to adjust level. Attach the unit to the
table using clamps.
5.3 The height of the autocorrelator can be changed relative to the optical table.
5.4 The polarization of the laser input to the autocorrelator must be horizontal.
(Vertical 杘ptional).
5.5 Direct laser beam through the aperture A1 onto the centre of the mirror M2
(through beam splitter BS) (Figure.2). For visible beams, this alignment can
be done visually; for infrared beams, a viewer should be used.
5.6 Move the correlator to direct reflected beam (from M2) nearly back to the
center of the aperture A1 (The beam reflected exactly back can cause
feedback in the laser). The Autocorrelator has been aligned at factory. Under
careful transportation conditions you can try to continue alignment from 5.15.
5.7 Adjust beam splitter (BS) to direct beam (reflected from M2 and BS) to the
center of aperture A2.
5.8 Direct beam reflected from M1 to the center of A2 too. If the beam reflected
from the M2 propagates far from the center of A2 or M2, you should change
the direction of the input beam with external mirrors or turn autocorrelator as
a whole. Repeat the above mentioned procedure.
5.9 Remove M3 with holder.
5.10 Place white screen at 30 - 50 cm distance from the autocorrelator and watch
spots of light reflected from M1 and M2.
5.11 Adjust BS and M1 to reach coincidence of spots from two beams on the
screen.
5.12 Watch spot movement while adjusting mirror M1 with smooth stepping of the
micrometric screw (One step <=10mkm). Interferometric fringes will appear
when arms of the autocorrelator become equal. The picture should have
7
good contrast and no more than one interferometric maximum should be
watched.
5.13 Place M3 mount into the autocorrelator.
5.14 Adjust M3 to direct beam onto the sensitive area of photodiode (PD).
5.15 Connect optical unit to terminal 揈XT?at rear panel of control unit.
5.16 For details we suggest you consult the Optical Delay Control Unit User Manual
(see Appendix A).
5.17 Connect oscilloscope input to 揂mplifier out?BNC connector of control unit.
Synchronization could be performed from 揟rig?signal (see Appendix A).
5.18 Make sure that no pump laser radiation propagates inside the autocorrelator.
5.19 Switch power on.
5.20 Set the control 揌i??揕ow?as follows: For 10?1000kHz input pulse
repetition rate (amplifiers) ?揕ow? For >1000kHz input pulse repetition rate
(generators) ?揌i? Push knobs 揂mplifier?and 揋enerator?to select mode of
operation.
5.21 Set-up oscilloscope sensitivity - 100 mV/cm scale.
5.22 To adjust retro-reflector M2 oscillation amplitude and frequency, adjust
揂mpl.?and 揊req.?with the 揋enerator?knob. You should see
autocorrelation functions on the oscilloscope screen (50-100msec/div) like
Fig.3.
5.23 To maximize signal rotate focusing screw SF. (It may be necessary to reduce
intensity of the input light to keep the signal less than 10V).
5.24 Moving retro-reflector M1 with micrometric screw find position in the middle
between two end positions (correlation functions disappear in the end
positions).
5.25 Find good position of the single autocorrelation function on oscilloscope
screen using 揇ELAY?control.
5.26 To display interferometric autocorrelation function (Figure.4) set low
frequency (揊req?2-5Hz) and big bandwidth (揊ilter?-max). The
autocorrelation function must be symmetric. If the function is observed to be
asymmetric, the autocorrelator is either misaligned relative to the input beam
or internally misaligned.
8
L O N G R A N G E S C A N N I N G ( U P T O 6 P S )
5.27 The scanning range of the M2 is adjustable to 1mm which means a time
delay >6ps.
5.28 The electrical signal (揝in? is proportional to the mechanical position of M2.
5.29 Use oscilloscope in X-Y mode to increase linearity.
----------------------------------------------------------------------------------------------
(Optional: The measuring converter of linear movements can be installed into AA-10D. An
electronic copy signal (use out 揇X? exactly proportional to the mechanical position of the
M2 allows one to operate the oscilloscope in X-Y mode. In this case linearity is better than
1% over the whole scan).

 

DETERMINATION OF PULSE
DURATION
Move the distance calibrator until arrow is in the middle of scale. Attach calibrator (MH)
to the optical plate using screw. Move M1 with screw (ST) and simultaneously measure
change of distance with help of distance calibrator and displacement of autocorrelation
function on the oscilloscope screen. Find two distances corresponding to positions of
autocorrelation function at 1/2 of the maximum. Calculate the difference DL. You can
measure the width (Dt) of autocorrelation function taking into account that 15mm measuring
distance corresponds to 100 fs. To get pulse width (Dt) you should divide Dt by 1.41
assuming Gaussian pulse shape or divide by 1.55 assuming sech2 shape.
(Roughly Dt (fs) = DL (mkm) x 4).
In measuring a pulse that has a width corresponding to <30fs, it is important to consider
possible sources of error. Due to the effect of group velocity dispersion within the BS fused
silica plate 1mm-thick the measured pulse width to be broadened by <1fs. To keep this
broadening at a minimum value compensation plate (CP) should be installed. To minimize
dispersive effects only gold mirrors are employed.

References
1. E. P. Ippen and C. V. Shank, in Ultrashort Light Pulses, S.L.Shapiro, ed. (Springer-Verlag, New York, 1977), pp.89-92.
2. J.-C. Diels and W. Rudolph, Ultrashort Laser Pulse Phenomena: Fundamentals, Techniques, and Applications on a Femtosecond Time Scale (Academic, San Diego, Calif., 1996), pp. 365-399.
3. J.-C. Diels, J. J. Fontaine, I. C. McMichael and F. Simoni, Appl. Opt. 24, 1270 (1985).
4. P. Heinz, A. Reuther, A. Lauberean, Optics Comms, 97, 35 (1993).
5. D.H. Auston, Appl. Phys. Letts, 18, 249 (1971).
6. W. Leupacher, A. Pezkofer, Appl Phys. B, 29, 263 (1982).
7. A. J. De Maria et al , IEEE, 57, No1, 2 (1969).
8. H. Schulz, H. Schuler, T. Engers and D. Von Der Linde, IEEE J.Quantum Electron. 25, 2580 (1989).
9. J.-C. Diels, J. J. Fontaine and W. Rudolph, Revue Phys Appl., 22, 1605 (1987).
10. J. G. Fujimoto, A. M. Weiner and E. P. Ippen, Appl. Phys. Lett., 44(9),832 (1984).
11. J.K.Ranka, A.L.Gaeta, A.Baltuska, M.S.Pshenichnikov, D.A. Wiersma, Optics Letters, 22, No.17, 1344 (1997).

 

 

Figure 3

Figure 4

 

Appendix A
G E N E R A L I N F O R M A T I O N
The Reef RT control unit is intended for driving moving mirror actuator and sensing mechanical
mirror position.
Control unit consists of sine generator of variable frequency, variable gain amplifier, output driver
and inductive position sensor (optional), see Figure. A1.
Output signal amplitude and frequency are controlled by an on board CPU. Parameters are
stored and recalled in/from flash memory. Generator frequency is variable in a range of 0.1...20 Hz
in 42 steps; output amplitude is variable in a range 0...+8V in 50 steps. . Two front panel rotary
knobs labeled 獳mplifier?and 獹enerator?control gain, bandwidth and output sine frequency and
amplitude respectively. Red LED digital display shows actual v alue of amplitude (0...50 in arbitrary
units) and frequency (0.1...20 in Hz). Output power driver can output up to +8 V to 10 Ohm load
and is short-circuit protected.
Generator outputs are:
?玈in?- fixed amplitude + 10V sine wave, BNC type connector, placed on a rear panel.
?玊rig?- 12V amplitude logical signal. Its phase relative to output sine can be adjusted by
?獶elay?potentiometer from a front panel.
Optional: Mechanical position sensor consists of 32 kHz generator driving fixed coil and
synchronous detector, sensing sample signal. Output value +10 V at 獶X? connector corresponds to
+ 1 mm movement. Position sensor is static and can measure any steady mirror shift.

 

 

Figure A1

 

 

E X T E R N A L C O N N E C T I O N S A N D O P E R A T I O N C O N T R O L
Check that power switch is set to off. Connect power cord to 100V 50-60Hz main outlet. Connect
measurement equipment to output BNC connectors. Switch power on. Red display shows current
values of a gain (or bandwidth) and frequency (or amplitude).
To adjust mirror oscillation amplitude, push 揋enerator?knob to set 揂mpl?mode and rotate the
knob. Right displayed value will change in a range of 0-50, corresponding to about 0-8 V output
amplitude. Set value is immediately stored in a nonvolatile memory.
To adjust mirror oscillation frequency, push 揋enerator?knob to set 獸req?mode and rotate the
knob. Right displayed value will change in a range of 0.1-20 Hz. Set value is immediately stored in a
nonvolatile memory.
To adjust the gain of signal amplifier, push 揂mplifier?knob to set the 揋ain? mode and rotate
the knob.
To adjust the bandwidth of amplifier, push 揂mplifier?knob to set the 揊ilter? mode and rotate
the knob.
Setting amplitude and frequency value; check mirror movement to not reach position limits,
disturbing parasitic oscillations.
Synchronization could be performed either from 玊rig? 揝in?or 獶X?(optional) signals.
Synchronization from first two signals is rather convenient in a case of large mirror movement range,
but may be unstable due to mechanical vibrations of mirror elastic holder. Synchronization from
獶X?signal avoids this error, but could be unstable in case of small oscillation amplitude. The best
data acquisition method consists of PC-based 2-channel ADC board collecting data array and
averaging them aft er acquisition.

 

Reef RT femtosecond autocorrelator manual

护士被强奷系列视频_黄色电影片_黄色视频网